To Build or to Buy Time: That is the Question

By Nicole Hemsoth

August 11, 2010

Generally, when one thinks about the vast array of small to medium-sized businesses deploying a cloud to handle peak loads or even mission-critical operations, the idea that such a business might be designing the future of missile defense strategy isn’t the first thing that comes to mind. After all, SMB concerns have historically not had much in common with those of large-scale enterprise and HPC users. The cloud is creating a convergence of these spaces and smaller businesses that were once unable to gain a foothold in their market due to high infrastructure start-up costs are now a competitive force due to the availablity of shared or rented infrastructure and a virtualized environment. This convergence creates new possibilites but can complicate end user decision-making about ideal options for mission-critical workloads.

Analytical Services, Inc. (ASI), a U.S. Department of Defense Missile Defense Agency subcontractor recently used Sabalcore’s high performance computing (HPC) on-demand services to design aerospike nozzles for use in missile systems. These developments in aerospikes represent a significant improvement from a design perspective but required enormous compute power to bring them to market. Orlando, Florida-based Sabalcore, a relatively small company, was able to provide the Linux cluster required for the task while allowing ASI to eliminate the overhead of investing in their own hardware to meet the design challenges.

According to Joseph D. Sims, Technical Director of Engineering at ASI, “Computational fluid dynamics (CFD) is critical to our design efforts, which means we cannot complete that design without Sabalcore’s Linux cluster. We, like many small businesses, cannot afford the luxury of buying and maintaining our own.” Sims went on to note that as with other design projects requiring high levels of compute power, ASI’s goals meshed well with the Linux clusters on-demand because “we could not hope to support our design efforts with CFD running on a serial computer (e.g., a desktop or workstation).” ASI’s Technical Director stated that following comparisons of buying and maintaining a cluster versus buying the access to the Linux cluster, there was “a huge cost savings” that could be realized.

Dividing Line on Building Versus Buying Time?

Gauging from conversations with vendors and end users alike, it is this investment avoidance, coupled with the on-demand nature that makes HPC on-demand services like those offered by Sabalcore and a handful of others (Cycle, Penguin, rSystems, SGI, etc.) attractive. This, along with the fact that HPC on-demand providers tout their high level of personalized support makes this an attractive option—sometimes more attractive than a public cloud.

One has to wonder where the dividing line is for those making decisions about buying versus renting time via an on-demand service—all coupled with the added possibility of the cloud. For some it is about price, for others, it’s rooted in performance goals, for others security. There are no hard and fast rules of thumb for end users but it might seem more attractive to take someone else’s cluster for a guided spin versus tweak applications to suit a cloud that might not yet have proved itself as a viable option.

So where does the cloud fall short when it’s decision time for end users to make the crucial build or buy decision in a case like ASI’s? In an email interview, co-founder of Sabalcore, John Van Workum was asked if there was any tension or cause for competitive concern between HPC on demand services like his company’s and a service like the newly-announced Cluster Compute Instances from Amazon, which are aimed at the same market—those who require HPC-like capacity to run complex or particularly resource-hungry applications. Van Workum stated:

Providers like Amazon have the advantage when it comes to sheer size. They have vast web, storage, and compute resources that a user can tap into. But, HPC boils down to performance. How fast will my application run and how much will it cost are the two biggest questions. It will be interesting to see if Amazon’s new HPC instances will be popular with the HPC user base community.

Because of Amazon’s virtualization layers, the end user is not getting near 100% of the bare-metal performance from a server. Their upgraded 10GigE network for the  HPC instances is an improvement over previous offerings, but DDR and QDR InfiniBand are proven faster. Also, I believe Amazon has restrictions in place when it comes to the number of cores an HPC instance can have at any given time.  Sabalcore, on the other-hand, has a purpose built HPC systems with very few restrictions. Of course, customer service and technical support sets us apart from large HPC cloud providers.

HPC On-Demand Versus an HPC Cloud

ASI like many other small to mid-sized enterprises who have occasional spikes in need for HPC resources are faced with the decision between building or buying time. Performing a careful cost analysis of such a decision is difficult and fraught with uncertainty for new users when there is a cloud option available to contend with as well. However, the problem is that many HPC on-demand companies like Sabalcore are taking the cloud approach with their marketing message and might be adding to confusion by muddling the concept of what a cloud is—and is not.

In fact, the very term “cloud” is problematic for a company like Sabalcore since what they’re providing is not really a cloud at all. While they certainly recognize this, companies with essentially the same offerings are putting the word “cloud” on HPC on-demand services, which adds to confusion, especially for new users who are far more concerned with keeping with their research and time-to-market goals than arguing over complex, hotly-debated definitions. In Van Workum’s view;

Cloud is such a broad term and it’s definition has been discussed in detail and I don’t believe it has one, all encompassing, definition.

One could consider us cloud simply because we host services on the internet. But it pretty much ends there. HPC has very little to do with web-based desktop tools, virtual storage, virtual servers, cloud files, and nebulous virtual  environments which are synonymous with “cloud” these days. We are none of those things either. So therefore we avoid using the term “cloud” when describing Sabalcore.

With this in mind, Workum also provided some commentary on those who are offering the same HPC on-demand service and how a company can differentiate itself in the face of new cloud offerings and competitors. While his detailed response is below, it should be noted that he hits on exactly the same core themes that have emerged in recent conversations with companies like Penguin about its P.O.D service, rSystems, and a host of others. On Sabalcore and the landscape for HPC on-demand companies Workum noted:

HPC users that are familiar with traditional Linux cluster environments will find our environment very similar. We have a very low learning curve. The end user is not hassled by managing instances, insufficient web interfaces, or third party products. Often, a customer is running their job in a matter of hours after logging in for the first time.

Not every application fits nicely into an HPC environment. We provide each new customer with adequate evaluation time and hand holding assistance should they require it.

Our engineers have experience working with hundreds of different applications and can usually make the required modifications in a matter of hours. It is important to note that we almost always adjust the customer’s computing environment in such a way that the changes are as transparent as possible to the customer. It is very uncommon for us to require that the customer make more than superficial changes to their applications or data. But when that does occur, we have the experience to either do it for them or to guide them with the modifications.

Experience and exceptional technical and customer support define us. Sabalcore is a 100% HPC as a service provider and has been since its inception in 2000. We focus solely on our service rather than also selling hardware unlike some recent HPC cloud participants.

In his line of thinking, the cloud is hindered by its lack of support, which is part of the reason why some companies opt for HPC on-demand services versus a public cloud like Amazon’s EC2—even with its new HPC-geared instance type.

Sabalcore has experienced solid growth in the last four years, in part because it has been able to appeal to those who rejected the cloud as an option and who have certainly rejected the option of investing in their own clusters for more obvious reasons. As the cloud, especially public cloud offerings, are developed to be more in tune with the needs of companies like ASI, however, the cloud might push HPC on-demand providers to emphasize even more fervently the support and personalization aspects that go hand-in-hand with their alternative.

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!

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing power it brings to artificial intelligence.  Nvidia's DGX Read more…

Call for Participation in Workshop on Potential NSF CISE Quantum Initiative

March 26, 2024

Editor’s Note: Next month there will be a workshop to discuss what a quantum initiative led by NSF’s Computer, Information Science and Engineering (CISE) directorate could entail. The details are posted below in a Ca Read more…

Waseda U. Researchers Reports New Quantum Algorithm for Speeding Optimization

March 25, 2024

Optimization problems cover a wide range of applications and are often cited as good candidates for quantum computing. However, the execution time for constrained combinatorial optimization applications on quantum device Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at the network layer threatens to make bigger and brawnier pro Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HBM3E memory as well as the the ability to train 1 trillion pa Read more…

Nvidia Appoints Andy Grant as EMEA Director of Supercomputing, Higher Education, and AI

March 22, 2024

Nvidia recently appointed Andy Grant as Director, Supercomputing, Higher Education, and AI for Europe, the Middle East, and Africa (EMEA). With over 25 years of high-performance computing (HPC) experience, Grant brings a Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HB Read more…

Nvidia Looks to Accelerate GenAI Adoption with NIM

March 19, 2024

Today at the GPU Technology Conference, Nvidia launched a new offering aimed at helping customers quickly deploy their generative AI applications in a secure, s Read more…

The Generative AI Future Is Now, Nvidia’s Huang Says

March 19, 2024

We are in the early days of a transformative shift in how business gets done thanks to the advent of generative AI, according to Nvidia CEO and cofounder Jensen 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…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Houston We Have a Solution: Addressing the HPC and Tech Talent Gap

March 15, 2024

Generations of Houstonian teachers, counselors, and parents have either worked in the aerospace industry or know people who do - the prospect of entering the fi Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. 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…

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…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a 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…

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…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

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…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t 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…

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…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

Intel Won’t Have a Xeon Max Chip with New Emerald Rapids CPU

December 14, 2023

As expected, Intel officially announced its 5th generation Xeon server chips codenamed Emerald Rapids at an event in New York City, where the focus was really o Read more…

IBM Quantum Summit: Two New QPUs, Upgraded Qiskit, 10-year Roadmap and More

December 4, 2023

IBM kicks off its annual Quantum Summit today and will announce a broad range of advances including its much-anticipated 1121-qubit Condor QPU, a smaller 133-qu Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire