ST2605

STORAGE

MAGAZINE

The UK’s number one in IT Storage

May/June 2026

Vol 26, Issue 3

MANAGEMENT:

Built Just for the 20%?

AI RETHINKING:

CASE STUDY:

Improved Recovery for NHS Trust

For Longterm Storage

LTO: NEW ECONOMICS:

Disk vs Tape Constraints

COMMENT - RESEARCH - INTERVIEWS - CASE STUDIES - OPINIONS - PRODUCT REVIEWS

Our Latest Innovation

LTO-10 Tape Media

TIME TO WAKE UP

Set your data preservation strategy

FLOORSPACE

REDUCTION

FASTER

ACCESS*

*hrRAO with the 40TB version

LOWER

TCO

Please contact us for more information

Sarah Taylor

Business Development Manager

Telephone: +44 (0) 7785 505992

sarah.taylor@fujifilm.com

CLICK TO

discover more

MANAGEMENT:

May/June 2026

Vol 26, Issue 3

CONTENTS

STOR

MAGAZINE

STORAGE

CONTENTS

AI RETHINKING:

CASE STUDY:

LTO: NEW ECONOMICS:

COMMENT….....................................................................4

CASE STUDY: WD CASE STUDY ON BUILDING AI AT

EXABYTE SCALE …..............................................................…6

How Dropbox is using WD's Ultrastar® DC HC690 HDD for AI

06

MANAGEMENT: WHY THE ECONOMICS OF

ENTERPRISE STORAGE NO LONGER ADD UP.................8

Nutanix contemplates if we have seen the end of predictable storage hardware costs

STRATEGY: RETHINKING LONG-TERM STORAGE IN

THE AGE OF AI......................................................................................10

Spectra Logic details why cost predictability, energy efficiency and data control are

reshaping archival strategy

10

MANAGEMENT: BUILT JUST FOR THE 20%............12

PEAK:AIO details why enterprise AI storage is failing for the majority

THOUGHT LEADERSHIP: THE MULTI-CHIP ERA:…..................16

Vertiv explores why data centre infrastructure must adapt to a new era of compute

Q&A: TAPE'S NEXT CHAPTER ………...................................….20

Fujifilm details why LTO-10 could prove to be a turning point

20

STRATEGY: THE TRUE COST OF DELAYING

BACKUP INVESTMENTS…...............................................…22

Object First explains why delaying recovery infrastructure may prove one of the most

expensive decisions organisations can make

OPINION: WHEN DISK GETS EXPENSIVE..…….........…24

Qualstar argues the case for LTO in a constrained market

26

PARTNER INSIGHT: BEYOND HONEYPOTS.......................26

How Stonesthro, Zadara, ZaveIT and Lupovis combined to power a new cyber deception

model service

31

CASE STUDY: NHS TRUST SIMPLIFIES BACKUP

OPERATIONS AND IMPROVES RECOVERY

PERFORMANCE...............................................................................28

Royal Wolverhampton NHS Trust replaces multiple legacy backup systems with a unified

ExaGrid and Veeam environment

AWARDS: THE 2026 STORAGE AWARDS FINALISTS.....31

Somehow, it's that time again....

www.storagemagazine.co.uk @STMagAndAwards May/June 2026

STORAGE

MAGAZINE

03

COMMENT:

EDITOR: Sharon Munday

editor@storagemagazine.co.uk

SUB EDITOR: Mark Lyward

mark.lyward@btc.co.uk

REVIEWS: Dave Mitchell

EDITORS COMMENT:

SHARON MUNDAY, EDITOR,

STORAGE MAGAZINE

PUBLISHER: John Jageurs

john.jageurs@btc.co.uk

LAYOUT/DESIGN: Ian Collis

ian.collis@btc.co.uk

SALES/COMMERCIAL ENQUIRIES:

Lucy Gambazza

lucy.gambazza@btc.co.uk

Stuart Leigh

stuart.leigh@btc.co.uk

MANAGING DIRECTOR: John Jageurs

john.jageurs@btc.co.uk

DISTRIBUTION/SUBSCRIPTIONS:

Christina Willis

christina.willis@btc.co.uk

PUBLISHED BY: Barrow & Thompkins

Connexions Ltd. (BTC)

Suite 2, 157 Station Road East

Oxsted. RH8 0QE

Tel: +44 (0)1689 616 000

Fax: +44 (0)1689 82 66 22

SUBSCRIPTIONS:

UK £35/year, £60/two years,

£80/three years;

Europe: £48/year, £85 two years,

£127/three years;

Rest of World: £62/year

£115/two years, £168/three years.

Single copies can be bought for £8.50

(includes postage & packaging).

Published 6 times a year.

No part of this magazine may be

reproduced without prior consent, in

writing, from the publisher.

©Copyright 2026

Barrow & Thompkins Connexions Ltd

Articles published reflect the opinions

of the authors and are not necessarily those

of the publisher or of BTC employees. While

every reasonable effort is made to ensure

that the contents of articles, editorial and

advertising are accurate no responsibility

can be accepted by the publisher or BTC for

errors, misrepresentations or any

resulting effects

DSIT have just published their latest Cyber Security Breaches Survey 2025/2026 and it's

another essential, if slightly depressing, read on the volatile digital world in which we live.

While much of the cyber security report summary still focuses on prevention, AI threats

and perimeter defence, the Government's latest findings reinforce a more uncomfortable reality:

organisations are increasingly being judged not simply on whether they are attacked, but on how

effectively they can recover, continue operating and protect critical data when incidents occur.

With 43% of UK businesses reporting a cyber breach or attack in the past year, alongside

continuing ransomware concerns, the report highlights why storage infrastructure, backup

strategy, resilience planning and recoverability are becoming central to modern cyber defence. I

would strongly encourage all our readers to download and review the full report: Cyber security

breaches survey 2025/2026 - GOV.UK

Notably, the survey also lands against a much more serious regulatory and operational

backdrop than the previous year's edition. Since the 2024/2025 report, the UK has seen a sharp

increase in "nationally significant" cyber incidents, growing concern around AI-enabled attacks

and perhaps also accelerating scrutiny linked to the proposed Cyber Security & Resilience Bill

now in committee stage. Last year DSIT highlighted cyber protection and awareness. This year's

report feels much more focused on operational continuity, recoverability and infrastructure

resilience, particularly as organisations face increasing pressure to demonstrate not just security

controls, but the ability to recover quickly and continue operating during disruption.

This ability places backup integrity, immutable storage, replication, disaster recovery and

infrastructure resilience firmly at the centre of modern business strategy. We've spent years talking

about capacity, performance and efficiency. Increasingly, in 2026, resilience seems to be the

most important metric of all.

Talking of 2026 importance, the 23rd Storage Awards are right around the corner. Voting

closes on June 4th, with the Ceremony back at the iconic De Vere Grand Connaught Rooms on

the 18th June. You'll find all this years finalists at the back of this issue. So get voting at storageawards.com.

You're back on stage!

Enjoy the issue.

Best

Sharon

Sharon Munday, Editor, Storage Magazine

04 STORAGE

MAGAZINE

May/June 2026

@STMagAndAwards

www.storagemagazine.co.uk

The future is here.

Tiered Backup Storage

• Fastest backups

• Fastest restores

• Scalability for fixed-length backup window

• Comprehensive security with ransomware recovery

• Comprehensive disaster recovery from a site disaster

• Low cost up front and over time

Thank you so much

to all who voted, and

congratulations to our fellow

Storage Awards 2025 winners!

Visit our website to learn more

about ExaGrid’s award-winning

Tiered Backup Storage.

LEARN MORE

CASE STUDY: DROPBOX STUDY:

BUILDING AI AT

EXABYTE SCALE

HOW DROPBOX IS USING WD'S

ULTRASTAR® DC HC690 HDD

TO RETHINK STORAGE DENSITY,

SCALABILITY, AND ECONOMICS

FOR THE AI ERA

The AI era has been framed as a race

for compute. But AI is fundamentally a

data system, not a compute one.

Models are only as powerful as the data

they ingest, store, and continuously access.

Leading hyperscalers like Dropbox that are

operating exabyte-scale infrastructure for

more than 700 million users already

demonstrate this reality. As AI-driven

applications scale, these data centres face

new constraints: not just processing power,

but the ability to store, manage, and retrieve

exponentially growing datasets efficiently,

cost-effectively, sustainably, and reliably.

Legacy storage approaches cannot meet

this demand at hyperscale levels. What's

needed is customer-centric, field-proven

innovation that delivers predictable

performance, high density, and cost

efficiency without compromise.

THE SOLUTION

Dropbox decided to deploy WD's Ultrastar®

DC HC690 HDD as the cornerstone for

their future-proof, scalable IT infrastructure.

Built on the world's first 11-disk platform

and delivering up to 32TB, the HC690

leverages field-proven UltraSMR technology

combined with energy-assisted PMR to

maximise areal density and unlock

significant capacity gains.

UltraSMR is already deployed in real-world

hyperscale environments and comes with a

06 STORAGE May/June 2026

@STMagAndAwards


MAGAZINE

CASE CASE STUDY: DROPBOX

proven track record of meeting customer

requirements at scale. It enables

significantly higher tracks per inch,

increasing capacity without introducing

uncertainty or operational risk.

For data-intense workloads, UltraSMR

architectures can be transformative: storing

more data per drive reduces the number of

drives required. This helps data centre

operators like Dropbox to increase storage

density in the existing storage footprint,

lower power and cooling demands, and

improve total cost of ownership - all while

maintaining enterprise-grade reliability.

physical footprint

Lower TCO: Fewer drives and racks

translate to reduced capital and

operational costs

Energy efficiency: Improved watts per

terabyte supports more sustainable data

centre growth

Proven scalability: UltraSMR is already

deployed at hyperscale, eliminating the

risk of untested alternatives

Enterprise reliability: Built on field-proven

hardware and firmware design to ensure

easy qualification, seamless integration,

and rapid adoption while maintaining

superior dependability and reliability.

Critically, the HC690 is engineered not

just for capacity, but for predictable, cloudscale

deployment like Dropbox's exabytescale

Magic Pocket. Its firmware,

architecture, and host-managed design

ensure seamless integration into modern

data pipelines and hyperscale

environments.

OUTCOMES AND BENEFITS

The HC690 delivers measurable impact

across the metrics that matter most for AI

infrastructure:

Increased data density: More capacity

per disk reduces racks, hardware, and

Dropbox's deployment of HC690 within

its exabyte-scale Magic Pocket

infrastructure demonstrates these benefits in

practice, validating UltraSMR as a

foundational technology for modern cloud

and AI storage.

As the industry shifts from compute-centric

thinking to data-centric AI infrastructure, the

Ultrastar HC690 stands out as a proven,

scalable, and economically transformative

solution. It doesn't just store data, it enables

AI at scale, today.

More Info: www.westerndigital.com


@STMagAndAwards May/June 2026

STORAGE

MAGAZINE

07

MANAGEMENT: STORAGE COSTS

WHY THE ECONOMICS OF ENTERPRISE STORAGE

NO LONGER ADD UP

STUART HEADE, EMEA SALES DIRECTOR, NUTANIX UNIFIED STORAGE (NUS), NUTANIX ARGUES THAT

WE MAY HAVE SEEN THE END OF PREDICTABLE STORAGE HARDWARE COSTS

Imagine signing off on a storage refresh in

early 2025 based on a hardware quote

that looked perfectly reasonable at the

time. But by the end of that same year, the

components underpinning that decision had

more than doubled in price. The quote itself

was not wrong. The market simply moved

faster than any procurement cycle was

designed to handle.

That scenario is playing out across enterprise

IT right now, and it is forcing a rethink that

goes well beyond how organisations buy

storage hardware. Why? Costs, demands,

supply chain issues, and new technologies are

playing havoc with your storage investments.

Let's look at some numbers to put this all

into context. NAND flash prices have risen

246% since the start of 2025. Enterprise SSD

contract prices are forecast by TrendForce, the

semiconductor research firm, to increase up

to a further 58% in a single quarter. Server

DRAM is not far behind, with prices

projected to surge more than

60% quarter-onquarter

in early

2026. These are

not

the conditions under which any organisation

should make long-term bets on proprietary

storage hardware, yet many still do.

The cause is structural rather than cyclical.

AI infrastructure has consumed global

memory and storage capacity at a pace the

market was simply not ready for, with

semiconductor manufacturers diverting wafer

production towards the high-bandwidth

memory demanded by GPU clusters. The

result is that standard enterprise components

have been crowded out. TrendForce further

projects that NAND demand in 2026 will

grow 20-22% year-on-year, while supply rises

only 15-17%.

The gap is not closing, and the bill is

landing squarely with the organisations that

have not yet rethought how they buy storage.

THE HIDDEN COST OF STAYING STILL

Proprietary storage architectures concentrate

value in the hardware itself.

Features, performance

improvements

and scalability

are tied to

appliance generations, which means the

vendor controls both the product roadmap

and, effectively, your refresh timeline. In stable

market conditions, that is just an

uncomfortable reality. In the current

conditions, it is a serious financial exposure.

The acquisition cost of storage is only part of

the picture. Support renewals on ageing

hardware routinely outpace inflation, and

organisations often find themselves paying

premium maintenance rates on systems

whose remaining business value no longer

justifies the expenditure. Add the operational

overhead of running separate silos for block,

file and object storage, and the true cost of

the traditional model becomes considerably

harder to defend than the original capital

investment suggested.

Siloed architectures also create inefficiency

that compounds over time. Capacity stranded

in one silo while another runs critically short is

common these days and is becoming more

costly to manage as workloads diversify. The

organisations now supporting AI-driven

pipelines alongside conventional enterprise

applications are finding that infrastructure

designed for a simpler era bends poorly

under the new load.

SEPARATING THE INTELLIGENCE

FROM THE IRON

Software-defined storage addresses the

problem at its root by decoupling storage

intelligence from the underlying hardware.

Data services, including scalability, resilience,

performance optimisation and security, are

delivered through software and can therefore

evolve independently of any particular

hardware generation. The organisation is no

longer locked to a vendor's refresh cycle.


@STMagAndAwards


MAGAZINE

MANAGEMENT: STORAGE COSTS

"The bill is landing squarely with the organisations that have not

yet rethought how they buy storage." Stuart Heade, Nutanix

Hardware can be sourced, scaled and

replaced according to economic logic rather

than appliance roadmaps, which restores

both flexibility and negotiating leverage.

It is worth remembering that modern

software-defined platforms are built to fully

exploit contemporary hardware, including

NVMe flash and high-speed networking

fabrics, delivering consistent, predictable

performance at scale. The decoupling is

architectural, not operational, and what

changes is who controls the value, not the

quality of the service.

The lifecycle argument is equally compelling.

Software-defined environments allow different

hardware generations to operate under a

common control plane, which means

organisations can refresh infrastructure

incrementally rather than through the

wholesale forklift upgrades that traditional

architectures typically demand.

This is a major factor for customers in

renewal cycles, especially given that

replacement hardware is both expensive and,

now with the shortages, difficult to procure

within preferred timelines. The ability to extend

the useful life of existing assets while

introducing new hardware alongside them is

becoming increasingly valuable to

infrastructure managers.

CAN THE CLOUD HELP?

When hardware prices are rising sharply

and lead times are stretching into months,

the public cloud becomes a practical

pressure valve rather than just a long-term

strategic option. The advantage of a

software-defined approach is that it makes

this genuinely straightforward.

For example, running the same software

across core datacentre, edge locations and

public cloud means organisations can

absorb demand spikes or bridge supply gaps

in the cloud without re-learning a new

environment or managing a different

operational model. The experience is

consistent wherever the workload runs, which

means the cloud stops being a separate

conversation and becomes a natural

extension of the same infrastructure strategy.

SECURITY AND AI HAVE RAISED THE

STAKES

Beyond shortages and costs, two industry

pressures are raising the stakes for

infrastructure. The first is security.

Ransomware resilience and regulatory

compliance now require capabilities that sit

within the storage layer itself: immutable

snapshots, orchestrated recovery and policybased

controls that can be audited and

demonstrated. Software-defined platforms

increasingly embed these natively, which is

more logical than retrofitting security onto

proprietary hardware not designed with

security in mind.

The second is AI. The volume of

unstructured data that AI pipelines generate

and consume is growing at a rate that

challenges fixed-capacity architectures, and

the performance consistency that model

training and inference demand is unforgiving

of infrastructure bottlenecks. Organisations

building AI capability on storage

infrastructure that cannot scale or adapt

quickly enough are discovering that the

infrastructure limitations are far more than

just a "technical inconvenience" but a

competitive liability with a measurable cost

and a roadblock to innovation.

THE REAL QUESTION HAS ALREADY

BEEN ANSWERED

The debate about whether software-defined

storage is mature enough for enterprise use

has long been settled. The question now is

whether there is any defensible argument for

continuing to anchor storage strategy to

hardware-centric models in an environment

where component prices can move 50% or

more in a single quarter, and supply gaps are

structural rather than temporary.

Software-defined storage does not make

hardware cheap. What it does is remove your

organisation's exposure to hardware volatility

as a strategic risk. When storage value

resides in software, refresh decisions are

made on your terms, not the vendor's. Costs

become predictable and defensible. In

conditions like these, that is not a minor

operational detail, it is the whole point.

More Info: www.nutanix.com



STORAGE

MAGAZINE

09

STRATEGY: SPECTRA LOGIC

RETHINKING LONG-TERM STORAGE IN THE AGE OF AI

TED OADE, DIRECTOR OF PRODUCT MARKETING, SPECTRA LOGIC, DETAILS WHY COST PREDICTABILITY,

ENERGY EFFICIENCY, AND DATA CONTROL ARE RESHAPING ARCHIVAL STRATEGY

As data volumes continue to surge,

long-term storage strategy is

becoming a real challenge for IT

teams. From AI pipelines and IoT systems to

video, analytics, and regulatory retention

requirements, organisations are generating

and preserving more data than ever before.

But while data growth accelerates, budgets,

power availability, and operational simplicity

are not keeping pace.

For years, public cloud storage has been

positioned as a flexible and scalable

solution for backup and archival. Its

accessibility and ease of deployment

made it an attractive option for

organisations looking to extend

beyond on-premises infrastructure.

However, as data footprints

expand and retention periods

stretch into decades, many

organisations are taking a

closer look at the long-term

implications of their storage

decisions. Cost predictability,

data control, energy consumption,

and performance are all becoming

increasingly important

considerations in designing

sustainable archival strategies.

Advances in object-based

storage are reshaping

how traditional

media

can be integrated into modern architectures.

By enabling archival storage to operate

within an object storage framework,

organisations can better align long-term

data retention with the same interfaces and

workflows used across cloud and

application environments. To assist, we've

mapped out six key considerations shaping

how IT teams are rethinking long-term

storage and archival strategies:

1. Data Growth Is Outpacing Traditional

Storage Models

The scale of modern data environments is

unprecedented. By the end of the decade,

global secondary storage demand is

expected to reach zettabyte levels, driven by

AI, machine-generated data, and long-term

retention requirements.

Traditional file-based systems often

struggle at this scale, introducing complexity

in how data is organised, discovered, and

accessed. Object storage offers a more

scalable alternative by decoupling data

from hierarchical structures and enabling

metadata-driven access.

When applied to long-term archival

environments, this approach allows

organisations to manage vast datasets more

efficiently while simplifying how data is

indexed, searched, and retrieved over time.

2. Not All Data Requires Always-On

Infrastructure

A significant portion of enterprise data

becomes "cold" shortly after it is created.

Backup copies, compliance

archives, historical datasets, and

intermediate AI data may be

retained for years but accessed

infrequently. Despite this, much


@STMagAndAwards


MAGAZINE

STRATEGY: SPECTRA LOGIC

"The most effective long-term storage strategies are those designed not just for today's

workloads, but for decades of continued data growth."

of this data continues to reside on alwayson

infrastructure, consuming power and

increasing operational overhead. As data

volumes grow, this model becomes

increasingly inefficient.

Tiering cold data to lower-power

storage solutions allows organisations to

reduce energy consumption, ease

pressure on data centre resources, and

support more sustainable infrastructure

growth without compromising long-term

durability or accessibility.

3. Long-Term Storage Economics Require

Predictability

Public cloud storage offers flexibility and

scalability, but long-term archival use

cases introduce additional considerations.

As data volumes increase, organisations

must account for the full lifecycle cost of

storing and retrieving data over time.

Complex pricing models that include

charges for data retrieval, API requests,

replication, and egress can make it

difficult to forecast costs accurately.

Industry research shows that many

organisations exceed their planned cloud

storage budgets as usage scales.

For long-term archives, predictability

becomes just as important as scalability.

Storage strategies that provide stable,

transparent cost structures can help

organisations better manage budgets and

avoid unexpected expenses over time.

4. Access Patterns Should Drive Storage

Design

Not all archived data is equal. Some

datasets may need to be accessed quickly

for analytics, compliance audits, or model

retraining, while others are rarely retrieved

but must be preserved for extended periods.

Lower-cost storage tiers can introduce

delays or additional steps when retrieving

data, which may impact workflows if not

properly aligned with access requirements.

