OpenACC Starts to Gather Developer Mindshare

By Michael Feldman

May 17, 2012

PGI, Cray, and CAPS enterprise are moving quickly to get their new OpenACC-supported compilers into the hands of GPGPU developers. At NVIDIA’s GPU Technology Conference (GTC) this week, there was plenty of discussion around the new HPC accelerator framework, and all three OpenACC compiler makers, as well as NVIDIA, were talking up the technology.

Announced at the Supercomputing Conference (SC11) last November, OpenACC is an open standard API developed by NVIDIA, PGI, Cray, and CAPS, to provide a high-level programming framework for programming accelerators like GPUs. OpenACC uses compiler directives, which programmers insert into high-level source (e.g., C, C++ or Fortran), to tell the compiler to execute specific pieces of the code on the accelerator hardware.

GTC conference-goers had plenty of opportunity to encounter OpenACC this week. There two OpenACC tutorials for would-be developers, one by NVIDIA, and the other by CAPS enterprise. In addition, there were four other sessions hosted by Cray, CAPS, and PGI throughout the week. That’s not counting the numerous mentions OpenACC got during other presentations involving GPGPU programming.

The technology is still in its infancy though. The PGI and Cray compilers are pre-production versions. CAPS first commercial offering is just two weeks old.

The initial goal of OpenACC is to bring more developers (and codes) into GPU computing, especially those not being served by the lower-level programming frameworks like CUDA and OpenCL. While CUDA is widely used in universities and in the technical computing realm, and OpenCL is emerging as an open standard for parallel computing, neither is particular attractive to commercial developers.

Most programmers are used to writing high-level code that focuses on the problem at hand, without have to worry about the vagaries of the underlying hardware. That hardware independence is also what makes OpenACC attractive for codes that need to span different processor architectures.

That assumes, of course, that compiler will support multiple accelerator chips. The first crop of OpenACC-enabled compilers from PGI, CAPS and Cray only generate code for NVIDIA GPUs — not too surprising when you consider NVIDIA’s current dominance in HPC acceleration. However all of the compiler efforts plan to widen the aperture of hardware support.

CAPS is perhaps most aggressive in this regard. According to CAPS CTO François Bodin, his company plans to add OpenACC support for AMD GPUs, x86 multicore CPUs and even the Tegra 3 microprocessor, an ARM-GPU design that will be used to power an experimental HPC clusters at the Barcelona Supercomputing Center (BSC). Bodin also said that they have an Intel MIC (Many Integrated Core) port of OpenACC in the pipeline. All of these compiler ports should be available later this year.

PGI is keeping its OpenACC development plans a little closer to the vest. But according to PGI compiler engineer Michael Wolfe, they have received requests for OpenACC support for nearly every processor and co-processor used in high performance computing. The compiler maker will undoubtedly be developing some of these over the next year.

Likewise for Cray, although its OpenACC compiler support is focused on the underlying accelerators of its own XK6 supercomputers. At this point, that’s confined to NVIDIA GPUs. Cray (which also carries CAPS and PGI compilers for its customers) has a unique OpenACC offering in that it supports those directives in PGAS languages Co-Array Fortran and Unified Parallel C (UPC) on the XK6.

Besides its applicability to multiple hardware platforms, OpenACC is just plain easier to use when you have lots of existing code. For one thing, OpenACC lets you attack the acceleration in steps. CUDA and OpenCL ports usually require code rewrites of at least a sizeable chunk of the application being accelerated, using low-level APIs. With OpenACC, the programmer just has to insert high-level directives into existing source, and this can be done iteratively, gradually putting more and more of the code under OpenACC control. This, say, PGI’s Wolfe, is “a hell of a lot more productive” than the low-level approach.

Even at the national labs and research centers, where there are computer scientists aplenty, OpenACC is starting to be recognized as an easier path to bring acceleration to hundreds of thousands of line of legacy codes. NASA Ames is already using PGI’s compiler to speed up some of their CFD codes on one of their GPU clusters. And the upcoming deployments of multi-petaflop GPU-based supercomputers like “Titan” at Oak Ridge National Lab, should provide a lot more opportunities for OpenACC-based application development. Titan project director Buddy Bland is on record endorsing the technology for software development on that machine.

As with all parallel programming though, there’s no free lunch to be had. In general, the programmer is probably going to sacrifice some runtime performance (compared to CUDA, for example) for the sake of programmer productivity. But there seems to be a general consensus that intelligent use of directives can easily get you to within 10 or 15 percent the performance of a low-level implementation. But as CAPS’ Bodin explains, to get in that close, “you have to know what you’re doing.” On the other hand, as the compiler technology matures and developers get more adept with OpenACC, the performance gap could narrow even further.

The other problem is just a lack of accelerator diversity at the moment. With Intel MIC waiting in the wings, and AMD still pretty much a no-show with server-side GPUs, there’s no immediate need to support anything but NVIDIA’s GPU architecture right now. Worse, both Intel and AMD are backing other parallel computing frameworks that they are rolling into to their accelerator programs: OpenMP, Cilk Plus, and TBB for Intel; OpenCL and C++ AMP for AMD.

Fortunately, it probably doesn’t matter that Intel and AMD haven’t hopped on the OpenACC bandwagon. PGI and CAPS can still produce compilers targeting Intel MIC or AMD GPUs, or whatever else comes along. And as long as there are at least two compiler vendors offering such support, the community should be satisfied.

The end game, though, is to fold the OpenACC capabilities into OpenMP. If and when that happens, both Intel, AMD will throw their support behind it. OpenMP has been around for 15 years and is a true industry standard.

There is currently a Working Group on Accelerators in the OpenMP consortium, which is looking at incorporating accelerator directives into the next OpenMP release. And while those directives will be based on the OpenACC directives, they are not likely to be adopted as is. There’s a real risk that if the process gets drawn out much longer and OpenACC captures a critical mass of users, there will end up being two directive-based accelerator standards to choose from.

Twas ever thus.

Related Articles

CAPS Entreprise Now Supports OpenACC Standard

OpenMP Announces Improvements for Multicore and Accelerators

OpenACC Support Available With New PGI Accelerator Fortran and C Compilers

NVIDIA Announces Initial Results of Directives-Based GPU Computing Program

NVIDIA Eyes Post-CUDA Era of GPU Computing

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