Company Offers a New Way to Parallelize Applications

By Michael Feldman

September 30, 2010

Startups always begin with big ambitions and Massively Parallel Technologies (MPT) is no exception. This week, the company unveiled “Blue Cheetah,” which is described as a “total application ecosystem” that aims to revolutionize the traditional software development and distribution model, especially for highly parallel codes. The idea is to turbo-charge the ROI model by automating the application development process, bringing software to market faster, and making it widely reusable.

MPT is not exactly a startup, though. The Colorado-based company has been around since 2000, and is on its third CEO. The first two, company founder Scott Smith and HPC luminary John Gustafson, are now on the board of directors. Former Linux Networx exec Bobbi Hazard is the current CEO and will oversee the reboot of MPT as it rolls out the Blue Cheetah suite of tools. Kevin Howard, the CTO, who has been with the company since its inception, has driven the technical innovation behind the products, especially in regard to the parallelization techniques.

During the company’s early years, MPT offered BLAST-based bioinformatics products and services, based on the company’s “HOWARD” technology. Using some of that early work, as well as additional development performed under a DARPA HPCS program contract, the company developed a new parallel communication technology that Hazard claims is “a huge improvement over MPI and PVM.” According to her, more than half of their 400-plus patent filings are based on this area of the technology.

But Blue Cheetah extends far beyond its novel communication scheme. It encompasses the whole software lifecycle, from design, testing, and development to deployment, licensing and revenue distribution. As Hazard puts it: “Usually you get one set of products to do one thing and another set of products to do another,” says Hazard. “I don’t know of anywhere else where you can get them all together.”

Not all of this is available today, though. What MPT announced this week is a beta version of Blue Cheetah’s software development platform, called Cub. With it, developers can design and develop applications, as well as share code with others in a collaborative fashion. Cub also automatically generates application documentation based on the design. The output from this tool is an executable that can run in a uni-processor environment. If parallelization is desired, that is performed separately further down the toolchain.

The key to Blue Cheetah software development is teasing out an application’s fundamental process elements, which they call kernels, from the control functions. Basically, the idea is to separate the math from the program control logic. During design, the system translates the developer’s kernel specifications directly into executable code. This is not revolutionary in itself; pseudcode-to-code translation has been done with varying degrees of success for decades. In this case though, there is no intermediary programming language like C or Fortran to deal with. The design itself represents the program source.

The control functions encompass the if-then-else and looping constructs that wrap around the kernel invocations. Conveniently, Cub automatically generate all the control functionality itself, again, based on the original design. This functional decomposition not only frees the developer from maintaining any of the control software, it also removes the dependency of the kernel algorithms on the underlying hardware and subsequent parallelization schemes.

Another important side effect to this decomposition is that the algorithms are easier to share among applications. Code reuse is a core element of MPT’s software monetization scheme, and during the design phase, the system points the developer to existing kernels that may apply to his or her application. Matches are based on keyword searches, input/output parameters, dataset similarities, and so on. Anything from an individual FFT algorithm to a complete application library can be shared across applications, taking with it the licensing agreement associated with the original code. The choice of reusing existing kernels versus designing new ones is up to the developer, though.

Once the application design is complete, its licensing is set up. The developer determines the fee structure and sets up the payment scheme for their own code. Blue Cheetah offers both a pay-per-use model and a more traditional licensing model. The intention is to offer applications on-demand via their own 256-node “cloud” cluster, but MPT will also license Blue Cheetah to customers who want to take the whole system in-house.

If kernels are reused in multiple applications, the original developers will get paid for each instance of use. It’s essentially the opposite of the open source licensing model. MPT is hoping to attract both commercial and academic developers, especially those frustrated by the “free software” business model. “A whole ecosystem will be formulated over time, and get larger as more kernels and algorithms become available,” says Hazard.

Blue Cheetah includes special capabilities to help developers parallelize their applications across the multicore/multiprocessor/cluster/grid/cloud computing landscape. That certainly covers high performance computing, but all software applications that require a large-scale computing infrastructure (e.g., cloud computing, business analytics, math-intensive applications, etc.) are fair game. Hazard says they have had early interest from organizations who develop nanotech, biotech and multi-player gaming applications.

Parallelizing applications will be performed by the upcoming Blue Cheetah product called Coalition, which is scheduled for release in the January 2011 timeframe. The tool will take the code developed under Cub and automatically restructure it in such a way as to maximize parallelism, be it for multicore platforms or clusters. How it actually accomplishes this is not clear, although an auto-parallelization feature that bypasses MPI and promises better performance should pique the interest of HPC developers.

Further down the road, a separate product called Savannah will also be available to put a Blue Cheetah app into firmware. This is targeted at users who want maximum performance or are running the types of embedded applications that requires the application to be executed locally.

Another future Blue Cheetah tool, called Spots, will consume existing source code and perform process-control decomposition so that it can be fed into the Cub platform. Once in the system, the application can go through the rest of the Blue Cheetah toolchain, including auto-parallelization. How this code transformation occurs, and what types of source code are deemed consumable, is not defined, but Hazard implied that even legacy MPI codes could be restructured by Spots. This tool also checks incoming code for malware and plagiarism.

Redefining the software development ecosystem is certainly a lot for one small company to take on. MPT has no venture capital money behind it. But the company has attracted a large number of angel investors to fund the Blue Cheetah development.

They’ve also managed to catch the attention of Gene Amdahl, a computer science icon who developed Amdahl’s law of software parallelism. He is on MPT’s board of advisors and appears to be thoroughly impressed by the Blue Cheetah products. In a video on MPT’s website, he talks about the importance of parallel computing and the opportunity afforded by the company’s technology. “It will revolutionize the world of computing,” he says.

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!

Illinois Considers $20 Billion Quantum Manhattan Project Says Report

May 7, 2024

There are multiple reports that Illinois governor Jay Robert Pritzker is considering a $20 billion Quantum Manhattan-like project for the Chicago area. According to the reports, photonics quantum computer developer PsiQu 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 what it is like to orbit and enter a black hole. And yes, it c Read more…

2024 Winter Classic: Meet the Mentors Round-up

May 6, 2024

To make navigating easier, we have compiled a collection of all the mentor interviews and placed them in this single page round-up. Meet the HPE Mentors The latest installment of the 2024 Winter Classic Studio Update S Read more…

2024 Winter Classic: The Complete Team Round-up

May 6, 2024

To make navigating easier, we have compiled a collection of all the teams and placed them in this single page round-up. Meet Team Lobo This is the other team from University of New Mexico, since there are two, right? T Read more…

How Nvidia Could Use $700M Run.ai Acquisition for AI Consumption

May 6, 2024

Nvidia is touching $2 trillion in market cap purely on the brute force of its GPU sales, and there's room for the company to grow with software. The company hopes to fill a big software gap with an agreement to acquire R Read more…

2024 Winter Classic: Oak Ridge Score Reveal

May 5, 2024

It’s time to reveal the results from the Oak Ridge competition module, well, it’s actually well past time. My day job and travel schedule have put me way behind, but I am dedicated to getting all this great content o 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…

How Nvidia Could Use $700M Run.ai Acquisition for AI Consumption

May 6, 2024

Nvidia is touching $2 trillion in market cap purely on the brute force of its GPU sales, and there's room for the company to grow with software. The company hop Read more…

Hyperion To Provide a Peek at Storage, File System Usage with Global Site Survey

May 3, 2024

Curious how the market for distributed file systems, interconnects, and high-end storage is playing out in 2024? Then you might be interested in the market anal Read more…

Qubit Watch: Intel Process, IBM’s Heron, APS March Meeting, PsiQuantum Platform, QED-C on Logistics, FS Comparison

May 1, 2024

Intel has long argued that leveraging its semiconductor manufacturing prowess and use of quantum dot qubits will help Intel emerge as a leader in the race to de Read more…

Stanford HAI AI Index Report: Science and Medicine

April 29, 2024

While AI tools are incredibly useful in a variety of industries, they truly shine when applied to solving problems in scientific and medical discovery. Research Read more…

IBM Delivers Qiskit 1.0 and Best Practices for Transitioning to It

April 29, 2024

After spending much of its December Quantum Summit discussing forthcoming quantum software development kit Qiskit 1.0 — the first full version — IBM quietly Read more…

Shutterstock 1748437547

Edge-to-Cloud: Exploring an HPC Expedition in Self-Driving Learning

April 25, 2024

The journey begins as Kate Keahey's wandering path unfolds, leading to improbable events. Keahey, Senior Scientist at Argonne National Laboratory and the Uni Read more…

Quantum Internet: Tsinghua Researchers’ New Memory Framework could be Game-Changer

April 25, 2024

Researchers from the Center for Quantum Information (CQI), Tsinghua University, Beijing, have reported successful development and testing of a new programmable 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…

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…

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…

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…

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…

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…

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…

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…

Leading Solution Providers

Contributors

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…

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…

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…

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…

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…

A Big Memory Nvidia GH200 Next to Your Desk: Closer Than You Think

February 22, 2024

Students of the microprocessor may recall that the original 8086/8088 processors did not have floating point units. The motherboard often had an extra socket fo Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire