ISC 2024: Hyperion Research Predicts HPC Market Rebound after Flat 2023

By John Russell

May 13, 2024

First, the top line: the overall HPC market was flat in 2023 at roughly $37 billion, bogged down by supply chain issues and slowed acceptance of some larger systems (e.g. exascale), according to Hyperion Research’s annual pre-ISC HPC market update presented last week. On-prem server revenue was down 2.7%. In stark contrast, HPC-related spending in the cloud grew robustly. In 2024, Hyperion expects on-premise server sales to rebound to around $16.3 billion.

No surprise, Hyperion emphasized HPC’s continued aggressive AI adoption, which is rapidly transforming HPC into a blended HPC/AI hybrid discipline. That’s prompted Hyperion to add a new expanded AI tracking program. Storage, a perennial strong market in HPC, remained robust. It was ~$6.2 billion in 2023 and is expected to grow to nearly $10 billion in 2028 (9.3% CAGR).

Commenting on the trends, Earl Jospeh, CEO, Hyperion Research, said, “In 2023 we saw a lot of orders placed, but not quite as many as acceptances as we thought. We think a lot of buyers were taking longer to make their decisions on what to purchase, how much to spend on the cloud versus on prem. That kind of dynamic is working its way through. So there’s growth we’re expecting here.

“At the same time, we still expect to see the supply chain issues, and the delays and exascale acceptances throughout this period. So there’s many growth drivers but there are also some constraints and we’re trying to balance how that’s going to land but we expect 2024 to be a very strong.” It’s important to note that Hyperion’s cloud numbers (below) are for user spending on HPC-related work in the cloud, and do not include spending by cloud providers on HPC-related infrastructure.

Hyperion packs a lot into its mid-year presentation with video-recorded ~10-minute snapshots of individual market segments (storage, interconnect, AI drivers, etc) presented by different analysts. It’s best to watch them directly and Hyperion plans to make the videos and slides available from its web site.

Joseph handled the opening overview with others doing segment drill-downs. Joseph also presented a list of 11 predictions (shown below) and discussed three of them.

Hyperion’s 2024 Predictions

  1. Utilization of HPC resources in the cloud will experience accelerating growth as users augment their AI focus on training with inferencing
  2. While “speeds and feeds” will continue to be important to buyers of storage systems, the primary value point and competitive advantage for data storage solutions will shift to the “Data Platform”
  3. System vendors will struggle to a greater degree than hyperscalers in absorbing the accelerated cadence of NVIDIA’s GPU roadmap
  4. Interest in procuring on-premises quantum computing (QC) systems will grow, augmenting but not replacing access to QC through the cloud
  5. Installations of HPCs built by large end users, not traditional HPC suppliers, will become more commonplace, particularly for those targeting high-end AI computational workloads
  6. RISC-V will continue to gain ground as a viable processor option
  7. The cost of energy will more frequently outweigh HPC performance increases, as some
    sites will settle for “good enough” speeds in order to reach energy efficiency goals
  8. The importance of monolithic HPC benchmarking will diminish
  9. LLM frameworks will gain prominence within the overall trend of LLM usage
  10. Generative AI growth will continue, but adoption growth rates will stabilize as applicability and limitations become more well understood
  11. Arm-based processors will rise sharply in adoption, with expected revenues for Arm- based HPC systems doubling compared with the previous year

Number three (accelerated GPU roadmaps) and number five (so-called dark HPC) are fascinating.

Latest vendor share ranking from Hyperion. Source: Hyperion Research.

Joseph said, “[The] acceleration of GPU roadmaps is presenting some serious problems for system vendors and users. The announcement by Nvidia, closely followed by AMD and Intel of shifting their roadmaps cadence from 18 months to something closer 12 months means for vendors that they have to incorporate a new processor type every 12 months.” He noted they all have multiple processor types, “which really means it’s almost every quarter [system vendors] have a new processor type to incorporate in their systems. The vendors have to make their systems and the software work across a much greater diversity of hardware. That’s causing some tremendous pain.”

Users, he said, are also being caught in the introduction cycle speed-up.

“When you buy a system, it takes time to do the RFP, to do make the decision, then get the system ordered and installed — [that] can exceed a year easily. By the time you turn your system on, you might be a generation or two behind in technology already, and as I mentioned, the price of the systems are growing dramatically. So users have to keep them longer to justify the higher amount of spending, perhaps 4-5 years or longer. Throughout the life of the system, you can easily fall three to four generations behind the leading edge technologies,” said Joseph.

About the rise of larger, less visible to the public, HPC installations, he said, “You know, so many years ago, we saw this as a standard practice, [but] it really disappeared for over the last 20 years or so. Now it’s being revitalized announcements from Google, and Microsoft, and Tesla, and the large Chinese sites are building truly gigantic systems. In addition, they’re looking at designing their own processors for the system. So it’s a whole new resurgence in the market, although it doesn’t show up in our market numbers, because the system hasn’t been sold from, you know, a buyer to a seller.”

Joseph noted the HPC market has always been prone to big swings but has grown even challenging to forecast. One result is Hyperion has begin updating its forecast more frequently. The charts below are illustrative. One shows the historical worldwide HPC server data showing dips and rise. The others show how Hyperion has begun adapting its forecasts more frequently.

Like virtually everyone in the HPC market watcher/consultant community, Hyperion is beefing up its AI practice.

“I’m happy to announce a new program. We’ve expanded our whole AI program and focus…building on our last 10 years of tracking AI, Big Data HPBA,” said Joseph, citing: more frequent identification of the key issues including drivers and inhibitors, best case practices such as how to make trade-offs on hardware, architecture, software, and new language models.

“To help us do this, we’re creating an AI experts Advisory Committee. We’re looking to the committee to give us suggestions and advice and guidance on what are the most critical issues right now?” He suggested interested potential participants could contact Hyperion.”

Bob Sorensen, Hyperion, cited changing system needs, suggesting there’s no longer a one-size fits all architecture in HPC; this the extension of an ongoing trend perhaps, but growing faster as AI workloads become infused with HPC.

Analyst Mark Nossokoff reviewed the HPC interconnect and  storage segments where AI’s changing workflow demands are shifting buyers’ file system preferences.

“Many of the file systems were developed for the traditional mod-sim workloads, particularly the checkpoint/restart type of items. The AI workloads are providing a much, much broader variety of profiles, whether the small block random, in addition to the large block sequential, different different access methods, different access frequency for the types of data that’s being required. We’re seeing shifting in the file system adoption preferences,” he said.

“We have a study and process that is taking stock of what’s happening in the file system adopted space today, and we’ll have results of those for you in the coming months. We’re seeing that while that traditional file systems continue to be used, there is an emerging class of file systems coming out by by several vendors that started with greenfield opportunity and with AI and addressing those needs from the from the start. In addition to that, we are seeing the broader business models supporting the these file systems from Hammerspace, VAST, and now VDURA, which was formerly Panasas, and Weka, all delivering software solutions as a business model.”

Sorensen gave a brief summary of the quantum market, which Hyperion predicts will pass the $1 billion mark next year.

“We’re pretty confident in this this relatively aggressive growth rate simply because we see a number of factors on the supply side that contribute to this. We’re seeing continued revenue growth by the traditional quantum computing supplier organizations — I say traditional in that they have been in business for perhaps more than two or three years — but we’re also seeing a lot of first revenue appearances by new to market players,” said Sorensen.

User interest is also increasing. “There’s widespread interest in accelerating critical compute jobs, many of them being constrained by the increasing cost complexity and power consumption of classical HPC. We’re seeing use case exploration of all over the place in terms of how can quantum help address some of my most vexing computational jobs. And we’re seeing sustained government programs, not only for continued funding to support R&D but also upticks in government procurements, the the acquisition of quantum computing hardware in order to make those systems available to leading research facilities, both within the government, commercial and even within academia as well. So we’re seeing lots and lots of demand for both supply and demand side.”

Hyperion’s market update contains too much to cover in a single article. Hyperion is making the presentation and slides available from its web site.

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

ISC 2024 Takeaways: Love for Top500, Extending HPC Systems, and Media Bashing

May 23, 2024

The ISC High Performance show is typically about time-to-science, but breakout sessions also focused on Europe's tech sovereignty, server infrastructure, storage, throughput, and new computing technologies. This round Read more…

HPC Pioneer Gordon Bell Passed Away

May 22, 2024

Legendary computer scientist Gordon Bell passed away last Friday at his home in Coronado, CA. He was 89. The New York Times has a nice tribute piece. A long-time pioneer with Digital Equipment Corp, he pushed hard for de Read more…

ISC 2024 — A Few Quantum Gems and Slides from a Packed QC Agenda

May 22, 2024

If you were looking for quantum computing content, ISC 2024 was a good place to be last week — there were around 20 quantum computing related sessions. QC even earned a slide in Kathy Yelick’s opening keynote — Bey Read more…

Atos Outlines Plans to Get Acquired, and a Path Forward

May 21, 2024

Atos – via its subsidiary Eviden – is the second major supercomputer maker outside of HPE, while others have largely dropped out. The lack of integrators and Atos' financial turmoil have the HPC market worried. If Atos goes under, HPE will be the only major option for building large-scale systems. Read more…

Core42 Is Building Its 172 Million-core AI Supercomputer in Texas

May 20, 2024

UAE-based Core42 is building an AI supercomputer with 172 million cores which will become operational later this year. The system, Condor Galaxy 3, was announced earlier this year and will have 192 nodes with Cerebras Read more…

Google Announces Sixth-generation AI Chip, a TPU Called Trillium

May 17, 2024

On Tuesday May 14th, Google announced its sixth-generation TPU (tensor processing unit) called Trillium.  The chip, essentially a TPU v6, is the company's latest weapon in the AI battle with GPU maker Nvidia and clou Read more…

ISC 2024 Takeaways: Love for Top500, Extending HPC Systems, and Media Bashing

May 23, 2024

The ISC High Performance show is typically about time-to-science, but breakout sessions also focused on Europe's tech sovereignty, server infrastructure, storag Read more…

ISC 2024 — A Few Quantum Gems and Slides from a Packed QC Agenda

May 22, 2024

If you were looking for quantum computing content, ISC 2024 was a good place to be last week — there were around 20 quantum computing related sessions. QC eve Read more…

Atos Outlines Plans to Get Acquired, and a Path Forward

May 21, 2024

Atos – via its subsidiary Eviden – is the second major supercomputer maker outside of HPE, while others have largely dropped out. The lack of integrators and Atos' financial turmoil have the HPC market worried. If Atos goes under, HPE will be the only major option for building large-scale systems. Read more…

Google Announces Sixth-generation AI Chip, a TPU Called Trillium

May 17, 2024

On Tuesday May 14th, Google announced its sixth-generation TPU (tensor processing unit) called Trillium.  The chip, essentially a TPU v6, is the company's l Read more…

Europe’s Race towards Quantum-HPC Integration and Quantum Advantage

May 16, 2024

What an interesting panel, Quantum Advantage — Where are We and What is Needed? While the panelists looked slightly weary — their’s was, after all, one of Read more…

The Future of AI in Science

May 15, 2024

AI is one of the most transformative and valuable scientific tools ever developed. By harnessing vast amounts of data and computational power, AI systems can un Read more…

Some Reasons Why Aurora Didn’t Take First Place in the Top500 List

May 15, 2024

The makers of the Aurora supercomputer, which is housed at the Argonne National Laboratory, gave some reasons why the system didn't make the top spot on the Top Read more…

ISC 2024 Keynote: High-precision Computing Will Be a Foundation for AI Models

May 15, 2024

Some scientific computing applications cannot sacrifice accuracy and will always require high-precision computing. Therefore, conventional high-performance c Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Atos Outlines Plans to Get Acquired, and a Path Forward

May 21, 2024

Atos – via its subsidiary Eviden – is the second major supercomputer maker outside of HPE, while others have largely dropped out. The lack of integrators and Atos' financial turmoil have the HPC market worried. If Atos goes under, HPE will be the only major option for building large-scale systems. Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Some Reasons Why Aurora Didn’t Take First Place in the Top500 List

May 15, 2024

The makers of the Aurora supercomputer, which is housed at the Argonne National Laboratory, gave some reasons why the system didn't make the top spot on the Top Read more…

Leading Solution Providers

Contributors

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Eyes on the Quantum Prize – D-Wave Says its Time is Now

January 30, 2024

Early quantum computing pioneer D-Wave again asserted – that at least for D-Wave – the commercial quantum era has begun. Speaking at its first in-person Ana Read more…

The GenAI Datacenter Squeeze Is Here

February 1, 2024

The immediate effect of the GenAI GPU Squeeze was to reduce availability, either direct purchase or cloud access, increase cost, and push demand through the roof. A secondary issue has been developing over the last several years. Even though your organization secured several racks... Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

The NASA Black Hole Plunge

May 7, 2024

We have all thought about it. No one has done it, but now, thanks to HPC, we see what it looks like. Hold on to your feet because NASA has released videos of wh Read more…

Intel Plans Falcon Shores 2 GPU Supercomputing Chip for 2026  

August 8, 2023

Intel is planning to onboard a new version of the Falcon Shores chip in 2026, which is code-named Falcon Shores 2. The new product was announced by CEO Pat Gel Read more…

GenAI Having Major Impact on Data Culture, Survey Says

February 21, 2024

While 2023 was the year of GenAI, the adoption rates for GenAI did not match expectations. Most organizations are continuing to invest in GenAI but are yet to Read more…

How the Chip Industry is Helping a Battery Company

May 8, 2024

Chip companies, once seen as engineering pure plays, are now at the center of geopolitical intrigue. Chip manufacturing firms, especially TSMC and Intel, have b Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire