NVIDIA and Folding@home have teamed up to find a cure for diseases using distributed computing. Find out what NVIDIA has in store for the future by reading our Q&A at Bjorn3D.com.
Folding@home is a distributed computing project run by Vijay Pande and the Pande Group at Stanford Univerity where “people from throughout the world download and run software to band together to make one of the largest supercomputers in the world.” The Folding@home Executive Summary’s stated goal is to “understand protein folding, protein aggregation, and related diseases” by using “novel computational methods and large scale distributed computing, to simulate timescales thousands to millions of times longer than previously achieved. This has allowed us to simulate folding for the first time, and to now direct our approach to examine folding related disease.”
Folding@home’s Executive Summary explains the folding of proteins this way:
The Proteins are biology’s workhorses — its “nanomachines.” Before proteins can carry out their biochemical function, they remarkably assemble themselves, or “fold.” The process of protein folding, while critical and fundamental to virtually all of biology, remains a mystery. Moreover, perhaps not surprisingly, when proteins do not fold correctly (i.e. “misfold”), there can be serious effects, including many well known diseases, such as Alzheimer’s, Mad Cow (BSE), CJD, ALS, and Parkinson’s disease.
NVIDIA and CUDA
So where does NVIDIA come into play with all this folding and what is CUDA? CUDA (Compute Unified Device Architecture) is a C programming language compiler and set of devolpment tools developed by NVIDIA that enables programmers to unlock the processing power of GPU’s.
CUDA allows developers to solve the most complex compute-intensive challenges using the native instruction set and memory of the parallel computational elements of CUDA GPU’s. CUDA can also be used to accelerate non-graphical applications, one such being Folding@Home.
NVIDIA and Folding@home
Last June, Stanford University released a Folding@home client specifically for NVIDIA GPU’s. The NVIDIA GPU client was developed using NVIDIA CUDA and it quickly delivered more processing power than any other architecture in the history of the project.
Some folders have taken the project to heart and have built folding farms that are comprised of numerous PC’s containing multiple GPU’s. Many of those folders prefer NVIDIA GPU’s due to their massive processing power and their high output of work units (WU’s) for the team at Stanford.
Click to Enlarge
Folding with Bjorn3D
Bjorn3D.com has many forum members that fold for a cure. They do it for many reasons. Some do it for fun or for the prizes that they might win in our Distributive Computing Forum. Some do it just for themselves or for someone they love. And some do it for the benefit of all mankind. For whatever the reason is that they fold, we at Bjorn3D are proud of our folding members.
We are also happy and proud to announce that Michael Steele from NVIDIA has taken the time to answer a few questions for us that we here at Bjorn3D have about NVIDIA’s contribution to folding and its plans for the future. Before we learn a little bit about Michael and get to the Q&A, let’s take a moment to learn a little bit about NVIDIA first.
NVIDIA (Nasdaq: NVDA) is the world leader in visual computing technologies and the inventor of the GPU, a high-performance processor which generates breathtaking, interactive graphics on workstations, personal computers, game consoles, and mobile devices.
NVIDIA serves the entertainment and consumer market with its GeForce® products, the professional design and visualization market with its Quadro™ products, and the high-performance computing market with its Tesla™ products.
These products are transforming visually-rich and computationally-intensive applications such as video games, film production, broadcasting, industrial design, financial modeling, space exploration, and medical imaging.
About Michael Steele
“Michael Steele is General Manager of the NVIDIA’s Visual Consumer Solutions group. Steele has nearly twenty years of high-tech marketing experience in the PC and networking hardware and semiconductor industry. Prior to NVIDIA, Steele was VP of Marketing at AGEIA leading PhysX software and hardware technology enabling immersive video gaming with real-time physical simulations.”
|What are NVIDIA’s future plans or goals with Folding@Home and GPU computing in general?
We have made a big impact in the Folding@home community in a very short period of time. It is easy to forget that folding on NVIDIA GPUs is still in its infancy. Near term, we are looking to fine tune our performance and make it easier to use multiple GPUS for Folding@home. We continue to work closely with the Pande Lab at Stanford University to improve parallel processing performance with CUDA. We also want to work with them to raise awareness and encourage participation.
In addition to Folding@home, we are also looking at how the power of the GPU can benefit other worthwhile distributed computing programs, such as the ones that run under the BOINC program. The GPU is a natural fit for distributed computing applications and we expect to see many more take advantage of NVIDIA technologies.
GPU computing with CUDA technology is taking off. We have seen breakthrough performance from a variety of visual computing applications from innovative software providers such as Adobe, Elemental Technologies, MotionDSP, Cyberlink and Arcsoft. Adobe just recently released its new CS4 (Photoshop, After Effects, Premier), which for the first time enables GeForce GPU accelerated features that are getting rave reviews. We have seen great success with distributed computing projects such as Folding@home as well as many other exciting scientific and socially responsible research projects that take advantage of parallel processing. We will continue to lead the graphics industry in to this new era of GPU computing and we hope the other graphics companies join the cause soon.
|Could additional emphasis be given in future graphics card designs specifically regarding CUDA and GPU computing performance, in addition to just pure gaming performance?
|We have been emphasizing the compute potential of NVIDIA GPUs for several generations already by enabling CUDA on every GPU we ship. As of today, NVIDIA has shipped over 100 million CUDA-accelerated GPUs. And NVIDIA has been consistently enabling the CUDA platform for GPU computing for several years. There are now over 250 developers who have publicly announced CUDA projects and more than 40 universities worldwide teaching CUDA as a key element of their curriculum. CUDA is a big factor in our GPU designs going forward. The GPU is the processing architecture for the future of visual computing.
|NVIDIA hasn’t been shy about talking about their upcoming Forceware 180 “Big Bang 2” drivers in regards to the features and performance improvements they will deliver. Might Folding@home projects also see any sort of moderate or substantial improvements from the Forceware 180 drivers?
|It is commonly held that NVIDIA GPUs are the fastest processors for folding. Folding@home is already optimized with existing drivers and beats CPU and ATI GPUs by a wide margin. As we said before, we continue to work closely with the Pande Lab at Stanford University to improve parallel processing performance with CUDA.
|One topic of extreme interest to our team and Folding@Home users in general is multi-GPU folding. Are there any plans in the works (or future plans) with Pande Group to help allow easier use of multiple GPU folding on individual systems?
This is an area we know the community wants us to improve in so we are exploring it, and we are allocating resources to address it. We are working with the Panda Group to tune the client and our driver.
However, the folding community is great and they have stepped up to the plate for us. There are a lot of very good guides out there that will walk users through the required steps to fold with multiple GPUs like the ones on HardOCP or overclockers.co.uk, just not in SLI mode yet.
NVIDIA SLI is a great extension to parallel processing and we’re looking at methods to take advantage of it with Folding@home. Stay tuned.
One thing users in particular are hoping for is an easier way to fold on multiple GPUs, or even two similar GPUs in SLI. The Forceware 180 drivers are expected to bring many needed improvements to SLI’s capabilities; could these drivers benefit SLI users that wish to fold on both (or more) of their NVIDIA GPUs?
|Again, this is an area we are looking to improve on. Until then, good guides to accomplish this are out there. Like this one.
|It is somewhat of an odd curiosity, but Folders have long noticed that with Windows XP Folding@home requires 100% of a CPU core dedicated for driver overhead. When folding under Vista only a small fraction of a single CPU core is needed for the same driver overhead, yet both operating systems offer comparable folding performance. Can you shed any light on this?
|Even though we made a huge splash and have been very impactful to performance, NVIDIA and our GPUs are still very new to Folding@home and we have a lot to learn. We’re constantly optimizing across both platforms, but we can’t say specifically yet how this issue will be addressed.
Is there anything you might wish to add or enumerate on to the Bjorn3D folding team and other Folding@home users?
We are thrilled with the reception NVIDIA’s folding efforts has had in the folding@home community. We are pleased to see the impact our work has had on the top folding teams, and we are pleased to have our own team approaching the top 30 after only 6 months. We’re glad that the compute power of the GPU can be put to such good use with distributed computing grids that help find cures and solve scientific problems that matter to society.
CUDA is not just a raw performance enabler; it’s also a proven technology that allows developers to create and scale in performance and features. It’s evolved consistently for several years and we’ve seen amazing industry-changing applications in High Performance PC workstations applications and SuperComputing markets. And now we’re starting to see many astounding new visual computing applications for consumers powered by CUDA-accelerated GPUs. This is just the beginning of a new era in computing.
Folding@home was born out of our CUDA technology and looking for new ways to harness the power of the massively parallel processing power of our GPUs. We hope that consumers will place a value on the things NVIDIA GPUs do outside of adding more frame per second to games. Things like CUDA, SLI, stereoscopic 3D and Physx should be a factor when they are deciding what GPU to buy or to recommend. Graphics is not enough in the visual computing age, you need “graphics plus”.
In just a short period NVIDIA has contributed immensely to the advancement of the Folding@home project as a whole, and it is great to see companies such as NVIDIA supporting these efforts. We wish to thank both NVIDIA and you for your time!
Do You Want To Know More?
NVIDIA Achieves Monumental Folding@Home Milestone With Cuda
“SANTA CLARA, CA—AUGUST 26, 2008—NVIDIA GPUs are contributing over 1 petaflop of processing power to Stanford University’s Folding@home distributed computing application as of last week, according to the statistics published by Stanford. Active NVIDIA® GPUs deliver over 1.25 petaflops, or 42% of the total processing power of the application which seeks to understand how proteins affect the human body…”
To read more, click here.
NVIDIA Taps Processing Power Of GeForce GPUs For More Than Graphics
“SANTA CLARA, CA—August 12, 2008—Consumers want blazing fast performance—whether blasting their way through the latest game or being socially responsible and sharing their PC’s processing power to help find cures for diseases. Today, NVIDIA Corporation, the worldwide leader in visual computing technologies, just made this easier by releasing a set of non-graphics applications that utilize the power of its GeForce® graphics cards. Included in the GeForce Power Pack are Stanford University’s Folding@home distributed-computing, protein-folding client and a trial version of Elemental Technologies’ Badaboom video transcoder. Available for download today at no-cost at www.nvidia.com/theforcewithin, these are part of a growing number of applications that use the power of NVIDIA GeForce® graphics processing units (GPU) and NVIDIA® CUDA™ C-programming technology to significantly improve the performance of non-graphics applications by transferring the workload from the CPU to the more efficient GPU…”
To read more, click here.
NVIDIA Dramatically Accelerates the Search for a Cure
“SANTA CLARA, CA—JULY 24, 2008— Stanford University’s distributed computing program Folding@home has become a major force in researching cures to life-threatening diseases such as cancer, cystic fibrosis, and Parkinson’s disease by combining the computing horsepower of millions of processors to simulate protein folding. The Folding@home project is the latest example in the expanding list of non-gaming applications for graphics processing units (GPU). By running the Folding@home client on an NVIDIA® GeForce® GPU, protein-folding simulations can be done 140 times faster than on some of today’s traditional CPUs…”
To read more, click here.
 A flop is a floating point operation per second, a standard measure of processing power. A teraflop is 1,000 billion flops and a petaflop is 1,000 trillion flops.