Startup Aims to Transform Computing with Probability Processing

By Michael Feldman

August 16, 2010

MIT spin-out Lyric Semiconductor Inc. has launched a new breed of integrated circuits that replaces the binary logic of traditional computing with probabilistic logic. The aim is to deliver a much more efficient architecture for applications based on probability computing. For these types of workloads, the company is promising orders-of-magnitude improvement in energy efficiency, performance and cost.

Essentially, what Lyric has come up with is a fifth processor architecture, following CPUs, GPUs, DSPs, and FPGAs. The technology grew out of Ben Vigoda’s Ph.D. thesis work at MIT more than 10 years ago. In 2006, Vigoda founded the company, along with IC designer David Reynolds and Analog Devices CEO Ray Stata. Vigoda became the CEO, Reynolds the VP of product development, and Stata took the post of chairman of Lyric’s board (with Stata Venture Partners listed as the lead investor). The company has filed 50 patents to protect its intellectual property and raised over $20 million to get the business off the ground. About $18 million of that funding was injected from DARPA and other government agencies to spur the research and development effort and make probability processing a viable technology.

The goal is to construct hardware circuitry and software purpose-built for probability applications. With conventional digital technology, processing has to follow a strictly linear path. This is fine for software like operating systems, spreadsheets, word processing, and database transactions, where the computing consists of straightforward calculations or data movement. “But most of the interesting things happening nowadays don’t really fit into that model,” says Reynolds. From his perspective, the goal of more and more software today is to find the best fit or most likely answer (or answers) to a problem.

Most data mining, predictive analytics, pattern recognition, and modeling/simulation codes fall into this category. That encompasses a wide range of applications including Web searching, financial modeling, genome sequence analysis, speech recognition, climate modeling, credit fraud detection, spam filtering, and financial modeling, among many others. People tend to associate these probability-based applications with human-like intelligence, and this is clearly where software, in general, is moving.

Even today, the average person uses probability computing on a fairly regular basis. For example, when you hop onto Amazon’s website, the software behind the scenes attempts to predict what product you might be interested in buying, basing its suggestions on past shopping behavior and what you’re currently clicking on. If you do make a purchase, the credit card company’s software tries to determine if the charge is fraudulent or not, again based on past buying habits. Both of these pieces of software are searching for the most likely result, rather than a specific answer. “So these applications are not a great fit for the traditional digital processor as we know it,” says Reynolds.

Probability computing has been around for decades, but most of the work has involved developing languages and algorithms, which are subsequently applied to vanilla digital computers. Lyric’s founders were convinced that they needed to rethink the design of the underlying circuitry to get the optimal solution. What they came up with was the concept of pbits (probability bits), which unlike Boolean bits can represent a lot more than two states. Lyric is not divulging exactly how many states that might be, but according to Reynolds: “Suffice to say that we have all the states we need for the applications we’ve looked at so far.”

To process pbits, Lyric has designed a set of standard probability logic gates that can be connected to one another multi-directionally. So instead of stepping through an application in a linear sequence using Boolean gates, more complex operations, such as getting the intersection of data, can be performed on multiple variables in parallel. That, says Reynolds, is a much more natural way of implementing parallelism than CPUs provide with multicore and multi-threading architectures.

According to him, for probabilistic-type operations, what took 500 transistors with conventional digital logic can be distilled down to just a handful of probability transistors. Lyric predicts that a single probability processor will be able to increase computational capability by two or three orders of magnitude compared to today’s server chips, with commensurate savings in cost, power and space.

The company’s first commercial product will be aimed at advanced error correction for NAND flash memory. Called Lyric Error Correction (LEC), it’s designed to relieve the flash memory ECC bottleneck, which is becoming a constraint as semiconductor process geometries shrink. On the 30nm process, flash memory errors are generated at the rate of 1 in 1,000. On the next generation technology, it will be 1 in 100. That means the ECC component of the controller, which is already the largest piece of the device, may become impractical to implement.

As it turns out advanced ECC logic is based on probabilities, so it’s a nice fit for Lyric-style circuits. The company is promising a 30-fold reduction in die real estate (for 1 Gbps bandwidths) and a 12-fold improvement in power. The company has ported the device to different process nodes using TSMC as the fab partner, and is claiming yields typical for this class of application. LEC is available for licensing today, with product integration expected in 12 months.

But the real payoff for Lyric is at least a few years down the road. The company is developing the GP5, which stands for general-purpose programmable probability processing platform. The technology is aimed at the broader set of probability applications mentioned above, and, according to the company, will be up to 1,000 times more efficient at these types of tasks than current x86 CPUs. The first samples of the GP5 are slated to appear in 2013.

According to Mira Wilczek, Lyric’s director of business development, the initial commercial implementations of the GP5 are likely to be packaged inside an appliance or embedded device, such as a Web search server or a handheld speech recognition device. A more general-purpose use case could involve a GP5 used as a coprocessor in conjunction with a CPU.

For high performance computing work, the latter configuration could be a way to accelerate performance on applications such as materials modeling and whole genome analysis, in much the same way as GPGPUs are being employed today. “Just like a GPU exploits the vector nature of graphics computation, we exploit the typical structures that our algorithms take for probability computation,” explains Wilczek.

The downside to the new architecture is that conventional programming languages, like C, Fortran and Java, are not very good at expressing probability algorithms. Lyric has come up with its own language, called PSBL (Probability Synthesizes to Bayesian Logic) that can be compiled to their hardware. The language is a rule-based language, where the programmer specifies the constraints of the problems, rather than how to solve it. Other probability-type programming languages, like R and the Microsoft’s Infer.NET could also presumably be targeted to Lyric’s architecture.

PSBL 1.0 will be licensed to select partners in Q4 2010 with a second version slated for Q4 2010. In the absence of Lyric hardware, the PSBL code can be compiled to run in a simulator on a conventional computer, albeit much more slowly. When the GP5 arrives in 2013, the hope is to have the foundation of a probability computing ecosystem already in place.

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