After contributing to the fight against Alzheimer's, Sony's Folding@home lead reflects on 'overwhelming' user support

This month, Sony's partnership with Stanford University's Folding@home program comes to an end with a software update for the PlayStation 3. During its involvement, Sony contributed both computational power to research aimed at fighting a degenerative disease and leveraged its gaming expertise to add visualizations to the project, Noam Rimon, director of software engineering at Sony Computer Entertainment Research and Development, told Polygon via email.

Sony's contribution to the project was "aimed at creating pharmaceuticals that would fight Alzheimer's disease," Rimon told Polygon. As that aspect of the Folding@home project wound down, Sony decided to sunset its involvement.

"To keep the project interesting, we would have to create new engines for the new types of research Stanford is currently working on," Rimon wrote. "The engine on PS3 was aimed to cover certain aspects of the research, and that will soon be completed. After much discussion, we decided it was time to retire the project on PS3."

Since 2006, the organizations had collaborated to leverage the console's Cell processor across a distributed network of PlayStation 3 systems, which were supported by users who opted into the program. The systems would crunch complex numbers and help model the complicated "folding" process wherein proteins assemble themselves to perform biological functions.

By increasing their understanding of protein behavior, the researchers hoped to discover new treatments for related diseases. For example, in one study to which Sony's involvement contributed, the Folding@home project researched the "aggregation of amyloid beta (Aβ) peptides [which] plays an important role in the development of Alzheimer's disease" by modeling processor-intensive "molecular dynamics simulations."

After much discussion, we decided it was time to retire the project on PS3.

That was the year that Sony unveiled the PlayStation 3. Underneath the hood, the new console sported something unique. Its Cell microprocessor, born of a collaboration between Sony, Toshiba and IBM, supported an architecture unlike any of its competitors. Sony touted the processor as revolutionary, destined for use in a wide variety of applications and products beyond the home gaming space.

At the same time, Rimon saw an opportunity to put the processor's technology to work outside of consumer electronics. He had been a "longtime fan" of Stanford University's Folding@home project, which since 2001 had been harnessing the power of distributed networked computing. The Cell processor's capabilities seemed uniquely suited to help Stanford in its endeavor.

So in 2006, Rimon picked up the phone and called Professor Vijay Pande, the head of Stanford's Folding@home project, to discuss a partnership.

"I rang up Professor Pande at Stanford University, who was really straightforward on working with us, and we began testing the viability of running protein folding algorithms on our new hardware," he wrote.

Working with Stanford's protocols, Rimon and his team at SCEA R&D whipped up "some quick R&D" projects that soon proved that their efforts could help the Folding@home project.

"The collaboration was pretty simple," he wrote. "Everything that was done on PS3 (including the graphical interface, for example) was pretty much handled by Sony R&D whereas the data collection and distribution was handled by Stanford. We didn't even need too many meetings to accomplish that."

"The Cell processor is unique..."

By the end of 2007, Rimon told Polygon that the distributed network was able to provide a sustainable 1 petaFLOP of active network computational power through networked PlayStation 3s, a feat that the Guinness Book of World Records recognized as the world's most powerful distributed network. At its peak, the Folding@home project on the PlayStation 3 had 50,000 concurrent users.

The Cell processor's technical specifications made Sony's involvement possible, Rimon told Polygon.

"The Cell processor is unique in the way that it has eight auxiliary processors all baked onto one silicon," Rimon wrote.

"Those auxiliary processors (also called SPU — 'Synergistic Processing Unit') are very good at stream processing, as long as you can run [the tasks] in parallel and stream data in and out quickly, which is exactly what we did in our version of Stanford's code," he wrote.

"We also had a lot of fun in the graphical visualization of the project."

Sony treated the project as part of its Corporate Social Responsibility efforts, which aim to "contribute to the realization of a sustainable society." During its involvement, the team at Sony R&D continued to optimize the code it had written. Sony's involvement had ramifications outside of its involvement.

"The same technology was later used to build the project on top of specialized hardware, such as GPUs," Rimon says.

Sony's team also used its gaming prowess to create new visualizations for the project that it hoped would incentivize participants.

"We also had a lot of fun in the graphical visualization of the project," he wrote. "Although similar visualization ideas existed at the time, as video game designers we pushed hard on getting all the visualization in real-time and to allow the user to have a ‘virtual flight' through the field of folded proteins. We also added the globe of the world with dots for each participating machine, spreading a feeling of ‘togetherness,' so users could see they were not alone in the folding world. We had one of our designers come up with that idea, and once it was out, it got really good reception. Sometimes people just accept that as a given and don't realize how much work went into it."

Though Sony is always looking at new opportunities for collaboration, its work with Folding@home has to an end. As he looks back, Rimon is proud of the work that his team accomplished.

"On a more personal note," he wrote, "I would like to say that the participation from our PS3 community has been overwhelming, and I feel really honored to have been able to work on this project. The support I received from within Sony was absolutely positive the entire way, starting from the engineers working on this all the way up to our CEO."