| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: OpenPandemics - COVID-19 Project
Thread: Let the final security review come second - let us run the GPU project now! |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 15
|
|
| Author |
|
|
Dayle Diamond
Senior Cruncher Joined: Jan 31, 2013 Post Count: 452 Status: Offline Project Badges: 
|
So in Los Angeles, it's estimated that a third of the population has COVID-19 antibodies, either because they are sick or were sick.
The crematoriums got emergency authorization to surpass emissions standards and burn bodies at all hours because the corpses were piling up and couldn't fit in cold storage. And a study came out that nearly 20 percent of all COVID positive people released from hospitals ended up dying of complications. I get that running less than fully secure programs on our GPUs comes with inherent risks. But every day without functional cures comes with much greater risks. And accelerating this project with GPU folding could make a difference to many people's lives. Anyhow, BOINC applications can be updated after they're released to fix security flaws. The chance of a malcontent finding and exploiting a security vulnerability between now and whenever the project would otherwise be ready is pretty small. There's just not enough users to target and the timeline before the application updates to a secure version is fairly short. So why not let us opt-in to run our GPUs on the project now? I'm sure the option to participate early could be placed below a waiver form. |
||
|
|
kamaradski
Cruncher Joined: Dec 16, 2010 Post Count: 12 Status: Offline Project Badges: 
|
no thanks, better wait for a little and not kill my hardware or open up security problems on my network.
|
||
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3315 Status: Offline Project Badges: 
|
The GPU application was already run at the Oak Ridge National Laboratory, at least the CUDA version that was developed with Nvidia. That's where IBM's Summit Supercomputer is. I doubt security is an issue.
----------------------------------------But I don't see this happening.  - AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W - AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W - AMD Ryzen 7 7730U 8C/16T 3.0 GHz [Edit 1 times, last edit by Falconet at Jan 20, 2021 4:58:01 PM] |
||
|
|
Jim1348
Veteran Cruncher USA Joined: Jul 13, 2009 Post Count: 1066 Status: Offline Project Badges: 
|
There is something they are not telling us. Probably they can't analyze the data any faster, and are happy with CPUs. Or else the GPUs introduce some uncertaintly in the data, and the results are not entirely comparable to the CPUs.
I don't expect an answer soon, so I will just crunch with CPUs. The Ryzen 3000's are very good, and the Ryzen 5000 will be better. |
||
|
|
Mike.Gibson
Ace Cruncher England Joined: Aug 23, 2007 Post Count: 12594 Status: Offline Project Badges: 
|
If things are not ready then I am happy to use my GPU to crunch Einstein GRP4. That doesn't need a CPU thread to be vacated.
Mike |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Nothing would make me drop this entire site faster than skipping security measures.
|
||
|
|
Speedy51
Veteran Cruncher New Zealand Joined: Nov 4, 2005 Post Count: 1326 Status: Offline Project Badges:
|
I agree. I would also drop the project if security checks were skipped.
----------------------------------------Jim1348 wrote: There is something they are not telling us. Probably they can't analyze the data any faster Personally I find this hard to believe. I hope what you are saying is not true, time will tell  |
||
|
|
Jim1348
Veteran Cruncher USA Joined: Jul 13, 2009 Post Count: 1066 Status: Offline Project Badges:
|
Personally I find this hard to believe. I hope what you are saying is not true, time will tell Even if they completed the computer analysis instantly, they have to synthsize the most promising molecules and then do lab tests to see if they really have anti-viral properties. Then, tests on laboratory animals to see if they work there and don't have toxic affects. Finally, one might make it to human trials, which could last for months. But do they have a commercial partner to produce and sell it? A lot of studies just result in papers being published in academic journals. I would like to see some idea of the overall program myself. |
||
|
|
Dayle Diamond
Senior Cruncher Joined: Jan 31, 2013 Post Count: 452 Status: Offline Project Badges:
|
I agree. I would also drop the project if security checks were skipped. Falconet has a point. Is your computer really that much more high security than Oak Ridge National Laboratory and IBM? What is it about giving users the option to begin GPUs early that would make you want to withdraw from secure CPU tasks? I think you're falling for a Loss Aversion fallacy. Nothing happening to you could be worse than yet another month without a cure for COVID. https://en.wikipedia.org/wiki/Loss_aversion Even if it takes more time that we'd like to finish the rest of the project, there's somebody's neck on the chopping block here. A death in a year is just as tragic as the deaths happening now - better to accelerate the research as best we can. Heck, they could just open the public version up early, say its ready, and than patch it later as an update. I'm sure, given infinite time and attention, that prior weaknesses could be found in prior projects. Nothing was ever risk free - Medicine has to move forward. |
||
|
|
Danathar
Cruncher Joined: Jul 28, 2011 Post Count: 7 Status: Offline Project Badges: 
|
If you guys want to use your GPU for Covid...while waiting for WCG to get the GPU system up and running just run folding@home but pause the CPU work units (only run GPU). You can run that at the same time you run BOINC (It's what I do).
|
||
|
|
|