ROI: Is HPC Worth It? What Can We Actually Measure?

By Addison Snell, Intersect360 Research

October 15, 2020

Editor’s Note: Intense disagreement and debate are powerful forces, which when productively applied, can often clarify issues. Though not always. At HPCwire, we’re hoping the arguments presented here clarify. Measuring ROI, it turns out, is often tricky and subject to dispute. This is very true for HPC whose deployment is rarely inexpensive or easy and whose influence on favorable outcomes may be difficult to quantify. The HPC payoff, we’re told, is worth the effort, and intuitively, this certainly seems the case. Think of breakthroughs (LIGO) enabled by HPC in science and payoffs delivered by HPC for industry (successful oil exploration and hedge fund trading).

In this commentary, Intersect360 Research CEO Addison Snell, disagrees with Hyperion Research’s approach to measuring ROI. Both Hyperion (formerly IDC’s HPC practice) and Intersect360 have long standing and significant influence in the HPC world. Both are well-respected. Hyperion’s ROI study – an ongoing effort begun with a government contract – is widely circulated including by HPCwire. It is because the Hyperion ROI measure is so widely circulated that HPCwire agreed to publish this commentary at Intersect360’s request. Hyperion argues its ROI tracking measure is well-considered and a good guide for HPC users.

HPC enables innovation and discovery. We all seem to agree on that. Is there a good way to quantify how much that’s worth?

Thanks to a sponsored white paper citing it, the annual “ROI of HPC” study from Hyperion Research is making the rounds again. It breathtakingly reports that organizations investing in HPC can expect a staggering, average return on investment of 4,300%: $44 dollars in profits and $463 in revenue for every dollar spent on high performance computing.

Let’s get this part out of the way now: Nothing in the corporate world has an average ROI of 4,300%. If there were such a thing, companies would do little else. The real goal of the study isn’t to measure actual ROI, but rather to make the case that HPC has quantifiable value. That brings up the interesting question: Does it?

The Real ROI of HPC

Addison Snell, CEO, Intersect360 Research

The true ROI of HPC investment isn’t nearly so high. The problem is, it is also terribly difficult to measure. The formula for ROI is simple: net profit over total investment. We have a good sense of total HPC investment, but for the types of projects in which HPC is deployed, this isn’t the same as total investment. There are costs beyond the HPC itself. I can’t sell an ice cream cone without ice cream, but I also can’t sell one without a cone, and what about the cost of the freezer?

ROI is therefore typically measured at the project or business unit level. We can assess the ROI of, say, opening a new store, or launching a new product line, but it’s hard to figure the ROI of individual expenses like computers.

For this reason, many companies will measure large investments against a corporate hurdle rate, maybe 12% or 15%, and will okay investments that are projected to beat it. Such calculations take into account the organizational discount rate, related to cost of capital, to account for the notion that a dollar today beats a dollar tomorrow.

Furthermore, any true ROI study measures marginal investment; i.e., the expected return of the next dollar spent. Going from zero to $100,000 will have a very different return than going from $100.0 million to $100.1 million.

Additionally, there’s a problem with addressing ROI at an industry level, because secondary demand stimulation (i.e., stealing market share) is zero sum. For certain investments, this can even lead to prisoner’s dilemmas.

Say Katherine and Mitchell operate competing lemonade stands on a beach, earning equal profits. If Katherine buys a sunshade, she can steal some of Mitchell’s customers, more than enough to pay for the sunshade. But if Mitchell also buys a sunshade, the profit shares don’t change, and no additional customers exist. Both spend money for no additional return.

All this is to suggest, (a) measuring ROI is complicated, (b) for something like HPC, it may be a flawed question, and (c) assuming companies are rational, the true ROI of HPC investments may be a healthy 20% to 40% (more than two orders of magnitude less than reported by Hyperion’s study).

What Can We Measure?

Companies that invest in HPC tend to continue to invest in HPC. There must be a financial reason why. This importance of HPC to U.S. industry was addressed in 2014 in a landmark Intersect360 Research study through the U.S. Council on Competitiveness, “Solve. The Exascale Effect: The Benefits of Supercomputing Investment for U.S. Industry.”

In that study, 86% of respondents agreed with the statement, “HPC is critical to the future direction of our business.” 79% agreed with “For some of our problems, HPC methods are the only practical solution.” (See Figure.)  But despite the significant assertion of value, companies often found it challenging to justify HPC investments internally, and they put forward a diversity of arguments for the justification. 19% of the time it’s ROI, but this is mixed in with time to solution, inability to solve the problem by other means, and lowered costs, among others. (See Figure.)

This left me to ponder other evidence to prove the value of HPC. It turns out, Intersect360 Research has already been running a study on this, for more than ten years.

Intersect360 Research has been surveying HPC users regularly since our founding (as Tabor Research) in 2007. In most surveys, we gather HPC budget data, in bands of spending, for demographic purposes. There are hundreds of organizations that have answered our surveys multiple times. Many of them are publicly-traded companies. In total, we identified 58 public companies that have provided HPC budget data to Intersect360 Research in more than one year, 2007 through 2018. (This excludes several Indian companies traded on the NSE, BSE, or OTC of India, which resulted in dramatic outliers. Furthermore, we exclude 2019 and 2020 responses to allow analysis of post-survey performance.)

In the charts below, we map these companies’ share prices, taking one data point per year, against the Dow Jones Industrial Average, grouping the companies into three categories: those that reported increases or temporary increases in HPC spending, those that reported decreases or temporary decreases in HPC spending, and those with stable HPC spending, according to the budget bands in our surveys. The results are intriguing.

Relative Stock Price by Year with HPC Budget Changes, Intersect360 Research, 2020

Relative Stock Price by Year with HPC Budget Changes – Increasing Budgets Only, Intersect360 Research, 2020

Relative Stock Price by Year with HPC Budget Changes – Stable Budgets Only, Intersect360 Research, 2020

Relative Stock Price by Year with HPC Budget Changes – Decreasing Budgets Only, Intersect360 Research, 2020

I’ll cut to the punch line: Some companies do well, and some not so well. Visually, there’s no correlation between increasing HPC budgets and increasing stock prices. We need to look at the numbers more carefully.

As a group, these are all HPC users. If you invested $1,000 across this portfolio of companies beginning in 2005, it would be worth $2,671 today. Investing in the Dow Jones portfolio would get you $2,590. The HPC users pictured here beat the DJIA by 0.2% per year. On average, the group tracks the market.

The best subgroup is those companies whose budgets increased over the time span. $1,000 invested in that subgroup would be worth $3,347—that’s a measurable 1.7% better than the DJIA.

Can we conclude the increase in budget is the cause of the stock price growth? Not yet.  We can also look at when the growth came: before or after the change in budget?

Across this data set, there are 82 points at which HPC budgets change from their previous bands—41 upward, and 41 downward. (This even distribution, it itself, implies there is no dramatic profit directly gained from HPC investment.) For each group, we looked at what happened to stock price in the two years prior to the reported change, and in the three years afterward.

On average, companies’ share prices do well—better than market performance—in the two years prior to a reported HPC budget increase. Conversely, average share prices are flat—worse than market performance—in the two years prior to a reported HPC budget decrease. There is a correlation between HPC budget changes and recent market performance.

This is no surprise. How many readers of this article have experienced this? Company does poorly one year; budgets get cut the next year. It’s commonplace. But here’s an interesting tidbit: Of those companies who reported budget increases, the recent stock price change was heavily polarized; some companies invested more in HPC when their stock prices were down sharply. This could be seen as an additional investment in HPC in order to change processes or to reduce costs.

Looking forward, however, there is no correlation. The three-year average stock performance of companies after an increase in HPC budget is no different from those following a decrease, a little below average market performance for the relevant years.

What It All Means

If there’s no indication that companies that increase their investments in HPC do better than companies that don’t, does that mean HPC isn’t worth it? Of course not. Just look at the Solve report. HPC is critical to those who use it. For many of these companies, HPC is part of the cost of doing business. At worst, it is the sunshades of the lemonade stands in the example above, spurring innovative competition that benefits consumers.

That said, any report that implies there is inherent automatic profitability in HPC investment (let alone 4,300%) is not only wrong, it’s irresponsible. Fortunately, most companies are managed intelligently, and we can assume that increases in HPC investment, on average, clear their hurdle rates. Companies keep buying HPC because they know it’s worth it, and more companies adopt HPC every year.

Subscribe to HPCwire's Weekly Update!