Designing storage architectures based on

actual data access patterns allows

organisations to balance performance and

cost more effectively, ensuring that

frequently accessed data remains readily

available while long-term archives are

stored efficiently.

5. Security and Resilience Demand a

Multi-Layered Approach

As cyber threats continue to evolve,

protecting backup and archival data has

become a top priority. Ransomware and

other attacks increasingly target data

repositories, making recoverability just as

important as storage.

Incorporating isolated or offline storage

layers can significantly enhance resilience

by reducing exposure to network-based

threats. This approach helps ensure that

critical data remains protected and

recoverable, even in the event of a breach.

A hybrid storage strategy that combines

multiple storage types can strengthen

overall data protection while maintaining

operational flexibility.

6. Data Sovereignty and Long-Term

Control Are Critical

As regulatory requirements continue to

expand globally, organisations must

maintain greater visibility and control

over where their data resides and how it

is managed.

Public cloud environments may distribute

data across multiple geographic regions,

which can introduce challenges related to

compliance, governance, and data

residency. Regulations such as GDPR,

HIPAA, and industry-specific mandates

further increase the need for transparency.

Maintaining control over long-term

archives supports compliance efforts

while also enabling organisations to

implement best practices such as

geographic redundancy and multi-tiered

storage strategies.

THE BOTTOM LINE

The continued rapid growth of data is

reshaping how organisations approach

long-term storage and archival. While

public cloud remains an important part of

modern infrastructure, it is not always

sufficient on its own to meet the demands

of large-scale, long-term data retention.

Emerging approaches that combine

object-based storage frameworks with a

mix of storage media are enabling

organisations to better balance

scalability, cost, performance, and

control. By aligning storage strategies

with data access patterns and long-term

retention requirements, IT teams can

reduce complexity while improving

resilience and sustainability.

As data volumes continue to expand and

infrastructure demands increase, the most

effective storage strategies will be those

that take a balanced, hybrid approach

designed not just for today's workloads,

but for the decades of continued data

growth ahead.

More Info: https://spectralogic.com/



STORAGE

MAGAZINE

11

MANAGEMENT: AI STORAGE

BUILT JUST FOR THE 20%: WHY ENTERPRISE AI

STORAGE IS FAILING THE MAJORITY

ROGER CUMMINGS, PRESIDENT AND CEO OF PEAK:AIO WARNS THAT HYPERSCALE-FIRST STORAGE

ARCHITECTURES ARE STRUCTURALLY EXCLUDING THE 80% OF ORGANISATIONS DRIVING TODAY'S

INFERENCE-LED AI WORKLOADS

When a well-known research

institution received its first GPU

cluster last year, it faced an

unexpected bottleneck before it even started

using it: storage. As they talked to multiple

enterprise AI storage vendors, they were

quoted solutions starting at $400,000 to

$650,000 before they had connected a

single GPU. The maths simply didn't add up.

With inference workloads now

accounting for roughly two-thirds of all AI

compute, a clear structural gap is

emerging. While many enterprise vendors

have built exceptional products for largescale

training clusters, their architectures

can't economically serve the estimated

80% of organisations running smaller

inference workloads and cost-constrained

training deployments.

This isn't about product quality or features.

It's a market segmentation problem hiding in

plain sight, one that has existed for years,

but that the surge in inference demand is

finally bringing to the forefront.

THE 80/20 SPLIT

Training workloads currently account for

approximately 20% of GPU deployments but

have historically dominated vendors'

attention. These large clusters (128+ GPUs)

run synchronised training jobs demanding

massive bandwidth and exabyte-scale

capacity from day one, justifying $400,000-

$650,000 storage minimums. Companies

like WEKA and VAST Data excel here, with

proven hyperscale architectures.

The problem emerging in the market is that

inference workloads now account for 60-

70% of total GPU compute hours, up from

33% in 2023. Unlike training, inference

workloads are highly distributed; individual

tasks can be handled independently and

typically run on smaller configurations (up to

128 GPUs). It is estimated that by 2030,

approximately 70% of all data centre

demand will come from AI inference.

A utilisation paradox compounds this gap.

Current industry data show that only 7% of

organisations achieve more than 85% GPU

utilisation, with most stuck at 51-70%,

wasting $960,000-$1.6 million in a typical

64-B200 cluster. Many organisations need

better storage to improve utilisation but can't

justify $400,000-$650,000 without first

proving they can drive up utilisation. It's a

chicken-and-egg problem that locks out a

major portion of the market.

THIS IS ABOUT ARCHITECTURE,

NOT PRICING

These minimum deployment requirements

aren't arbitrary; they're architectural

constraints stemming from platforms

originally designed for workloads

different from those driven by modern AI.

For example:

WEKA requires a minimum of 8 servers

(6 in the cloud) for "optimal

performance, resilience, and scalability".

At ~$50,000 per server, that's a

$400,000+ entry point.

VAST's EBox requires a minimum of 8

nodes, with documentation noting that

fewer than 12 EBoxes reduce maximum

write performance; 8-10 EBoxes yield

approximately 15% less write bandwidth.

Their recommendation: start with 12 or

more EBoxes. At 13 servers, the minimum

entry point approaches $650,000.

These minimums exist for valid technical

reasons: erasure coding, distributed

metadata, and performance optimisation all

require a critical mass, but they create an

insurmountable economic barrier for what

now accounts for 80% of deployments.

Deployment timelines compound the

problem further. Traditional enterprise

storage takes 6-10 weeks to go from

purchase to production. When every GPUhour

costs more and lead times stretch to a

year, utilisation becomes the defining metric,

and slow storage deployment only makes

that worse.


@STMagAndAwards


MAGAZINE

Broadcom changed the rules.

Time

to change

your stack

Get a Free

with StorPool Storage

Ultra-Fast

Performance

Sub-100µs in-VM latency

and millions of IOPS, even at

high load and high capacity

utilization on standard

servers.

Predictable

Cost

Move to a transparent pricing

model and pay only for what

you need. Up to 67% lower

total cost vs VMware Cloud

Foundation.

Operational

Simplicity

StorPool's experts handle

design, deployment, upgrades,

and 24/7 monitoring so your

team stays focused on the

business.

StorPool offers two enterprise-grade

paths for replacing VMware.

Both come with no vendor lock-in and no migration headaches.

StorPool One

The complete turnkey KVM cloud platform

integrated and fully managed by StorPool.

Ideal for organizations that want to move fast

and remove operational complexity.

StorPool with

Oracle Virtualization

Enterprise-grade KVM with Oracle Linux

Virtualization Manager. Ideal for

organizations that need commercial-vendor

support and Oracle ecosystem alignment.

info@storpool.com

MANAGEMENT: AI STORAGE

THE INFERENCE MISMATCH

The shift to inference-centric infrastructure

is well documented but poorly addressed.

Deloitte projects inference spending will

reach $20.6 billion in 2026, up from

$9.2 billion in 2025. Gartner forecasts

65% of AI-optimised IaaS spending will

support inference by 2029. Yet most

infrastructure was optimised for training:

long, compute-heavy jobs in large,

centralised facilities. The divergence is

significant: training demands 100-200+

kW per rack and specialised

interconnects; inference requires 30-150

kW per rack, real-time responsiveness,

and distributed deployment.

Inference workloads don't need exabytescale

on day one. They need fast

deployment, hardware flexibility, and the

ability to start small and scale linearly. True

software-defined architectures running on

the commodity hardware of your choice

should deploy in hours and start at one

server rather than 8-13.

A DIFFERENT APPROACH

The requirements to meet the needs of the

underserved majority are clear: eliminate

minimum server requirements, enable

hardware flexibility beyond specific certified

configurations; enable rapid deployment

measured in hours rather than weeks; and

provide utilisation-first economics.

The research institution mentioned at the

outset is instructive here. By deploying on

existing white-box servers, they achieved a

5× increase in throughput, avoided a

$1M+ hardware refresh, and were up and

running within hours rather than days or

weeks. This isn't superior technology for its

own sake; it's the appropriate technology

for a different market segment.

As inference dominates compute hours

and GPU scarcity makes utilisation the

defining metric, the vendors that succeed

won't necessarily be those with the most

exabytes under management. They will be

the ones who democratise access to AI

storage infrastructure for the silent majority.

SERVING THE MARKET WE NEED

The enterprise AI storage market isn't

broken; it's incomplete. WEKA and VAST

Data have built remarkable platforms for

the 20% who can justify a $400,000-

$650,000 minimum investment. The

problem is that the other 80%,

organisations deploying inference

workloads on 8-128 GPUs, are stuck in the

utilisation paradox, unable to justify

certified hardware premiums and

structurally underserved.

The question for the storage industry is

whether we build for the market we want to

serve, where the headlines and multibillion-dollar

deals are, or the market we

need to serve.

The answer will determine which vendors

remain relevant as inference accounts for

65% of AI infrastructure spending by 2029.

More Info: https://www.peakaio.com


@STMagAndAwards


MAGAZINE

ST P

Running AI on data

you can’t trust.

ST RT

Processing from governed,

AI-ready data.

CTERA unifies unstructured enterprise files into a trusted, governed data fabric

that powers agentic AI — delivering consistent document processing where every

stage of the pipeline operates on the same data without copying or replication.

LEARN MORE

THOUGHT LEADERSHIP: VERTIV LEADERSHIP:

THE MULTI-CHIP ERA: WHY DATA CENTRE

INFRASTRUCTURES MUST ADAPT FAST

JON ABBOTT, TECHNOLOGIES DIRECTOR OF GLOBAL STRATEGIC CLIENTS AT VERTIV WARNS THAT

THE RAPID DIVERSIFICATION OF AI SILICON IS FORCING A FUNDAMENTAL RETHINK OF POWER

DENSITY, COOLING STRATEGIES, AND LONG-TERM INFRASTRUCTURE DESIGN

As data centre operators deploy a

broader mix of graphics processing

units (GPUs), custom applicationspecific

integrated circuits (ASICs),

inference accelerators, and traditional

central processing units (CPUs),

infrastructure planning must further

evolve towards more flexible, futureready

systems capable of handling rapid

silicon innovation.

THE SHIFT TO DIVERSIFICATION

Just as the automotive industry is moving

beyond petrol engines to embrace electric

vehicles, hybrids, and alternative fuels,

data centres are experiencing a parallel

transformation. Traditional facilities built

around x86 CPU servers for generalpurpose

workloads still command the

largest market share. Yet the explosive

growth of AI is driving rapid diversification

in the silicon powering modern AI

factories.

Hyperscalers, cloud providers, large

enterprises, and colocation operators are

deploying AI servers that combine multiple

processor types to optimise everything

from massive model training to highvolume

inference. This silicon

diversification demands a fresh approach

to power delivery, cooling, rack design,

and overall facility architecture, balancing

performance, efficiency, cost, and longterm

adaptability.

THE EXPANDING LANDSCAPE OF AI

ACCELERATORS

Several distinct categories of processors

are now shaping AI infrastructure

decisions:

GPUs remain the performance leader -

they continue to deliver industryleading

parallel processing power for

the most demanding AI workloads,

particularly large language model

(LLM) training. Their high core counts,

substantial memory bandwidth, and

versatility make them equally capable

for inference once models move into

production. Major hyperscalers and

the largest enterprises lead adoption,

though persistent supply constraints

and high costs have prompted some

organisations to explore alternatives.

Custom silicon and ASICs are gaining

momentum - leading cloud providers

have leveraged ARM-based

architectures to develop their own

ASICs. Google's Tensor Processing

Units (TPUs) emphasise strong

performance per watt, while Amazon

Web Services continues to scale its

Trainium and Inferentia families.

Microsoft has advanced its Maia series

and Meta has expanded its MTIA

inference accelerators as part of a

broader industry move towards

workload-optimised silicon. Even

OpenAI has partnered on custom

silicon development. These chips can

deliver superior efficiency for specific

workloads while reducing reliance on

general-purpose accelerators.

Specialised inference chips - these

represent a huge and rapidly growing

market and, in many deployments,

may exceed training volumes. Because

inference workloads are typically less

computationally intensive, dedicated

inference processors can deliver

energy efficiency without the full power

draw of high-end GPUs. Qualcomm

has entered this space with its AI200

and AI250 chips, targeting enterprise

deployments with a focus on

performance per watt. Numerous startups

are also bringing ASIC-based

inference solutions to market, enabling

operators to handle diverse AI

applications more cost-effectively.

The enduring role of CPUs - traditional

CPUs have not been sidelined. In AI

servers, they still manage

orchestration, networking, storage I/O,

and system-level operations, while

accelerators handle the heavy

computational lifting. For lighter or

smaller-scale inference tasks, modern

CPUs can deliver adequate


@STMagAndAwards


MAGAZINE

EUROPEAN DATASOVEREIGNTY

YOURRISKFREE

SOVEREIGNTY

JOURNEY

TigerTechnologyistheonlyEuropean-based

datamanagementsolutiontoguaranteea

seamlessandstress-freesovereigntytransition.

FULLCONTROL

EveryEuropean

S3CloudSupported

Zero

DataLock-In

Minimal

DowntimeMigration

RequestYourFulSovereignty

AssessmentToday

www.tiger-technology.com

THOUGHT LEADERSHIP: VERTIV LEADERSHIP:

performance,

avoiding the need for

specialised hardware in every

deployment.

Emerging architectures - wafer-scale

processors, such as those from

Cerebras, already support both

training and inference at massive

scale. Looking further ahead,

neuromorphic and quantum

computing approaches could

eventually influence data centre

design, though their practical impact

remains longer term.

POWER AND COOLING

The move towards heterogeneous silicon

directly affects power and thermal

management. AI factories often feature

extremely dense GPU clusters, while many

enterprise and colocation operators are

retrofitting existing halls to support mixed

workloads. Rack power densities have

also risen dramatically.

This escalation strains conventional aircooling

limits and power delivery systems.

Higher-density racks require upgraded

power distribution

units (PDUs), highervoltage

architectures, and often

direct current (DC) power distribution

for improved efficiency.

Cooling strategies have evolved

accordingly. Options now include:

Direct-to-chip liquid cooling

Immersion cooling

Hybrid air-liquid systems

The choice depends heavily on the

silicon mix. Low-power inference chips

from certain vendors allow operators to

retain much of their existing air-based

infrastructure. In contrast, dense GPU or

high-volume ASIC deployments often

necessitate liquid cooling from day one,

along with robust heat reuse strategies to

improve overall facility efficiency.

STRATEGIC APPROACHES TO

SILICON DIVERSIFICATION

Data centre operators have two options:

1. High specialisation - tailor power and

cooling infrastructure to specific server

configurations and chip types. This can

optimise short-term performance and

efficiency but risks reduced flexibility as

silicon roadmaps evolve.

2. Future-ready design - build modular,

scalable infrastructure with a 10+ year

facility lifespan in mind, even though

individual AI servers and chips may be

refreshed every two to four years. This

approach favours flexible power systems,

rack designs that support both air and

liquid cooling, and standardised

interfaces that accommodate

heterogeneous compute.

A diverse silicon strategy also influences

storage infrastructure. AI workloads

generate and consume vast amounts of

data, placing new demands on highbandwidth,

low-latency storage fabrics

that must integrate seamlessly with mixed

processor environments.

PREPARING FOR A MULTI-SILICON

FUTURE

GPUs are expected to retain dominance for

large-scale training and many highperformance

inference tasks. However,

custom ASICs and energy-efficient

inference chips are rapidly carving out

significant roles for cost-sensitive or powerconstrained

deployments. CPUs will

continue providing the foundational control

layer across AI systems.

Operators that invest in flexible power and

cooling architectures today will be better

positioned to adopt whatever mix of

processors proves most effective tomorrow.

By treating infrastructure as a strategic

enabler rather than a static backdrop, data

centre teams can turn silicon diversification

from a challenge into a competitive

advantage in the AI era.

More Info: www.vertiv.com


@STMagAndAwards


MAGAZINE

Stadium MK Transforms Crowd Management and

Security with Synology’s AI-Powered Solutions

The Challenge

Stadium MK relied on manual handheld

clickers at each gate to count guests. This was

not only time-consuming, but required

extensive training, increased the risk of human

error and also led to high labor costs.

The Benefits

Enhanced accuracy in guest trac king and crowd

management

Real-time insights into crowd flow with live data

monitoring

Scalable AI-driven solution without additional

hardware costs

The Solution

Stadium MK deployed Synology TC500 Al-powered cameras across all entry points, connected to a Synology DVA

Deep Learning NVR. This intelligent solution provides the operational and security team with real-time, accurate guest

counting and crowd flow insights to improve operational efficiency and safety.

Batch configuration Multi-gate coverage Real-time dashboards Reporting & analytics

Efficiently configure and

Accurate counting

Live occupancy and flow

Comprehensive reports to

deploy multiple cameras

across multiple

visuals for proactive

evaluate performance and

at scale

gates and entry points

operational decisions

plan for future events

About Stadium MK

Stadium MK is a state-of-the-art venue with a

seating capacity of 30,000+, hosting football

matches, concerts, conferences, and a wide

range of live events year-round.

Discuss your surveillance needs with Synology

Discover how Synology solutions can help enhance safety

and improve efficiency in your site

Learn more

Q&A: FUJIFILM FUJIFILM

WHEN DISK GETS EXPENSIVE: THE CASE FOR LTO

TAPE IN A CONSTRAINED MARKET

AS AI ACCELERATES DATA CREATION AND STORAGE COSTS CONTINUE TO CLIMB, TAPE IS FINDING

RENEWED RELEVANCE. RICHARD ALDERSON, EMEA SALES & MARKETING MANAGER AT FUJIFILM,

DISCUSSES WHY

Tape may be one of storage's oldest

technologies, but it is far from

standing still. As LTO-10 comes to

fruition and organisations reassess longterm

retention strategies, Storage

Magazine catches up with Richard

Alderson, EMEA Sales & Marketing

Manager at Fujifilm, to discuss why

demand continues to grow.

Storage Magazine: Many people still

associate tape with legacy infrastructure,

yet LTO continues to evolve. What does

the launch of LTO-10 tell us and why do

you believe this generation could prove to

be a turning point?

Richard Alderson: Tape demand

continues to grow significantly, particularly

as organisations rethink how they manage

rising data volumes and long-term

retention requirements. Looking at the

latest figures, more than 70 exabytes of

tape capacity were delivered globally in

2024, excluding hyperscale organisations,

representing year-on-year growth of

approximately 15%.

Much of this demand is being driven by

a combination of factors, including AIgenerated

data growth, increasing storage

requirements, and the need for more

cost-effective long-term preservation

strategies. Industry forecasts also

suggest tape demand could

continue to rise strongly over

the coming years, with some

projections estimating

compound annual growth of

between 30% and 40%

between fiscal years 2024

and 2027.

Against this backdrop,

the arrival of LTO-10

marks an important step

forward for tape

technology. Fujifilm now

has LTO-10 with

capacities of 30TB and

40TB, helping organisations address

density and floor space challenges.

The technology also introduces

advancements such as hrRAO (High

Resolution Recommended Access Order),

enabling faster access to archived data,

while enhancements to the tape base film

have been designed to improve durability

and support lower energy consumption

compared with continuously powered

storage environments.

Storage Magazine: What are the biggest

advantages tape still offers in a modern

storage environment, particularly as

organisations reassess cost, resilience and

sustainability?

Richard Alderson: Tape continues to offer

three core advantages: security, longevity

and cost efficiency.

From a cyber resilience perspective, tape

remains particularly valuable because of

its air-gapped nature, helping

organisations strengthen protection

against ransomware and other cyber

threats. For many businesses, offline

storage remains an important component

of broader recovery and resilience

strategies.

Longevity is another important

consideration. Tape media is designed for

long-term archival, with the ability to

preserve data for decades, making it well

suited to organisations with compliance,

governance or historical retention

requirements.


@STMagAndAwards


MAGAZINE

Q&A: Q&A: FUJIFILM

"The challenge is no longer generating data, it is storing it economically,

securely and sustainably for the long term." Richard Alderson, Fujifilm

Not all of this information needs to sit on

expensive, always-on primary storage.

As businesses generate and retain larger

data volumes, long-term preservation

strategies are becoming more important.

Tape provides a highly scalable and secure

option for storing data that must be

retained but is not regularly accessed,

helping organisations balance accessibility

with economics.

For this reason, we expect AI to become a

major driver of future tape demand.

There is also growing interest in tape

from a sustainability perspective. For

archival workloads, tape can offer lower

energy consumption than continuously

powered disk-based alternatives, while

high-capacity cartridges reduce physical

storage demands.

Storage Magazine: Who is investing in

tape today, and what business challenges

is it helping organisations solve?

Richard Alderson: Tape remains

particularly relevant for organisations

managing large volumes of infrequently

accessed data that must be retained

securely over long periods.

Industries including scientific research,

media and entertainment, financial services

and cloud providers continue to make

extensive use of tape technology, often

because of the sheer scale of data

involved.

Security is naturally a major

consideration, particularly as cyber threats

continue to evolve. However, organisations

are also facing increasing pressure from

rising energy costs, ongoing storage

expansion, floor space limitations and

fluctuations in hard disk pricing and

availability.

In many of these scenarios, tape provides

an effective long-term archival layer within

a broader storage strategy.

Storage Magazine: AI is dramatically

increasing data volumes. What impact do

you expect this to have on long-term

storage strategies, and where does tape

fit in?

Richard Alderson: AI is undoubtedly

accelerating data creation at an

unprecedented scale. But beyond model

training, organisations are also

increasingly recognising the importance of

retaining datasets for governance,

compliance, retraining and historical

analysis.

Storage Magazine: Looking ahead ten

years, what role do you think tape will play

in storage, and how far can the technology

still evolve?

Richard Alderson: Tape technology has

already evolved considerably, and there is

still significant headroom for innovation.

Back in 2020, Fujifilm demonstrated a

record capacity of 580TB on a single

cartridge under laboratory conditions,

highlighting the long-term potential of tape

density improvements.

Today, LTO-10 already delivers almost

double the native capacity of LTO-9, with

up to 40TB per cartridge. Future

developments are likely to focus on

increasing capacity further, improving

efficiency and enabling organisations to

manage growing archival requirements

more effectively.

As enterprise data volumes continue to

expand, tape is likely to remain an

important part of long-term storage

strategies for many years to come.

More Info: http://www.fujifilm.com/uk



STORAGE

MAGAZINE

21

STRATEGY:

STRATEGY: BACKUP INVESTMENT

THE TRUE COST OF DELAYING

BACKUP INVESTMENTS

AI-DRIVEN DEMAND, SUPPLY CHAIN DISRUPTION

AND TIGHTER IT BUDGETS ARE PUTTING PRESSURE

ON STORAGE INVESTMENT DECISIONS. YET AS

RANSOMWARE INCREASINGLY TARGETS BACKUP

ENVIRONMENTS, PETE HANNAH, VP SALES

WESTERN EUROPE, OBJECT FIRST EXPLAINS WHY

DELAYING RECOVERY INFRASTRUCTURE MAY

PROVE ONE OF THE MOST EXPENSIVE DECISIONS

ORGANISATIONS CAN MAKE


@STMagAndAwards


MAGAZINE

STRATEGY:

STRATEGY: BACKUP INVESTMENT

"Supply chain volatility won't slow down anytime soon. But

delaying backup investment is a false economy when it

compromises your security strategy." Pete Hannah, Object First

Let's address the elephant in the room:

prices are going up across the tech

industry and yet budgets aren't keeping

pace. The AI revolution is driving

unprecedented demand for hardware, while

geopolitical instability is adding to supply

chain uncertainty. Analysts are clear that

pricing pressure and supply constraints will

continue for the foreseeable future: industry

research points to NAND and DRAM

shortages extending through 2027, with

project delays and overruns likely as

traditional procurement cycles are impacted.

This price pressure is forcing leaders to

make tough decisions, including delaying

some investments in the hope that prices

stabilise and delivery times regain their

predictability. However, it's critical to

recognise that even if reluctantly accepting

delay is a viable strategy in some areas, it

increases risk and potential costs in

others. One important example is

cybersecurity investment.

The threat of cyberattacks, primarily

ransomware, continues to rise. Enterprise

Strategy Group (ESG) research found that

66% of organisations reported at least

one ransomware attack in the last two

years, and 96% of those attacks targeted

backups to prevent recovery and force a

ransom payment. And what did 81% of

IT professionals identify as their best

defence against ransomware? Immutable

backup storage.

So now is not the time to delay investment

in backup storage as a foundation for your

security strategy. Recent Veeam research

(Data Trust and Resilience Report published

in April 2026) shows businesses that

increased their cybersecurity investment

over the past year had measurably better

recovery outcomes after an attack, in terms

of not paying ransoms and recovering all

their data, than those that kept spending

flat or reduced. But how can businesses

achieve ransomware resilience based on

immutable backup architecture, despite the

challenging pricing and supply chain

pressures?

Optimisation is essential. Conventional

CapEx purchase models require upfront

investment based on long-term usage

assessments. Inaccuracies or sudden

changes in requirements can put

businesses at risk of over- or underprovisioning,

leading to wasted spend or

operational difficulties.

Consumption-based subscription models

are emerging as a practical, flexible

response to this problem. They allow

immediate access to secure backup

storage solutions, thus providing a reliable

route to recovery after a cyberattack, while

side-stepping CapEx challenges and

budget cycles.

That choice also extends to vendor

strategy. While shortages and extended

lead times are real, they are not universal.

Backup vendors that have invested in

forward supply planning, tighter

component forecasting, and dedicated

capacity can shield customers from the

worst of the supply chain squeeze. That

translates into predictable delivery timelines

and less pressure to compromise on

security posture simply to secure hardware.

Supply chain volatility won't slow down

anytime soon. But delaying backup

investment is a false economy when it

compromises your security strategy.

Growing ransomware threats, and the

demonstrable recovery advantages of

immutable backups, require decisive

action now. By choosing resilient backup

storage through flexible consumption

models, and by working with vendors that

can still deliver predictably despite market

disruption, organisations can secure

recovery readiness without losing control

of costs.

More Info: objectfirst.com



STORAGE

MAGAZINE

23

OPINION: LTO

WHEN DISK GETS EXPENSIVE: THE CASE FOR LTO IN

A CONSTRAINED MARKET

DOUG WILLIAMS, EMEA CHANNEL MANAGER AT QUALSTAR, EXPLORES HOW PRICING PRESSURE AND

SUPPLY CONSTRAINTS ARE RESHAPING THE ROLE OF TAPE IN ENTERPRISE STORAGE PLANNING

The supply side of the storage market

has quietly inverted. Early in 2026,

Western Digital CEO Irving Tan

confirmed that its entire HDD production for

2026 had been sold through, with longterm

agreements now extending into 2028.

Market tracker TrendForce recorded HDD

contract prices climbing roughly 4% quarter

on quarter through Q4 2025, the sharpest

sustained move in eight quarters. Enterprise

SSD pricing has moved further still. The

same market tracker forecast enterprise SSD

contract prices rising on average 58%

quarter on quarter in Q1 2026, with NAND

flash contract prices projected to rise up to

75% in Q2 2026. Pricing data reported by

Tom's Hardware shows a 30TB TLC

enterprise drive moving from roughly

$3,000 in Q2 2025 to $11,000 by early

2026, a 257% increase in nine months.

For IT leaders working on annual

planning, the takeaway is practical. The

primary storage tier has become materially

more expensive to refresh. Buyers who

approved a disk expansion in early 2024

are now budgeting for the same capacity at

a premium, (assuming they can secure

allocation at all.) At the same time, data

volumes continue to compound under AI

training data, surveillance video, IoT

telemetry, and retention schedules that now

stretch up to ten years or more.

This is where the operational question

becomes economic. Komprise's 2024 State

of Unstructured Data Management report

found that on average 70% of enterprise

data is cold and has not been accessed in

over a year. Others place the figure closer

to 80% once archive tiers are included. That

inventory is paying full freight for

performance it is not using.

A sensible response, and the one the

market is beginning to take seriously again,

is to move cold data off primary tiers and

onto modern LTO. For organisations

running the 2026 refresh

maths, the cartridge

becomes the most

useful tool in the

room. The LTO

Program specifies that

LTO-10, shipping since

August 2025, holds 30

TB natively and up to 75 TB with

compression, at a native transfer rate of

400 MB per second. Pricing places raw LTO

media at around $5 to $8 per terabyte, with

fully loaded archive TCO between $7 and

$15 per terabyte per year. That compares

with $60 to $150 per terabyte per year for

active enterprise disk. Cloud archive sits

between those two numbers on published

rate cards from AWS S3 Glacier Deep

Archive at $0.00099 per gigabyte per

month, with Azure Archive and Google

Cloud Archive at $0.0024 per gigabyte per

month (all before egress fees of $80 to

$120 per terabyte applied at retrieval).

Those numbers lead to a different planning

conversation. Rather than paying to grow

the primary tier at 2026 prices, many

operators are sweating their existing disk

estate for another two years and placing the

incremental petabytes on tape. This is the

quiet strategy the hyperscalers have followed

for years. AWS, Azure, and Google have

not retired their tape libraries. IEEE

Spectrum has documented that Microsoft,

Amazon, OVH, and Baidu all deploy tape at

hyperscale, which is the only way the subcent

per gigabyte monthly pricing for

Glacier Deep Archive, Azure Archive, and

Google Archive can remain profitable.

When the largest cloud buyers in the world

run tape underneath their archive services,

the case for enterprises running tape on

premises becomes harder to dismiss.

THE OFFLINE COPY AND THE

RANSOMWARE ECONOMY

In parallel the security argument has

strengthened. Veeam's 2024 Ransomware


@STMagAndAwards


MAGAZINE

OPINION: LTO

"When the largest cloud buyers in the world run tape underneath their

archive services, the case for enterprises running tape on premises becomes

harder to dismiss." Doug Williams, Qualstar

gap between primary and archive

performance remains substantial. The shift is

about right-sizing. 60 to 90% of data that

does not need to be instantly available

should not be sitting on the most expensive

tier in the stack, particularly when that tier is

the one currently being squeezed by AIdriven

supply constraints.

Trends Report, who interviewed more than

1,300 organisations that suffered attacks,

found that backup repositories were targeted

in 96% of successful attacks. Disk, flash, and

cloud systems share a common vulnerability.

They are network addressable. An attacker

with sufficient privilege can reach them. A tape

cartridge that has been written, ejected, and

placed in a slot or on a shelf cannot be

reached over any network.

That physical disconnection, often described as

an air gap, remains the only storage state that is

provably unreachable by a remote adversary.

Marsh's cyber risk practice has called out

immutable and offline backups as a control that

materially improves both recovery outcomes

and underwriting terms, and regulatory

guidance, including the European Union's NIS2

Directive, now references backup and resilience

requirements explicitly in Article 21.

CARBON, POWER, AND

THE ARCHIVE TIER

The sustainability argument has moved

from theory into corporate accounting.

Modelling shows that moving 10 PB of cold

data from HDD to LTO over a ten-year

horizon reduces associated carbon

emissions by roughly 87%, with total cost of

ownership falling by a similar margin. IEEE

Spectrum reached the same conclusion in

its 2023 piece on tape as a climate option

for data centres.

The mechanism is simple. A tape cartridge

on a shelf draws zero watts. A disk shelf

draws power continuously for spindles,

controllers, and cooling, whether or not the

data on it is being read. For companies with

binding ESG commitments, and for

operators in regions with stressed power

grids, the carbon profile of the archive tier

has become a board-level figure.

A TWO-YEAR BRIDGE FOR THE

PRIMARY TIER

None of this suggests disk or flash is going

away. Active workloads still belong on the

fastest medium the budget supports, and the

LTO has moved from the perimeter of the

storage conversation back to its centre

because the numbers have forced the move.

It now carries the role that HDD carried a

decade ago: the capacious, cost-efficient tier

that holds the bulk of an organisation's data

while faster media handles the working set.

For IT leaders building 2027 budgets,

that reframe is useful. It allows the existing

primary estate to be extended by two

years, removes the most volatile line item

from the capex forecast, and meets

security and sustainability targets that are

no longer optional.

As organisations rework storage plans

around tighter supply and rising costs, the

role of scalable archive infrastructure is

coming back into focus. Whether delivered

through large-scale libraries or modular

deployments, LTO provides a predictable

and controllable layer in an otherwise

volatile market. It enables IT teams to

extend existing disk investments, manage

long-term data growth, and meet resilience

and sustainability requirements without

overcommitting to constrained primary tiers.

In today's conditions, that makes tape less of

a legacy choice and more of a strategic one.

More Info: www.qualstar.com



STORAGE

MAGAZINE

25

PARTNER INSIGHT INSIGHT:

BEYOND HONEYPOTS: HOW EDGE STORAGE

POWERS A NEW CYBER DECEPTION MODEL

HOW STONESTHRO, ZADARA, ZAVEIT AND LUPOVIS COMBINED EDGE CLOUD AND STORAGE

INFRASTRUCTURE TO CREATE REALISTIC ATTACK SURFACES AND DELIVER CYBER DECEPTION AS A

SCALABLE SERVICE

Traditional cyber deception models have

long relied on centralised honeypots,

but in a threat landscape shaped by

automation, AI-driven attacks and

increasingly sophisticated adversaries, those

environments are becoming easier to detect

and avoid. For deception to deliver

meaningful intelligence, attackers must

believe the systems they are interacting with

are real and not isolated traps, but part of a

living, distributed infrastructure.

That creates a new challenge for storage

and infrastructure teams. How do you build,

replicate and manage realistic

environments across multiple locations,

while maintaining the performance,

resilience and scalability required to

support both operations and deception?

Stonesthro, working with Zadara, ZaveIT

and Lupovis, set out to address this by

building a cyber deception model

grounded in edge cloud and distributed

storage architecture.

CHALLENGE

Traditional honeypots continue to play a role

in cyber defence, but many are deployed in

centralised environments that can be

fingerprinted by attackers. As adversaries

become more automated and sophisticated,

these static environments struggle to generate

high-quality, actionable intelligence.

For organisations looking to adopt cyber

deception, the challenge is twofold. First,

decoy systems must closely mirror real

enterprise environments across networks,

applications and data. Second, those

environments must be deployable at scale

across distributed locations without

introducing excessive operational complexity.

This places new demands on infrastructure.

Storage must support replication,

distribution and consistency across edge

environments, while maintaining

performance and resilience. Without this

foundation, deception environments risk

becoming predictable and ineffective.

SOLUTION

To address this, Stonesthro developed a

sovereign micro-edge cloud model designed

to support distributed cyber deception at

scale. Zadara provided the underlying edge

cloud and storage infrastructure, delivering a

flexible and resilient foundation capable of

supporting geographically dispersed

deployments. This enabled realistic

environments to be created and maintained

across multiple locations, rather than

confined to a centralised system.

"This is a hugely exciting project, illustrating

the power of the 'edge cloud' or microscaler.

Collectively we are driving a truly innovative

new market."

Alasdair Hill, Sales Director, Zadara

ZaveIT supported the commercial and

operational layer, using its platform to

package, deliver and manage the service in a

scalable, partner-ready format.

"We're proud to support the commercial

delivery of this innovative edge service."

Andy Palmer, International Sales Director, ZaveIT

Lupovis used this infrastructure and service

model to deliver its cyber deception

capabilities, deploying decoy environments

across Stonesthro's edge cloud.

RESULTS

From initial deployment, adversaries began

interacting with the deception environments,

validating the effectiveness of a more realistic

and distributed approach. While sobering in

terms of the level of hostile activity, this

engagement confirmed the value of creating

authentic, distributed attack surfaces.

The model demonstrated that cyber

deception can move beyond isolated

implementations and become a scalable

service. By leveraging edge cloud and

distributed storage, environments can be

deployed more rapidly, replicated across

locations and managed more consistently.

"Cyber deception is the way forward in the fight

against sophisticated cyber crime. In the rise of

AI and sovereign concerns, this is precisely why

we were keen to work with Lupovis."

Andy Bates, CISO, Stonesthro

The aproach gives increased threat intelligence

and visibility into attacker behaviour.

"Reacting to cyber threats is no longer

sufficient. Organisations need visibility into

attacker intent before impact."

Xavier Bellekens, CEO, Lupovis

By combining infrastructure, commercial

delivery and advanced deception capabilities,

the four partners have shown how storage

and edge architecture can underpin a new

generation of cyber defence, turning

deception into a repeatable, scalable service.

Together, as a sovereign consortium, the

partners have built a trusted framework that

combines innovation, control, and shared

intelligence to strengthen resilience at scale.

More Info: www.zaveit.io


@STMagAndAwards


MAGAZINE

CASE STUDY:

CASE STUDY: ROYAL WOLVERHAMPTON NHS TRUST

NHS TRUST SIMPLIFIES BACKUP OPERATIONS AND

IMPROVES RECOVERY PERFORMANCE AFTER

MOVING TO EXAGRID

THE ROYAL WOLVERHAMPTON NHS TRUST REPLACES MULTIPLE LEGACY BACKUP SYSTEMS WITH A

UNIFIED EXAGRID AND VEEAM ENVIRONMENT, REDUCING BACKUP WINDOWS, EXTENDING RETENTION

TO 30 DAYS, AND ACCELERATING RESTORE PERFORMANCE ACROSS 485TB OF PROTECTED DATA

The Royal Wolverhampton NHS Trust

is one of the largest acute and

community healthcare providers in

the West Midlands, with more than 800

beds at New Cross Hospital, alongside

rehabilitation beds at West Park Hospital

and Cannock Chase Hospital. As the

largest employer in Wolverhampton, the

Trust employs more than 8,000 staff.

THE CHALLENGE

The Trust's IT team had been using a

mixture of backup platforms,

including Quest NetVault and

Veritas Backup Exec for physical servers,

alongside Veeam for virtual environments.

These were supported by multiple storage

targets including disk arrays,

deduplication appliances and LTO tape.

Managing this environment required

significant staff effort and made it difficult

to complete backup operations within the

available window.

"Our weekly full backups were starting

to take more than a week to complete,

and we didn't want to leave ourselves

exposed by not having our backups

available to restore from," said John Lau,

server engineer at the Trust.

Tape copy performance was another

constraint.

"We found that copying from disk to

tape was quite slow using the solutions

we had in place, so we decided we

needed to look for a better solution that

allowed us to copy to tape very quickly,"

added Mark Parsons, the Trust's

infrastructure manager.


@STMagAndAwards


MAGAZINE

Step into

a thriving

partnership

with an

innovative

industry

leader.

The Seagate Partner Program

helps you grow with the training,

enablement, demand generation,

and marketing support you need

to gain a competitive edge.

As a Seagate partner, you’ll get:

• Access to special pricing and up-front rebates

• Sales tools like deal registration, co-marketing,

and demand generation support

• Training and certifications to boost your expertise

• Access to evaluation units, demo equipment,

and a 30-day free trial of Lyve Cloud

Scan the QR code to explore the benefits per tier:

CASE STUDY:

CASE STUDY: ROYAL WOLVERHAMPTON NHS TRUST

THE SOLUTION

Working with its long-standing adviser

PrimeSys, the Trust evaluated ExaGrid

Tiered Backup Storage as a way to

simplify its environment while improving

performance and scalability.

"PrimeSys is a specialist in data

protection and recovery, and we've been

working in the backup industry for over 20

years," said Ian Curry, director at PrimeSys

Ltd. "ExaGrid's architecture with its highperformance

Landing Zone and very

efficient long-term data repository is quite

unique in the market. We knew it was

going to be a good solution to fix the

immediate issues, but also it would give

the Trust a cost-effective way to scale and

expand going forward, as well."

The modular scale-out architecture also

aligned well with public sector lifecycle

expectations.

"We've seen a lot of take-up in public

sector with ExaGrid, with its modular

approach that allows customers like the

Trust to scale in a way that allows for a

good return on investment from the asset.

Traditionally in backup storage you buy a

system, then after three to five years later

it reaches its end of life, so you have to

keep redesigning and replacing every

three to five years.

"We know ExaGrid provides a long-term

solution, because new appliances can be

added alongside existing ExaGrid systems

as time goes on, with a non-obsolescence

policy that means that customers are

getting five to seven or even more years of

utilisation," said Curry.

Security was another important factor in

the decision.

"At PrimeSys, we're very aware that our

customers in the NHS are concerned

about and security and ransomware

impacting backups. We wanted to present

a solution we could be confident would

secure the backups of the Trust. ExaGrid

has a great range of features to secure

the system, including role-based

administration, multifactor authentication,

encryption of data at rest and in transit,

and the Retention Time-Lock (RTL) feature

which makes backups immutable so they

can't be altered and so they can't be

affected by ransomware. That was

another key reason we suggested

ExaGrid," said Curry.

ExaGrid's architecture includes a networkfacing

Landing Zone for fast restores and a

non-network-facing repository tier that

supports ransomware recovery through a

tiered air-gap design. Deduplicated data

stored in the repository tier is protected by

delayed deletes and immutable objects,

helping ensure recovery availability in the

event of an attack.

IMPLEMENTATION AND SUPPORT

Following a pilot implementation, the

Trust standardised on a combined

ExaGrid and Veeam solution. This

replaced multiple backup tools with a

single platform and resolved previous

backup-window challenges.

"During the pilot test, our ExaGrid

support engineer was very helpful with

installation and configuring our ExaGrid

system with Veeam and setting up

ExaGrid's RTL feature, and that made the

whole process seamless," said Parsons.

"Our ExaGrid support engineer guided

me through Veeam settings that offer the

most benefits with ExaGrid to get the most

out of the solution, which was great," said

Lau. "Working with ExaGrid has all been

very positive. If I notice any alerts on our

ExaGrid system, I can simply email our

support engineer and he often replies to

me within a few minutes."

THE RESULT

The Trust currently protects approximately

485TB of data using daily incremental

backups and weekly full backups that are

also written to tape and stored offsite for

additional protection.

Since deploying ExaGrid, onsite retention

has been extended to 30 days while still

maintaining performance targets.

"We noticed a huge improvement in our

backup performance since switching to

ExaGrid," said Lau. "Even though we have

added more data to back up, our backups

still fit within the desired window, and

making copies to tape is quicker, too."

Restore performance has also improved.

"One of the less-experienced members

of our IT team had to restore an

enormous VM recently, and the restore

was quite quick considering the size, and

it was a very straightforward process so he

was able to manage doing the restore on

his own just fine," said Parsons.

PARTNER PERSPECTIVE

PrimeSys emphasised that long-term

service quality was as important as

technical capability.

"At PrimeSys, we put the customer

experience at the heart of everything we

do. This means the solutions we provide

have to deliver, in terms of the technology

but also its installation, integration with

existing systems, and ongoing support.

This is particularly important in healthcare

where lives are literally at stake. ExaGrid

provides industry-leading storage

efficiency, performance and security but

it's the level of post-sales customer service

and support that really sets them apart.

"Working with ExaGrid, we can be

confident that our solutions will deliver

and our customers will receive the highest

standards of service and support, through

the life of the solution," concluded Curry.

More Info: www.exagrid.com


@STMagAndAwards


MAGAZINE

EVENT:

EVENT: STORAGE AWARDS 2026

THE STORAGE AWARDS 2026:

AND SOMEHOW IT'S THAT TIME AGAIN

The finalists for the Storage Awards 2026

are officially announced, and the shortlist

is packed with categories for innovation,

resilience, AI, cloud, cyber security, backup,

recovery, and enough of our favourite industry

people to keep us all happily arguing well into

the summer.

Firstly though, a huge thank you to everyone

who entered and nominated and emailed us

"just one final update" at 11pm the night

before entries closed. We see you! (but we also

have you!).

So yes somehow, unbelievably, the Storries

STORAGE INNOVATOR OF THE YEAR

Arcserve

Boston

Cloudian

Cohesity

CTERA

Dell Technologies

Druva

ExaGrid

Fujifilm

Hitachi Vantara

Hornetsecurity by Proofpoint

IBM

Infinidat

Infoscale

Kioxia

Lenovo

Nasuni

Nexsan

Nutanix

Object First

PEAK:AIO

Phison

QNAP

Quantum

Seagate

Stonethro

StorMagic

StorONE

Storware

StorPool

Supermicro

Synology

Tiger Technology

Tintri

Toshiba

Veeam

Wasab

WD

Zadara

ZaveIT

ONE TO WATCH - COMPANY

Assured Data Protection

Cohesity

CTERA

EuroNAS

Hitachi Vantara


Infinidat

are back round again and we are thrilled to be

back in one of our favourite of all Storage

Awards venues over the years, the sparkling

De Vere Grand Connaught Rooms (yes the

one with the chandeliers and sweeping

staircase that somehow gets steeper on the

way out).

As always, most of the categories are voted

by you, with only a handful the reserve of

panels or Editor's Choice. So this really is a

Peoples Vote.

Voting is now open and closes on 4 June, so

now is the time to support your colleagues,

FINALISTS LINE-UP 2026:

Isystems

Lightbits

Nasuni

Nexsan

Object First

Phison

QNAP

Qualstar

Scale Computing

Seagate

Soft Iron

StorMagic

StorONE

StorPool

Storware

Tiger Technology

Tintri

VAST Data

Wasabi

ZaveIT

ONE TO WATCH - CHANNEL

101 Data Solutions

Academia

Astralis

Cameo

CDS

Climb Channel Solutions

Convergent Technology

Coolspirit

ICT Reverse

Insurgo

M2M

N2S

Primesys

Procurri

Relltek

Stonesthro

Ultra Support

IMMUTABLE STORAGE COMPANY

OF THE YEAR

Acronis

Arcserve

Cloudian

Cohesity

Everpure

ExaGrid


HPE

partners and favourite technologies. Vote here:

www.storage-awards.com

Good luck to all our finalists and see you on

the 18th.

Sharon Munday, Editor of Storage Magazine

PS. Thanks to our chief and diamond sponsors

for 2026, ExaGrid, Epaton and NGS.

IBM

Infinidat

Infoscale

NetApp

Nexsan

Object First

QNAP

Qualstar

Quantum

Rubrik

Scality

Seagate

Synology

Tintri

Wasabi

Zadara

STORAGE INDUSTRY CHAMPION -

COMMERCIAL

101 Data Solutions - Brett Edgecombe

Arcserve - Karl Fontanari

Barracuda - Giovanni Goduti

CDS - Ricky Patel

CMS Distribution - Frank Salmon

CTERA - Beky Cairns

Druva - Pam Napier

Epaton - Johnathan Lassman

ExaGrid - Andy Walsky

Nexsan - Adrian Hedges

Northamber - Alexander Philips

Nutanix - Stuart Heade

Object First - Pete Hannah

QNAP - Tom Jepps

Scale Computing - Gary Lynch

Stonesthro - Andy Bates

StorMagic - Scott Mann

Supermicro - Jonathan "Jonnie" Ray

Synology - Nic Metcalfe

Tintri - Mark Walsh

Titan Data Solutions - Ben Jackson

Ultra Support - Pete Hodgson

Virtual Effect - John Greenwood


TECHNICAL

101 Data Solutions - Dan Luxton

Arcserve - Carl Green

Climb Channel Solutions - Gary Morris

Cloudian - Neil Stobart

Epaton - Pete Aspinall


@STMagAndAwards


MAGAZINE

EVENT:


Everpure - David Milne

Hammer Distribution - Jason Daly

Nasuni - Chris Wales

Nexsan - Richard Hornsby

Nutanix - Ralf Frankemölle

QNA - Craig Reid

Scale Computing - Ian Smith

StorMagic - Mark Christie

Titan Data Solutions - David Treadwell

VAST Data - Ross Cooper-Smith

Wasabi - Neale "Nelly" Simpkins

Zadara - Nick Cambridge


MARKETING

101 Data Solutions - Natasha Waldron

Barracuda - Rebecca Bailey

Boston - Miodrag Relic

Climb Channel Solutions - Kyle King

Druva - Hayley Newlove

ExaGrid - Mary Domenichelli

Fujifilm - Sarah Taylor

Infinidat - Eric Herzog

Indoscale - Varun Verma

Kingston Technology - Miriam Brown

ObjectFirst - Anastasiia Stepanova

Open-E - Mikolaj Paszko

Nexstor - Catherine Osborne

Nutanix - Alex Almeida

Quantum - Tom Hassall

Seagate - Cassie Newman

StorMagic - Victoria Sapsford

StorPool - Beth Caltagirone

Tintri - Ken Man

CHANNEL EXCELLENCE AWARD

Arrow ECS - William "Billy" Bond

Assured Data Protection - Ron Mackle

Barracuda - Giovanni Goduti

CDS - Ricky Patel

Climb Channel Solutions - Gerard Brophy

CMS Distribution - Frank Salmon

Druva - Pam Napier

Epaton - Jonathan Lassman

ExaGrid - Andy Walsky

Hammer Distribution - Calum Collington

Infinidat - James Lewis

M2M - Patrick Mitchell

Nasuni - Kenz Mroue

Seagate - Andy Palmer

SoftIron - Philip Crocker

StorMagic - Matt Fearn

Titan Data Solutions - David Treadwell

AI DATA STORAGE INNOVATOR

OF THE YEAR

Cloudian

CTERA

Dell Technologies

Everpure

ExaGrid

HPE

IBM

Infoscale

Kioxia

NetApp

Nutanix

PEAK: AIO

Scality

Seagate

Stonesthro

Supermicro

Tiger Technology

WD

Zadara

CYBER RESILIENT STORAGE COMPANY

OF THE YEAR

Acronis

Arcserve


Barracuda

Carbonite

Cloudian

CTERA

Cohesity

CrashPlan

Data Barracks

Datto

Druva

Everpure

ExaGrid

Fujifilm


IBM

Infinidat

Infoscale

Nasuni

Nutanix

Object First

Quantum

Rubrik

Scality

Seagate

StorONE

Synology

Zerto

RANSOMWARE COMPANY OF THE YEAR

Acronis

Arcserve

Asigra

Cohesity

CTERA

Druva

Everpure

ExaGrid

Fujifilm


IBM

Infinidat

Infoscale

Nasuni

NetApp

Nutanix

Object First

Open-E

QNAP

Scality

Synology

Wasabi

Zerto

CLOUD BACKUP COMPANY OF THE YEAR

Acronis

Arcserve

Barracuda

Carbonite

Cohesity

Commvault

CrashPlan

Datto

Druva


IBM

QNAP

Quantum

Relltek

Rubrik

Seagate

StorPool

Veeam

Wasabi

Zerto

DATA PROTECTION COMPANY

OF THE YEAR

Acronis

Arcserve

Asigra

Barracuda

Carbonite

Cohesity

Commvault

CrashPlan

Datto

Druva

Everpure

ExaGrid


HPE

IBM

Infinidat

Kingston Technology

Nexsan

Open-E

Rubrik

Scality

StorONE

Storware

Supermicro

Synology

Veeam

ZaveIT

Zerto

ENTERPRISE BACKUP HARDWARE VENDOR

OF THE YEAR

Arcserve

Boston

Cohesity

CTERA

Dell Technologies

Everpure

ExaGrid

Hitachi Vantara

HPE

IBM

Infinidat

Lenovo

NetApp

Object First

QNAP

Qualstar

Quantum

Rubrik

Seagate

StorONE

Synology

Tintri

VAST Data

DATA STORAGE SOFTWARE COMPANY

OF THE YEAR

Acronis

Barracuda

Cloudian

Commvault

Databricks

DataCore

Druva

EuroNAS

Infinidat

Infoscale

Nutanix

Scality

Stonesthro

StorONE

StorPool

Synology

Veeam

TAPE COMPANY OF THE YEAR

Fujifilm

HPE

IBM

Lenovo

Overland Storage

Qualstar

Quantum

Sony

Spectra Logic

ENTERPRISE DATA STORAGE AND SERVICE

COMPANY OF THE YEAR

101 Data Solutions

Cameo

Hammer Distribution

Insurgo

Kingston Technology

Park Place Technologies

Procurri

Relltek

Stortrec

Ultra Support

Zadara

DATA MANAGEMENT & MONITORING

VENDOR OF THE YEAR


Blancco

Cirrus Data

Cohesity

CTERA

Databricks



STORAGE

MAGAZINE

32

EVENT:


DataCore

Dell Technologies

Hitachi Vantara

HPE

IBM

Infoscale

Komprise

NetApp

Open-E

Panzura


StorPool

StorMagic

VAST Data

ZaveIT

OBJECT STORAGE VENDOR

OF THE YEAR

Cloudian

CTERA

DataCore

Dell Technologies

EuroNAS

Everpure

ExaGrid

HPE

IBM

Nasuni

NetApp

Nutanix

Object First

QNAP

Quantum

Scality

Seagate

Spectra Logic

StorONE

Supermicro

Tintri

VAST Data

Wasabi

Zadara

HYPER-CONVERGENCE

VENDOR OF THE YEAR

Boston

DataCore

EuroNAS

Hitachi Vantara

HPE

IBM

Lenovo

NetApp

Nutanix

Open E

Scale Computing

StorMagic

StorPool

Supermicro

Synology

Tintri

Zadara

'AS A SERVICE' PLATFORM

OF THE YEAR

101 Data Solutions

Acronis

Arcserve

Arrow ECS

Asigra


Barracuda

Carbonite

Cohesity

Commvault

CrashPlan

Datto

Druva


HPE

Nutanix

Scality

Seagate

StorPool

Veeam

Wasabi

ZaveIT

FLASH STORAGE VENDOR

OF THE YEAR

Dell Technologies

Everpure

Hitachi Vantara

HPE

IBM

Infinidat

Infortrend

Kingston Technology

Kioxia

Lenovo

NetApp

Nexsan

Phison

QNAP

Quantum

Seagate

StorONE

Synology

Tintri

Toshiba

VAST Data

PERFORMANCE STORAGE

VENDOR OF THE YEAR

Boston

CTERA

Dell Technologies

Everpure

Hitachi Vantara

HPE

IBM

Infinidat

Lenovo

Nasuni

NetApp

Nexsan

Nutanix

PEAK:AIO

QNAP

Quantum

Scality

Seagate

Soft Iron

Tintri

Toshiba

VAST Data

Zadara

CLOUD ENABLER OF THE YEAR

Acronis

Arcserve

Arrow ECS

Barracuda

Carbonite

Cloudian

Cohesity

CrashPlan

CTERA

Databricks

DataCore


IBM

Infinigate

Infoscale

Nasuni

NetApp

Nutanix

Panzura

QNAP

StorPool

Tiger Technology

Wasabi

SOFTWARE DEFINED STORAGE

(SDS) VENDOR OF THE YEAR

Cohesity

CTERA

DataCore

EuroNAS

IBM

Infinidat

Infoscale

NetApp

Nutanix

Open-E

Panzura

PEAK:AIO

Scale Computing

Scality

Seagate

Soft Iron

StorMagic

StorONE

StorPool

Supermicro

Tiger Technology

Zadara

STORAGE OPTIMISATION

COMPANY OF THE YEAR

Accelstor

Cirrus Data

Cloudian

Cohesity

DataCore

EuroNAS

Everpure

ExaGrid

HPE

IBM

Infinidat

Infoscale

Komprise

Lightbits

Nasuni

NetApp

Nutanix

PEAK:AIO

Quantum

Seagate

StorONE

StorPool

Tiger Technology

Tintri

Toshiba

ELECTRONIC DATA

WAREHOUSING COMPANY OF

THE YEAR


Cloudian

Databricks

Dell Technologies

Hitachi Vantara

HPE

Natrinsic

Nasuni

NetApp

Quantum

Scality

StorONE

Supermicro

VAST Data

CAPACITY STORAGE VENDOR

OF THE YEAR

Boston

Cloudian

Cohesity

DDN

Dell Technologies

Everpure

ExaGrid

Hitachi Vantara

HPE

IBM

Infinidat

Infortrend

Lenovo

NetApp

Nexsan

Nutanix

QNAP

Qualstar

Quantum

Scality

Seagate

Supermicro

Synology

Toshiba

Tintri

VAST Data

CHANNEL PARTNER

PROGRAMME OF THE YEAR

Arcserve

Barracuda

Boston

Cohesity

DataCore

Druva

ExaGrid

HPE

Infinidat

Lenovo

Nexsan

Nutanix

Object First

QNAP

Quantum

Scality

Seagate

StorMagic

ZaveIT

MULTI-VENDOR SERVICE

PROVIDER OF THE YEAR

101 Data Solutions

Cameo

CDS

Insurgo

Natrinsic


Procurri

Relltek

Stortrec

Ultra Support

Storage Repair Specialist

Cameo

CDS

Gentronics

ICT Reverse

N2S

Procurri

Ultratec

SUSTAINABILITY CHAMPION

OF THE YEAR

Astralis Tech

AWS

Cameo

CDS

Coolspirit

Evernex

Gentronics

HPE

IBM

ICT Reverse

Infinidat

Insurgo

N2S

NetApp


PEAK:AIO

Procurri

Relltek

Seagate

Stortrec

Ultra Support

Ultratec

DATA DESTRUCTION

COMPANY OF THE YEAR

Astralis Tech

Blancco

Cameo

CDS

Evernex

Gentronics

ICT Reverse

N2S


Procurri

Relltek

Ultratec

CYBER SECURITY STORAGE

RESELLER OF THE YEAR

101 Data Solutions

Central Technology

ITEC


@STMagAndAwards


MAGAZINE

EVENT: STORAGE EVENT:

AWARDS 2026

Logicalis

MTI

Nexstor

NGS

Primesys

Proact IT

Softcat

Telefonica Tech

TruStack

SPECIALIST STORAGE RESELLER OF THE

YEAR

101 Data Solutions

CCK Tech

Cognitive Network Solutions


CSI (Computer Systems Integration)

Cristie Data

Datek Solutions

DSI

Epaton

Logicalis

Nexstor

Primesys

Proact IT

Techvertu

Virtual Effect

ENTERPRISE STORAGE RESELLER OF THE

YEAR

101 Data Solutions

Academia

Boxxe

Bytes

CCS Media

CDW

Centerprise

Computacentre


Coolspirit

Epaton

European Electronique

Insight

MTI

Nexstor

Softcat

Trustmarque

Viadex

MANAGED SERVICE PROVIDER (MSP) OF

THE YEAR

101 Data Solutions

Air IT

ANS Group


Centerprise

Data Barracks

Datastore365

Entrust IT

Exponential-E

Kascade

Meridian IT

Nexstor

Node4

OVHcloud

Stonesthro

Viadex

Xtravirt

Zen Internet

SPECIALIST STORAGE DISTRIBUTOR OF

THE YEAR

Arrow ECS

CMS Distribution

Ethos Technology

Exclusive Networks

Hammer Distribution

M2M

Titan Data Solutions

ENTERPRISE STORAGE DISTRIBUTOR OF

THE YEAR

Arrow ECS


CMS Distribution

Hammer Distribution

Infinigate

Ingram Micro

Northamber

Nuvius

TD Synnex

Titan Data Solutions

Westcoast

EDITOR'S CHOICE - PRODUCT

NOT OPEN FOR PUBLIC VOTING - EDITOR

OF STORAGE MAGAZINE WILL DECIDE

WINNER

EDITOR'S CHOICE - COMPANY

NOT OPEN FOR PUBLIC VOTING - EDITOR

OF STORAGE MAGAZINE WILL DECIDE

WINNER

STORAGE SOLUTION OF THE YEAR -

PUBLIC SECTOR

NOT OPEN FOR PUBLIC VOTING - PANEL

OF INDUSTRY EXPERTS WILL DECIDE

WINNER

STORAGE SOLUTION OF THE YEAR -

CORPORATE


OF INDUSTRY EXPERTS WILL

DECIDE WINNER

CLOUD SOLUTION OF THE YEAR


OF INDUSTRY EXPERTS WILL

DECIDE WINNER

CLOUD PRODUCT OF THE YEAR

Acronis - XDR

Arcserve - Cloud Direct

Asigra - Backup2Cloud

Barracuda - Intronis Backup

Cohesity - FortKnox

Commvault - Cloud Rewind

Druva - Data Security Cloud

Everpure - Everpure Cloud

Hornetsecurity - 365 Total Protection

HPE - GreenLake

IBM - Cloud Backup

Nutanix - Files on Public Cloud

Seagate - Lyve Cloud

Spectra Logic - Vail

StorPool - StorPool Storage

Tiger Technology - Tiger Bridge

Veeam - Veeam Data Cloud

Wasabi Cloud NAS

Zadara - zCompute

ZaveIT - ZaveIT Platform

STORAGE PRODUCT OF THE YEAR

Arcserve - Cyber Resilient Storage

Barracuda - EntraID Backup

Boston - Igloo AI+

Cloudian - Hyperstore

Cohesity - NetBackup Flex Appliance

CTERA - Intelligent Data Platform

DataCore - Puls8

EuroNAS - eEKAS

Everpure - Everpure//E

ExaGrid - EX540-SSD

Fujifilm - LTO Ultrium

Hornetsecurity - 365 Total Backup

HPE - Alletra Storage MP X10000

IBM - Storage FlashSystem 5600

Infinidat - InfiniBox G4 Family

Infoscale - Infoscale Enterprise

Kingston Technology - DC3000ME PCIe 5.0

NVMe U.2 SSD

Kioxia - AiSAQ

Nasuni - File Data Platform

NetApp - StorageGRID

Nexsan - VHR-Series

Nutanix - NCI

Object First - Ootbi

Open-E - JovianVHR

Panzura - CloudFS

QNAP - TS-h1077AFU

Quantum - ActiveScale Cold Storage

Scality - ARTESCA

Seagate - MACH.2

Spectra Logic - OSW-2400 Optical SAS Switch

StorMagic - SvHCI

StorONE - Smart Auto-Tiering Cloud

StorPool - StorPool Storage

Synology - HD6500

Tintri - VMstore

VAST Data - DataStore

Veeam - Veeam Vault

Zadara - VPSA

CLOUD COMPANY OF THE YEAR

Acronis

Asigra

Arcserve

Barracuda

Carbonite

Cloudian

Cohesity

Commvault

CrashPlan

CTERA

DataCore

Datto

Druva

Everpure

Hitachi Vantara


HPE

IBM

Infinigate

Nasuni

NetApp

Nexstor

Nutanix

Rubrik

Scality

Seagate

StorPool

Supermicro

Veeam

Wasabi

Zadara

STORAGE COMPANY OF THE YEAR

Acronis

Arcserve

Barracuda

Boston

Cloudian

Cohesity

Commvault

CTERA

DataCore

Dell Technologies

Druva

EuroNAS

Everpure

ExaGrid

Fujifilm


HPE

IBM

Infinidat

Infoscale

Nasuni

Nexsan

Nutanix

Object First

Open-E

QNAP

Quantum

Scality

Seagate

Soft Iron

Spectra Logic

StorMagic

StorPool

Supermicro

Tintri

Toshiba

Ultra Support

Veeam

WD

ZaveIT

www.storagemagazine.co.uk @STMagAndAwards May/June 2026

STORAGE

MAGAZINE

34

COMING SOON!

W W W .STORAGE-AW ARDS.COM

ST2605

Transform your PDFs into Flipbooks and boost your revenue!

Delete template?

Save as template ?