HPC Innovator’s Work Spans Two Continents

By Nicole Hemsoth

May 28, 2010

Over the past two decades, Dr. Ashwini Nanda has been at the center of some of the most cutting-edge HPC projects and initiatives in the world.  At IBM’s T.J. Watson Research Center, in New York, Dr. Nanda led the development of the Cell processor-based  systems (QS20, QS21, and QS22 blades) and software technologies for high performance computing. That work culminated in the construction of the Roadrunner supercomputer for Los Alamos National Laboratory. Later, as the head of Computational Research Laboratories (CRL) in Pune, India, he directed the development of the “Eka” HPC cluster that, in 2007, held the title of Asia’s fastest supercomputer. He also established the shared memory systems group at IBM Research, worked on the Amazon superscalar architecture at Texas Instruments, and developed parallel computers for India’s missile defense systems at Wipro, Bangalore.

Recently he moved back to India, where he founded HPC Links, a company that offers parallel programming software tools and services for the global high performance computing community. We got the opportunity to ask Dr. Nanda about how he sees the HPC industry today, how it’s changing, and what led him to start up his company.

HPCwire: Maybe we can start with some of your thoughts about the development of high performance computing in India. What are the principal challenges in building the country’s HPC capability? And how do you see the way forward for HPC and supercomputing in India — for both vendors and users?
 
Ashwini Nanda: India has less than 1 percent of the compute power in the global TOP500 supercomputing list today. This tiny share of the pie is not indicative of the financial and high-tech might of India. The gap indicates there is significant room for growth in HPC infrastructure in the country. The gap one sees today is perhaps due to the lack of broad awareness of the benefits of HPC to scientific/technical research and to business enterprises. This is despite the commendable progress made by organizations such as CDAC, CRL, SERC/IISc and HiPC in spreading HPC awareness in the country.

India’s per capita spending in high performance computing is negligible compared to the US, Japan and Germany, and for that matter compared even to China, Spain and Russia. The Unique ID (Adhaar) project of the government will likely fuel growth in HPC related to data mining, consummer security and national security. The government has procured quite a few HPC clusters for weather and climate prediction which would foster research and application development in these areas. India has very successful government enterprises in nuclear energy, space and defence, which would benefit tremendously from the use of HPC. The US counterparts in these sectors are the prime movers of the HPC industry there. The Indian government has the financial strength and must spend heavily in HPC in order to make these sectors globally competitive.
 
India has vibrant pharmaceutical, financial, entertainment and manufacturing industries, all of whom would gain significant productivity and competitiveness by using HPC. We built the Eka machine at CRL soley with private investment from Tata. The other industrial powerhouses in India have the financial means, and they will likely follow suit once they see the benefits of HPC to their business. Talking to leaders from government, academia and industry across the country, one can sense a growing awareness of the potential of HPC, which I believe will translate into a faster pace of growth in HPC infrastructure and services during the next five years or so in India. When that happens, it will benefit the multinational vendors as well as the local vendors.

HPCwire: What do you think will be the role of cloud computing for HPC users in India?

Nanda: Culturally, and economically, reuse and sharing of resources are well accepted in the Indian society. Once the users discover the utility of HPC, I think sharing resources through cloud platforms will come naturally to them, especially the small and medium scale users of HPC. But more importantly, India could also become a key global host of cloud computing infrastructure and services due to attractive low-cost operations and highly skilled technical manpower.

HPCwire: You had a rather prominent role in the development of Cell processor-based systems and software at IBM. The Cell helped usher in the petaflop era, but overall it looks like the impact of this technology in supercomputing is going to be relatively limited. What do you think the Cell brought to the HPC space and what lessons were learned?

Nanda: The Cell based systems from IBM made two important contributions to HPC technology, besides helping cross the petaflop barrier. First they set the new trend of using hybrid multicore clusters — with CPU-GPU combinations — to build the most cost-effective, power-efficient and best-performing supercomputers. Secondly, creating a software ecosystem to harness the compute power of a revolutionary processor like Cell seemed prohibitive in the beginning. But the Cell systems effort at IBM Research showed that an effective collaboration of government, academia and industry can indeed build a respectable software ecosystem for a new architecture.

It’s true that the momentum around Cell has has been lost, but credit should go to IBM for laying the foundations of a new era that would see the proliferation of hybrid multicore CPU-GPU combination clusters in solving key HPC problems. We are already seeing glimpses of this through the momentum building around such offerings from AMD, Intel and NVIDIA.

HPCwire: In HPC, which hardware and software technologies do you see becoming increasing important over the next, say, five years?

Nanda: I believe in terms of hardware, hybrid multicore GPU-CPU combination technologies and commodity InfiniBand and Ethernet technologies will continue to lead the way and take us through the exaflop mark. In terms of software we will see more emphasis being put on tools that make it easier to write parallel applications and increase productivity.

HPCwire: Switching to your current role as the founder and CEO of HPC Links: What is the company about and what was your motivation to launch this as a business?

Nanda: While doing the systems work at IBM and at Tata over the last few years, the obvious realization came to me that now the commoditization of HPC platforms is complete. We have been building affordable parallel machines all around the world, and almost any one who needs access to a parallel machine can access one today. But are these machines utilized well enough? Not really. Are most people, or industries, who could benefit from HPC, taking advantage of these platforms today? The answer is clearly no. So what is the problem? The US government Council on Competitiveness identified two years back that there are three primary barriers to mass adoption of HPC. Namely, lack of parallel programming skills, lack of parallel applications, and high cost of adoption.

HPC Links was formed last year to help address these customer pain points. Our goal is to help businesses stay competitive by alleviating these pain points and achieving high efficiency, faster time to market and enhanced product quality through innovative use of HPC, cloud and multicore solutions. We are addressing all the three key barriers in HPC adoption through our parallel application services offerings, system integration offerings and the software tools and packages under development.

HPCwire: What is unique about the company?
 
Nanda: Our uniqueness today, if I can point to the most significant one, is our interdisciplinary skill pool. We have Ph.D.’s and Masters in multitude of scientific and engineering disciplines, all adept at various flavors of parallel programming, on all kinds of hardware and software platforms. Mix that with the breadth of domain knowledge the HPC Links team has across industry verticals, and you get a really unique, comprehensive parallel programming skill pool in the world.

HPCwire: What do you see as the principal challenges in developing parallel applications for science and technical computing codes today?

Nanda: The hard challenges in parallel application development as I see are lack of skills, and productive tools. Tackling the challenge of productive tools for parallel programming, I believe, will take longer than tackling the issue of lack of skills. There is the general lack of parallel programmers in the world today, which the universities have started to address pretty effectively. And then there is the problem that people who are good in parallel programming are not necessarily trained in various application domains, and people who are domain experts are often times not proficient in parallel programming. The key is to bring domain experts and parallel programmers together and cross train them on the job.

HPCwire: Who do think will be your main customers for these services?

Nanda: In the near term we see majority of our customers outside of India where the HPC market has higher momentum and awareness — especially in the US and Europe. The Indian HPC market is in its nascent stage, but has great potential in the longer term. Our services are targeted toward any one who can benefit from an experienced parallel programming skill pool in research and development, scaling, testing and optimization of parallel applications in a wide range of domains. For example, recently we signed up with Microsoft to provide this kind of parallel application programming services for their HPC Server platform. This makes our services available to the Windows HPC Server users and ISVs in all industry segments. We have deep Linux cluster programming expertise, and are working with universities and national labs, as well as industry HPC users.

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!

Quantum Watchers – Terrific Interview with Caltech’s John Preskill by CERN

July 17, 2024

In case you missed it, there's a fascinating interview with John Preskill, the prominent Caltech physicist and pioneering quantum computing researcher that was recently posted by CERN’s department of experimental physi Read more…

Aurora AI-Driven Atmosphere Model is 5,000x Faster Than Traditional Systems

July 16, 2024

While the onset of human-driven climate change brings with it many horrors, the increase in the frequency and strength of storms poses an enormous threat to communities across the globe. As climate change is warming ocea Read more…

Researchers Say Memory Bandwidth and NVLink Speeds in Hopper Not So Simple

July 15, 2024

Researchers measured the real-world bandwidth of Nvidia's Grace Hopper superchip, with the chip-to-chip interconnect results falling well short of theoretical claims. A paper published on July 10 by researchers in the U. Read more…

Belt-Tightening in Store for Most Federal FY25 Science Budets

July 15, 2024

If it’s summer, it’s federal budgeting time, not to mention an election year as well. There’s an excellent summary of the curent state of FY25 efforts reported in AIP’s policy FYI: Science Policy News. Belt-tight Read more…

Peter Shor Wins IEEE 2025 Shannon Award

July 15, 2024

Peter Shor, the MIT mathematician whose ‘Shor’s algorithm’ sent shivers of fear through the encryption community and helped galvanize ongoing efforts to build quantum computers, has been named the 2025 winner of th Read more…

Weekly Wire Roundup: July 8-July 12, 2024

July 12, 2024

HPC news can get pretty sleepy in June and July, but this week saw a bump in activity midweek as Americans realized they still had work to do after the previous holiday weekend. The world outside the United States also s Read more…

Aurora AI-Driven Atmosphere Model is 5,000x Faster Than Traditional Systems

July 16, 2024

While the onset of human-driven climate change brings with it many horrors, the increase in the frequency and strength of storms poses an enormous threat to com Read more…

Shutterstock 1886124835

Researchers Say Memory Bandwidth and NVLink Speeds in Hopper Not So Simple

July 15, 2024

Researchers measured the real-world bandwidth of Nvidia's Grace Hopper superchip, with the chip-to-chip interconnect results falling well short of theoretical c Read more…

Shutterstock 2203611339

NSF Issues Next Solicitation and More Detail on National Quantum Virtual Laboratory

July 10, 2024

After percolating for roughly a year, NSF has issued the next solicitation for the National Quantum Virtual Lab program — this one focused on design and imple Read more…

NCSA’s SEAS Team Keeps APACE of AlphaFold2

July 9, 2024

High-performance computing (HPC) can often be challenging for researchers to use because it requires expertise in working with large datasets, scaling the softw Read more…

Anders Jensen on Europe’s Plan for AI-optimized Supercomputers, Welcoming the UK, and More

July 8, 2024

The recent ISC24 conference in Hamburg showcased LUMI and other leadership-class supercomputers co-funded by the EuroHPC Joint Undertaking (JU), including three Read more…

Generative AI to Account for 1.5% of World’s Power Consumption by 2029

July 8, 2024

Generative AI will take on a larger chunk of the world's power consumption to keep up with the hefty hardware requirements to run applications. "AI chips repres Read more…

US Senators Propose $32 Billion in Annual AI Spending, but Critics Remain Unconvinced

July 5, 2024

Senate leader, Chuck Schumer, and three colleagues want the US government to spend at least $32 billion annually by 2026 for non-defense related AI systems.  T Read more…

Point and Click HPC: High-Performance Desktops

July 3, 2024

Recently, an interesting paper appeared on Arvix called Use Cases for High-Performance Research Desktops. To be clear, the term desktop in this context does not Read more…

Atos Outlines Plans to Get Acquired, and a Path Forward

May 21, 2024

Atos – via its subsidiary Eviden – is the second major supercomputer maker outside of HPE, while others have largely dropped out. The lack of integrators and Atos' financial turmoil have the HPC market worried. If Atos goes under, HPE will be the only major option for building large-scale systems. Read more…

Everyone Except Nvidia Forms Ultra Accelerator Link (UALink) Consortium

May 30, 2024

Consider the GPU. An island of SIMD greatness that makes light work of matrix math. Originally designed to rapidly paint dots on a computer monitor, it was then 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…

Shutterstock_1687123447

Nvidia Economics: Make $5-$7 for Every $1 Spent on GPUs

June 30, 2024

Nvidia is saying that companies could make $5 to $7 for every $1 invested in GPUs over a four-year period. Customers are investing billions in new Nvidia hardwa Read more…

Nvidia Shipped 3.76 Million Data-center GPUs in 2023, According to Study

June 10, 2024

Nvidia had an explosive 2023 in data-center GPU shipments, which totaled roughly 3.76 million units, according to a study conducted by semiconductor analyst fir Read more…

AMD Clears Up Messy GPU Roadmap, Upgrades Chips Annually

June 3, 2024

In the world of AI, there's a desperate search for an alternative to Nvidia's GPUs, and AMD is stepping up to the plate. AMD detailed its updated GPU roadmap, w Read more…

Some Reasons Why Aurora Didn’t Take First Place in the Top500 List

May 15, 2024

The makers of the Aurora supercomputer, which is housed at the Argonne National Laboratory, gave some reasons why the system didn't make the top spot on the Top Read more…

Intel’s Next-gen Falcon Shores Coming Out in Late 2025 

April 30, 2024

It's a long wait for customers hanging on for Intel's next-generation GPU, Falcon Shores, which will be released in late 2025.  "Then we have a rich, a very Read more…

Leading Solution Providers

Contributors

Google Announces Sixth-generation AI Chip, a TPU Called Trillium

May 17, 2024

On Tuesday May 14th, Google announced its sixth-generation TPU (tensor processing unit) called Trillium.  The chip, essentially a TPU v6, is the company's l 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…

IonQ Plots Path to Commercial (Quantum) Advantage

July 2, 2024

IonQ, the trapped ion quantum computing specialist, delivered a progress report last week firming up 2024/25 product goals and reviewing its technology roadmap. 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…

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…

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…

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…

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…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire