Cloud Control: Outsourcing an HPC Cluster

By Scott Clark

August 5, 2010

So, thus far in this series of posts, we have discussed the following issues:

– IT is not a core competency of the business, so we should look to outsource if we can outsource without jeopardizing the business.

– We should look to cloud computing to bring costs under control and to deliver cost efficiencies over time, not as an immediate cost reduction activity.

– In order to outsource IT, we must trust the suppliers and vendors involved, which means developing relationships, not better bludgeoning weapons. And we have already done an extremely similar divestiture in our past, so we have a model to look at that says it can be done successfully

Now we need to talk about what an organization would need to look like in order to properly manage the outsourcing of your HPC cluster. So what would that look like? Well, we should assume that all technical and operational capabilities necessary to execute the infrastructure are included in the outsource. The supplier is expected to provide the entirety of the technical function and carry out all operational duties. That is not to say that the customer is off the hook technically, just the opposite. The customer needs to assemble a small team of technically savvy, business minded (specific to the core product of your company) individuals to measure and manage the outsource. This team needs to be very strong technically in order to vet and gauge any available technologies for potential use as well as identify flaws in solutions or methodology of solutions delivered. The size of the team would be dependent on the size of the company (and therefore the size of the outsource).

Functionally, the outsource management team is the control point for the outsourcing of your infrastructure. Through this group, you maintain control over your infrastructure, and therefore can have full trust in your outsource partner (because you know exactly what you want, and you know how to measure if you are getting it). The intent of this team is to stay abreast of the constantly changing needs of the business, understand the continuously evolving capabilities of technology, and combine the two awareness’ to understand how the company should be leveraging technology to maximize benefit to the business and control costs. With that combined awareness, you now hold the outsource accountable for delivering an appropriate solution to your company’s need.

This is not to say that all responsibility falls to the customer outsource team. The supplier will need to have a disciplined focus in the specific space that your company does business, and be innovating their solution to specifically solve the problems of that industry. If they do not, then they will probably not be a cost competitive, viable supplier long term.

You will see many functions that fall under the customer outsource team. And remember, this team needs to remain small in order to avoid paying too much for your solution. There will be a constant loop for the outsource team to:

1. Quantitatively measure the current solution

2. Analyze cost and benefits of the current solution

3. Assessment of best practices

4. Revision of current solution

5. Loop back to 1

There will be several technical responsibilities that the outsource team will participate in jointly. The supplier should be doing most of this work for the customer, but how do you know if the data they are presenting is 100% accurate or appropriate for your solution. When in doubt, the outsource team will generate their own data, and share that data with the supplier to derive a more accurate solution. In that, the outsource team will do some amount of, but not every facet of:

1. Technical and cost Benchmarking

2. Technical advisory / liaison (IT industry to customer business)

3. Technical architects – Designing architecture of applications and services that are appropriate for the company’s consumption

There are many responsibilities of the outsource team that will fall into the relationship management arena. This team will be the primary point of contact and control between the customer and the supplier, and I can’t say enough how important having a positive relationship with the supplier is to the quality of the product you consumer or the price you pay for that product. The outsource team will be responsible for communicating current and future requirements to the supplier, and many of those will take on the form of Service Level Agreements (SLAs) which we will talk about in a moment. The outsource team will also be responsible for how technology is being consumed by the customer company. The outsource team needs to make sure the company is getting the appropriate solution from the supplier company at an appropriate price with appropriate constraints / limitations / boundaries.

Another very important responsibility of the outsource team will be to maintain flexibility from a quality of solution as well as a cost perspective. In this, staying standards based is very important. It is not an absolute requirement, there may be solutions that are proprietary that solve a problem much more efficiently or cost effectively. What you need to consider in this case is when the vendor thinks they have you locked in, and start raising the price because they think you can’t get out of their solution, what is your plan for defeating them. So, where possible, use industry standards so that you can move from vendor to vendor without losing time, money, or critical features. Where that is not possible, what is the plan for using one vendor’s proprietary solution but being able to migrate to another vendor’s solutions without impact to maintain negotiating position.

Finally, there is the new component to infrastructure management. The outsource team will need to learn how to define and measure service level agreements (SLAs). The definition stage will have several components. What is the service level expectation (defines success and failure criteria)? This will sometimes have many different components for a single solution. An example would be storage: is there enough capacity, do we get enough IOPs, and is there enough throughput. All of these are different measurements, but critical to a storage infrastructure for HPC. How will this service level be measured and how often? We have all seen many improper SLA measurements where IT informs the engineer that they have 99.997% availability of the environment, but the engineer knows that there were several outages that had him or her non-productive for days at a time. So do you measure component level availability or solution level? How frequently are the polls for availability? Is availability the right measure? This is all part of the definition. And then, what happens when a failure criteria is met? This is where there is a lot of work happening in the industry. It is not sufficient to refund the months colo fees when a power outage cost the company 6 weeks worth of work. There is a cost to failure, and that is usually very specific to the industry. An outage on a cluster for an EDA company has different implications than an outage to a scientific computing cluster for a university. The recourse needs to be negotiated based impact. Does this at all sound familiar? Any insurance people reading this?? Well, that is one of the solutions the industry is exploring, is having insurance policies behind the supplier. Finally, we need to look at how service levels re-assessed over time. As the technology evolves, so should the service levels.

The fabless semiconductor industry is fairly mature in it’s process for outsourcing the fabrication function. They have cost models and laws (Rock’s Law for cost of a fab over time) that help decision processes, they have a collaborative (FSA) for arriving at better process, and they have an established track record that this can be accomplished very successfully and with cost benefit. The HPC Cloud industry needs to mature. That will just take time.

 

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!

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion XL — were added to the benchmark suite as MLPerf continues 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 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…

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion 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…

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