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

Energy Exascale Earth System Model Version 2 Promises Twice the Speed

October 18, 2021

The Energy Exascale Earth System Model (E3SM) is an ongoing Department of Energy (DOE) earth system modeling, simulation and prediction project aiming to “assert and maintain an international scientific leadership posi Read more…

Intel Reorgs HPC Group, Creates Two ‘Super Compute’ Groups

October 15, 2021

Following on changes made in June that moved Intel’s HPC unit out of the Data Platform Group and into the newly created Accelerated Computing Systems and Graphics (AXG) business unit, led by Raja Koduri, Intel is making further updates to the HPC group and announcing... Read more…

Royalty-free stock illustration ID: 1938746143

MosaicML, Led by Naveen Rao, Comes Out of Stealth Aiming to Ease Model Training

October 15, 2021

With more and more enterprises turning to AI for a myriad of tasks, companies quickly find out that training AI models is expensive, difficult and time-consuming. Finding a new approach to deal with those cascading challenges is the aim of a new startup, MosaicML, that just came out of stealth... Read more…

NSF Awards $11M to SDSC, MIT and Univ. of Oregon to Secure the Internet

October 14, 2021

From a security standpoint, the internet is a problem. The infrastructure developed decades ago has cracked, leaked and been patched up innumerable times, leaving vulnerabilities that are difficult to address due to cost Read more…

SC21 Announces Science and Beyond Plenary: the Intersection of Ethics and HPC

October 13, 2021

The Intersection of Ethics and HPC will be the guiding topic of SC21's Science & Beyond plenary, inspired by the event tagline of the same name. The evening event will be moderated by Daniel Reed with panelists Crist Read more…

AWS Solution Channel

Cost optimizing Ansys LS-Dyna on AWS

Organizations migrate their high performance computing (HPC) workloads from on-premises infrastructure to Amazon Web Services (AWS) for advantages such as high availability, elastic capacity, latest processors, storage, and networking technologies; Read more…

Quantum Workforce – NSTC Report Highlights Need for International Talent

October 13, 2021

Attracting and training the needed quantum workforce to fuel the ongoing quantum information sciences (QIS) revolution is a hot topic these days. Last week, the U.S. National Science and Technology Council issued a report – The Role of International Talent in Quantum Information Science... Read more…

Intel Reorgs HPC Group, Creates Two ‘Super Compute’ Groups

October 15, 2021

Following on changes made in June that moved Intel’s HPC unit out of the Data Platform Group and into the newly created Accelerated Computing Systems and Graphics (AXG) business unit, led by Raja Koduri, Intel is making further updates to the HPC group and announcing... Read more…

Royalty-free stock illustration ID: 1938746143

MosaicML, Led by Naveen Rao, Comes Out of Stealth Aiming to Ease Model Training

October 15, 2021

With more and more enterprises turning to AI for a myriad of tasks, companies quickly find out that training AI models is expensive, difficult and time-consuming. Finding a new approach to deal with those cascading challenges is the aim of a new startup, MosaicML, that just came out of stealth... Read more…

Quantum Workforce – NSTC Report Highlights Need for International Talent

October 13, 2021

Attracting and training the needed quantum workforce to fuel the ongoing quantum information sciences (QIS) revolution is a hot topic these days. Last week, the U.S. National Science and Technology Council issued a report – The Role of International Talent in Quantum Information Science... Read more…

Eni Returns to HPE for ‘HPC4’ Refresh via GreenLake

October 13, 2021

Italian energy company Eni is upgrading its HPC4 system with new gear from HPE that will be installed in Eni’s Green Data Center in Ferrera Erbognone (a provi Read more…

The Blueprint for the National Strategic Computing Reserve

October 12, 2021

Over the last year, the HPC community has been buzzing with the possibility of a National Strategic Computing Reserve (NSCR). An in-utero brainchild of the COVID-19 High-Performance Computing Consortium, an NSCR would serve as a Merchant Marine for urgent computing... Read more…

UCLA Researchers Report Largest Chiplet Design and Early Prototyping

October 12, 2021

What’s the best path forward for large-scale chip/system integration? Good question. Cerebras has set a high bar with its wafer scale engine 2 (WSE-2); it has 2.6 trillion transistors, including 850,000 cores, and was fabricated using TSMC’s 7nm process on a roughly 8” x 8” silicon footprint. Read more…

What’s Next for EuroHPC: an Interview with EuroHPC Exec. Dir. Anders Dam Jensen

October 7, 2021

One year after taking the post as executive director of the EuroHPC JU, Anders Dam Jensen reviews the project's accomplishments and details what's ahead as EuroHPC's operating period has now been extended out to the year 2027. Read more…

University of Bath Unveils Janus, an Azure-Based Cloud HPC Environment

October 6, 2021

The University of Bath is upgrading its HPC infrastructure, which it says “supports a growing and wide range of research activities across the University.” Read more…

Ahead of ‘Dojo,’ Tesla Reveals Its Massive Precursor Supercomputer

June 22, 2021

In spring 2019, Tesla made cryptic reference to a project called Dojo, a “super-powerful training computer” for video data processing. Then, in summer 2020, Tesla CEO Elon Musk tweeted: “Tesla is developing a [neural network] training computer... Read more…

Enter Dojo: Tesla Reveals Design for Modular Supercomputer & D1 Chip

August 20, 2021

Two months ago, Tesla revealed a massive GPU cluster that it said was “roughly the number five supercomputer in the world,” and which was just a precursor to Tesla’s real supercomputing moonshot: the long-rumored, little-detailed Dojo system. Read more…

Esperanto, Silicon in Hand, Champions the Efficiency of Its 1,092-Core RISC-V Chip

August 27, 2021

Esperanto Technologies made waves last December when it announced ET-SoC-1, a new RISC-V-based chip aimed at machine learning that packed nearly 1,100 cores onto a package small enough to fit six times over on a single PCIe card. Now, Esperanto is back, silicon in-hand and taking aim... Read more…

CentOS Replacement Rocky Linux Is Now in GA and Under Independent Control

June 21, 2021

The Rocky Enterprise Software Foundation (RESF) is announcing the general availability of Rocky Linux, release 8.4, designed as a drop-in replacement for the soon-to-be discontinued CentOS. The GA release is launching six-and-a-half months... Read more…

US Closes in on Exascale: Frontier Installation Is Underway

September 29, 2021

At the Advanced Scientific Computing Advisory Committee (ASCAC) meeting, held by Zoom this week (Sept. 29-30), it was revealed that the Frontier supercomputer is currently being installed at Oak Ridge National Laboratory in Oak Ridge, Tenn. The staff at the Oak Ridge Leadership... Read more…

Intel Completes LLVM Adoption; Will End Updates to Classic C/C++ Compilers in Future

August 10, 2021

Intel reported in a blog this week that its adoption of the open source LLVM architecture for Intel’s C/C++ compiler is complete. The transition is part of In Read more…

Intel Reorgs HPC Group, Creates Two ‘Super Compute’ Groups

October 15, 2021

Following on changes made in June that moved Intel’s HPC unit out of the Data Platform Group and into the newly created Accelerated Computing Systems and Graphics (AXG) business unit, led by Raja Koduri, Intel is making further updates to the HPC group and announcing... Read more…

Hot Chips: Here Come the DPUs and IPUs from Arm, Nvidia and Intel

August 25, 2021

The emergence of data processing units (DPU) and infrastructure processing units (IPU) as potentially important pieces in cloud and datacenter architectures was Read more…

Leading Solution Providers

Contributors

AMD-Xilinx Deal Gains UK, EU Approvals — China’s Decision Still Pending

July 1, 2021

AMD’s planned acquisition of FPGA maker Xilinx is now in the hands of Chinese regulators after needed antitrust approvals for the $35 billion deal were receiv Read more…

HPE Wins $2B GreenLake HPC-as-a-Service Deal with NSA

September 1, 2021

In the heated, oft-contentious, government IT space, HPE has won a massive $2 billion contract to provide HPC and AI services to the United States’ National Security Agency (NSA). Following on the heels of the now-canceled $10 billion JEDI contract (reissued as JWCC) and a $10 billion... Read more…

Julia Update: Adoption Keeps Climbing; Is It a Python Challenger?

January 13, 2021

The rapid adoption of Julia, the open source, high level programing language with roots at MIT, shows no sign of slowing according to data from Julialang.org. I Read more…

10nm, 7nm, 5nm…. Should the Chip Nanometer Metric Be Replaced?

June 1, 2020

The biggest cool factor in server chips is the nanometer. AMD beating Intel to a CPU built on a 7nm process node* – with 5nm and 3nm on the way – has been i Read more…

Quantum Roundup: IBM, Rigetti, Phasecraft, Oxford QC, China, and More

July 13, 2021

IBM yesterday announced a proof for a quantum ML algorithm. A week ago, it unveiled a new topology for its quantum processors. Last Friday, the Technical Univer Read more…

The Latest MLPerf Inference Results: Nvidia GPUs Hold Sway but Here Come CPUs and Intel

September 22, 2021

The latest round of MLPerf inference benchmark (v 1.1) results was released today and Nvidia again dominated, sweeping the top spots in the closed (apples-to-ap Read more…

Frontier to Meet 20MW Exascale Power Target Set by DARPA in 2008

July 14, 2021

After more than a decade of planning, the United States’ first exascale computer, Frontier, is set to arrive at Oak Ridge National Laboratory (ORNL) later this year. Crossing this “1,000x” horizon required overcoming four major challenges: power demand, reliability, extreme parallelism and data movement. Read more…

Intel Unveils New Node Names; Sapphire Rapids Is Now an ‘Intel 7’ CPU

July 27, 2021

What's a preeminent chip company to do when its process node technology lags the competition by (roughly) one generation, but outmoded naming conventions make i Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire