Mitrionics Brings High-Performance FPGAs to the Masses

By Tim Curns, Editor

August 26, 2005

In efforts to make FPGA computing easier and more accessible to the masses, Mitrionics announced a revolutionary new development platform based on the Mitrion Virtual Processor and Software Development Kit. The new technology allows FPGAs (Field Programmable Gate Arrays) to be programmed faster, easier, and more affordably than current development tools.  The Mitrionics solution does not require any hardware programming or hardware design experience and is the first to achieve true pushbutton software-to-hardware compilation.

Editor Tim Curns spoke with Mitrionics' CEO Anders Dellson to garner further insight into this technology.



HPCwire:
Please describe your company and its place in HPC. What sort of things are you recognized for? Where areas do you hope to make the most impact in the near future?

Anders Dellson: The HPC market and industry has shown very strong interest in the tremendous potential with FPGA-based High Performance Computing for several years. Until now, the main challenge and obstacle has been the difficulty and time required to program the FPGAs.

Mitrionics is the first company to truly make FPGA acceleration possible in a practical sense for HPC. Current and previous FPGA design tools other than Mitrionics are very good at general purpose hardware design, but actually very bad at programming FPGAs for HPC. They must be used by EE hardware designers and their abstraction level is far too low for HPC. And while some of them have come out with C-type languages for programming, they still require extensive hardware design knowledge. So this really isn't really making anything easier or faster.

In contrast, the Mitrion C language allows FPGAs to be programmed without any hardware design knowledge at all. Once the program is written, our platform features a unique, first-in-the-industry, pushbutton software-to-hardware compilation. Using one real-life example, 180 lines of Mitrion C code produced 150,000 lines of VHDL. We can't go into a great deal of technical detail here, but more  information including  a white paper can be found on our Web site at www.mitrionics.com.

What we believe is most exciting about the Mitrion Platform is that it removes one of the major obstacles to making HPC available and accessible to a much broader market and to a much larger group of developers and programmers. Our platform effectively lowers the price, complexity, and development time for HPC. The next six months to one year should be really interesting as new HPC applications are developed and implemented in a wide variety of areas including genomics, proteomics, seismic exploration, signal processing, and financial applications just to name a few.   

HPCwire: Describe the importance of FPGA technology in the HPC space?

Dellson: FPGAs can execute specific applications 10-100x faster than CPUs, which provides a huge cost to performance benefit. And FPGA power consumption is typically less than 1% of a CPU for the same job, providing another large benefit. Lastly, FPGA performance will continue to grow with Moore's Law in the future; whereas CPUs have already hit the wall in terms of clock frequency.

During the last few years, both Cray Inc. and SGI have made significant investments in FPGA-based HPC products. This testifies to their strong belief in the FPGA as an important computing device for the future.

HPCwire: Describe how your product is different than competitive tools?

Dellson: The Mitrion Platform is 100% targeted at supercomputing (software) type of applications, as opposed to general purpose low-level chip design. The Mitrion development environment is true software programming; which is very different from hardware design in terms of being much faster and easier.

The default tools for FPGA design are HDLs (Hardware Design Languages) – Verilog or VHDL. Both are very low abstraction level compared to software programming. They require the designer to be able to design his own application-specific computer on the FPGA surface. Other leading tools on the market require tens of thousands of lines of code and take many man months or even years to develop a relatively simple HPC application. In contrast, HPC applications are what the Mitrion Platform was designed for. These applications only take a few hundred lines of code and days or weeks to program.

While some FPGA programming tools have recently come up with a c-type language, the designer still needs to have detailed knowledge in hardware design. Programming in Mitrion C does not require any hardware design knowledge.

HPCwire: How will these differences make your product more accessible to the market?

Dellson: The true software programming is key. For most HPC programmers, the Mitrion platform is the only solution that will enable them to program FPGAs today. The Mitrion programmer does not need to have a background in FPGA design or reconfigurable computing. The Mitrion C language will be very familiar and easy to learn to programmers worldwide.

This is widely recognized by SGI, Cray, and other HPC vendors. The Mitrion platform will be a key element to make their FPGA products available to the wider HPC market.

HPCwire: What do current products in this field lack? Why is this detrimental to HPC users?

Dellson: Two things are needed for FPGA-based HPC to become a widespread reality. One, off-the-shelf computer hardware including FPGAs with very high I/O capacity. Two, relatively fast and easy-to-use solutions for programming the applications to the FPGAs.

The hardware side of the equation is happening right now. Traditional HPC vendors, Cray launched their XD1 last year and SGI are launching their RASC this year. We expect other vendors to make announcements in this area in the coming year.

The hardware side also includes companies like Nallatech, a leading developer of PCI acceleration boards with FPGAs.

And now with the release of the Mitrion Platform, there is also a fast, easy, and effective programming solution available.

HPCwire: What obstacles have you run into trying to invent your product? How did you overcome these obstacles?

Dellson: Technology — the realization that software and hardware are different things and there is no way making a direct translation between them. This obstacle was solved by introducing the Mitrion Virtual Processor.

Cultural — very few people know both software and hardware design. FPGA designs today are typically about glue logic or signal processing in embedded systems. For these developers using traditional HDLs, an FFT or and MP3 decoder is about the most complex application anybody would ever try to design. So the people that know FPGAs often have little appreciation for the need of higher level programming solutions. On the other hand, software programmers typically do not know the first thing about how hardware works. Most of the time, they do not even have any idea about how the CPU they are programming works. So they would never be willing to learn traditional FPGA design even if it could deliver huge performance wins. This is the gap that the Mitrion platform bridges.

HPCwire: What kind of applications will your technology most effectively advance? Can you give specific examples?

Dellson: Good application areas include genomics, proteomics, imaging, seismic exploration, cryptography, pattern recognition, financial simulations and much more.

Today FPGAs are particularly strong at integer operations and up to single- precision floating point operations. The chips today are at the brink of being too small to get very high performance for double-precision floating-point. However, FPGAs will grow out of this limitation in the next couple of years.

Recent Mitrion based implementations include Markov Chain Monte Carlo simulations of phylogenetic trees (searching for the genomic “Tree of Life”) and image analysis for deciphering protein gels.

HPCwire: Speak a bit about the specifics of the technology involved? How did you make it so much easier for programmers and HPC users?

Dellson: The real secret of the Mitrion platform is the Mitrion Virtual Processor, that enables the true software programming. It allows the solutions the be created in two steps: In the first step, the Mitrion Processor, which is an abstract machine, is programmed in software by the developer. In the second step, a blueprint for an adapted Mitrion Processor is created. The massively parallel processor design consists of exactly the operators needed for the application, highly optimised in an architecture where data streams through the whole design to allow all the units to execute each clock cycle. This blueprint is uploaded into the FPGA – and the acceleration hardware is ready to run.

HPCwire: How will you continue to improve FPGA design? What's in the works for Mitrionics?

Dellson: First of all, we are totally dedicated to FPGA programming for the HPC market. So our promise to HPC programmers is that we will continue to work with their toolflow, not the toolflow of the EDA market.

For HPC users, FPGAs offer a quantum leap in performance increase and a path to continue to gain from Moore's Law as CPUs fail to do so.

For FPGA manufacturers, the HPC market is a completely new market – for their highest-end product lines. We help them develop this new market.

Mitrionics has come up with a working solution to the programming issue, and our goal is to be able to help bring all the benefits of FPGAs to the HPC community.

HPCwire: Thanks for speaking with me, Anders, about this exciting new technology.

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