5 ms·

It would be nice if NVIDIA donated compute time on one of their large in-house machines the DGX SuperPOD [1] or the DGX SAturn V [2] (#20 and #67 on the last To

It would be nice if NVIDIA donated compute time on one of their large in-house machines the DGX SuperPOD [1] or the DGX SAturn V [2] (#20 and #67 on the last Top500, resp.) for COVID-19 research. Running simulations in a data-center is far more efficient. [1] https://www.top500.org/system/179691 https://www.top500.org/system/179691 [2] https://www.top500.org/system/178928 https://www.top500.org/system/178928

koheripbal 7y ago
These projects are cool and do contribute to our understanding, but they are not part of the critical path of fighting this pandemic. The vaccine(s) are already in phase one clinical trial testing, as are some important anti-viral treatments which are in phase three clinical trial testing. There is no potential output from this folding project that can accelerate those timelines.
- slizard 7y ago
  Sure, I'm aware that this is fundamental research (likely not on actual protein folding, but haven't looked at the details), which will probably not have short-term outcomes, not on the timescales required for a 1st gen vaccine (perhaps not even 2nd gen). That does however not take away from my main point, if anything it is in support of what I was trying to emphasize: such computational research is better suited for a compute cluster.
  - koheripbal 7y ago
    I guess my point was that corporations devoting large sets of their compute clusters to partake in activity that has no potential to help the current pandemic, isn't going to make much business sense.
    - slizard 7y ago
      I agree. In fact, donating in other ways, e.g. giving free programming, HPC etc. education to more researchers can have a higher impact on the long run.
dahart 7y ago
Sure, agreed. But Folding@Home is setup for home, and both of those machines put together add up to the equivalent of maybe 1k-2k gaming rigs. Putting aside looking for the maximum possible efficiency, crowd sourcing has the potential to scale to many orders of magnitude larger than what can be done in a data center.
- slizard 7y ago
  > But Folding@Home is setup for home, Nothing prevents running the fah client on nodes of a compute cluster -- in fact my colleagues did that (while running a local F@H server), though that was a number of years ago just because they wanted take advantage of the distributed computing facilities provided by the client-server setup and built-in algorithms. > crowd sourcing has the potential to scale to many orders of magnitude larger than what can be done in a data center. Potential it does have, but I am skeptical of the "many orders of magnitude" claim ever having a chance to materialize. I'd love to see a cost / benefit analysis on the effective amount of useful work contributed vs the cost of the same in a data center.
  - dahart 7y ago
    > I am skeptical of the "many orders of magnitude" claim ever having a chance to materialize. The many orders of magnitude has already materialized https://en.wikipedia.org/wiki/SETI@home#Statistics https://en.wikipedia.org/wiki/SETI@home#Statistics “On September 26, 2001, SETI@home had performed a total of 1021 floating point operations. It was acknowledged by the 2008 edition of the Guinness World Records as the largest computation in history.[22] With over 145,000 active computers in the system (1.4 million total) in 233 countries, as of 23 June 2013, SETI@home had the ability to compute over 668 teraFLOPS.[23] For comparison, the Tianhe-2 computer, which as of 23 June 2013 was the world's fastest supercomputer, was able to compute 33.86 petaFLOPS (approximately 50 times greater).”
    - Retric 7y ago
      This is 2013 with most of those computers likely even older than 2013 and likely single core vs 3,120,000 brand new cores on the supercomputer. So, in terms of “raw” flops it’s just a question of more and newer hardware on the supercomputer not really better architecture.

koheripbal 7y ago

These projects are cool and do contribute to our understanding, but they are not part of the critical path of fighting this pandemic. The vaccine(s) are already in phase one clinical trial testing, as are some important anti-viral treatments which are in phase three clinical trial testing. There is no potential output from this folding project that can accelerate those timelines.

slizard 7y ago
Sure, I'm aware that this is fundamental research (likely not on actual protein folding, but haven't looked at the details), which will probably not have short-term outcomes, not on the timescales required for a 1st gen vaccine (perhaps not even 2nd gen). That does however not take away from my main point, if anything it is in support of what I was trying to emphasize: such computational research is better suited for a compute cluster.
- koheripbal 7y ago
  I guess my point was that corporations devoting large sets of their compute clusters to partake in activity that has no potential to help the current pandemic, isn't going to make much business sense.
  - slizard 7y ago
    I agree. In fact, donating in other ways, e.g. giving free programming, HPC etc. education to more researchers can have a higher impact on the long run.

dahart 7y ago

Sure, agreed. But Folding@Home is setup for home, and both of those machines put together add up to the equivalent of maybe 1k-2k gaming rigs. Putting aside looking for the maximum possible efficiency, crowd sourcing has the potential to scale to many orders of magnitude larger than what can be done in a data center.

slizard 7y ago
> But Folding@Home is setup for home, Nothing prevents running the fah client on nodes of a compute cluster -- in fact my colleagues did that (while running a local F@H server), though that was a number of years ago just because they wanted take advantage of the distributed computing facilities provided by the client-server setup and built-in algorithms. > crowd sourcing has the potential to scale to many orders of magnitude larger than what can be done in a data center. Potential it does have, but I am skeptical of the "many orders of magnitude" claim ever having a chance to materialize. I'd love to see a cost / benefit analysis on the effective amount of useful work contributed vs the cost of the same in a data center.
- dahart 7y ago
  > I am skeptical of the "many orders of magnitude" claim ever having a chance to materialize. The many orders of magnitude has already materialized https://en.wikipedia.org/wiki/SETI@home#Statistics https://en.wikipedia.org/wiki/SETI@home#Statistics “On September 26, 2001, SETI@home had performed a total of 1021 floating point operations. It was acknowledged by the 2008 edition of the Guinness World Records as the largest computation in history.[22] With over 145,000 active computers in the system (1.4 million total) in 233 countries, as of 23 June 2013, SETI@home had the ability to compute over 668 teraFLOPS.[23] For comparison, the Tianhe-2 computer, which as of 23 June 2013 was the world's fastest supercomputer, was able to compute 33.86 petaFLOPS (approximately 50 times greater).”
  - Retric 7y ago
    This is 2013 with most of those computers likely even older than 2013 and likely single core vs 3,120,000 brand new cores on the supercomputer. So, in terms of “raw” flops it’s just a question of more and newer hardware on the supercomputer not really better architecture.