World Community Grid now has their Open Pandemics subproject running, With COVID-19 as the first pandemic it is working on.

Nice!

The increase in power is likely that Fahcore_a7 is utilizing AVX set instructions.

The increase in power is likely that Fahcore_a7 is utilizing AVX set instructions.

I concur. This is most likely.

You should implement an AVX offset in your BIOS if your motherboard has that feature. It will reduce the clocks when it detects AVX workloads.

---

The MMR vaccine offer some protection against COVID-19.

Analysis of Measles-Mumps-Rubella (MMR) Titers of Recovered COVID-19 Patients

https://mbio.asm.org/content/11/6/e02628-20

Do you have enough information on mumps antibodies that you could create workunits to dock them to parts of the virus?

Respectfully, you are misunderstanding the finding:

"Our results demonstrate that there is a significant inverse correlation between mumps titers from MMR II and COVID-19 severity."

That does not mean that the MMR vaccine can prevent COVID-19. It means that a statistical association was observed. Correlation is not causation.

Persons without medical contraindications are advised to be immunized with MMR (measles-mumps-rubella) vaccine because measles is a serious disease potentially fatal to children, it is perhaps the most transmissible virus known to man, and there was multi-year global outbreak prior to COVID-19. But not because the MMR vaccine has been proven to prevent COVID-19.

Hmmm... I hate to be a party pooper, but:
My personal opinion is that crunching for COVID-19 is not very useful.
The whole world is looking for drugs to fight that disease, and they are doing thousands of in vitro tests a day, which works a lot quicker than computer simulations.

So my humble opinion is this: (I may be wrong)
It is better to let our computers crunch data for diseases which either do not receive enough attention (child cancer, Tuberculosis, ...) or diseases of which we know that a lot of work is needed for years/decades to come like cancer, so long term projects.

So if GPUGRID stays without WUs for some more time and I decide to temporarily switch to Folding@Home - at least for my GPUs - I would rather let them do work for Cancer for example, rather than COVID-19.
We have vaccines for COVID-19 now, but a lot of work remains to be done before we can beat cancer definitively, or child cancer, or tuberculosis etc.

My CPUs do work for the Mapping Cancer Markers project, the tuberculosis project and the Child Cancer project at World Community Grid.
Alas these do not use GPUs.
So if you want to put your GPUs to work, I would choose diseases at Folding@Home which will need long term work (years/decades) before we can beat them.
Something like Covid-19 which is already being beaten by the first generation of vaccines... I just think it is less interesting for our species to put our GPUs to work on something which is about to get solved by the first vaccine generation, at least temporarily.

This is just my opinion.
But I respect everybody's choice of course. Use your GPUs the way you like. :-)
My choice is to put my GPUs to work on longer-term problems.

Greetings, and have a nice day;
Carl

Hmmm... I hate to be a party pooper, but:
My personal opinion is that crunching for COVID-19 is not very useful.
The whole world is looking for drugs to fight that disease, and they are doing thousands of in vitro tests a day, which works a lot quicker than computer simulations.

I think there is considerable merit to that, and have been wondering about it myself (with over 100 Ryzen 3000 virtual cores on WCG/OPN at the moment). The limiting factor at the moment is testing, with animal testing and then human testing being even more limited and expensive than in vitro testing.

However, there is the question of mutations. We need something that targets a broad range of viruses, and it might be possible that the computer simulations can do that. In particular, I think that Folding can find and target areas of the virus that the ordinary testing techniques will miss. I have 10M PPD riding on it in any case. I hope something works.

Hmmm... I hate to be a party pooper, but:
My personal opinion is that crunching for COVID-19 is not very useful.
The whole world is looking for drugs to fight that disease, and they are doing thousands of in vitro tests a day, which works a lot quicker than computer simulations.

I think there is considerable merit to that, and have been wondering about it myself (with over 100 Ryzen 3000 virtual cores on WCG/OPN at the moment). The limiting factor at the moment is testing, with animal testing and then human testing being even more limited and expensive than in vitro testing.

However, there is the question of mutations. We need something that targets a broad range of viruses, and it might be possible that the computer simulations can do that. In particular, I think that Folding can find and target areas of the virus that the ordinary testing techniques will miss. I have 10M PPD riding on it in any case. I hope something works.

Jim, I partially agree with you.
And I did not mean to say that computer simulations are useless against Corona.

But let us not forget that each year, the flu still also kills many people worldwide.
We just have the luxury that we have learned to make quite effective vaccines against the flu each year.
A luxury that we do not have yet against Corona, but the vaccines will get better very quickly: Like the flu, Corona is a rather "simple" virus, nothing compared to the complexity of the HIV virus:
Don't forget that - even after 30 years of research - we have still not found ANY vaccine against HIV. For Corona I took us less than a year to create the first vaccines.

So I do believe that these distributed computing projects about COVID-19 will be quite short-lived:
Compared to the extremely intensive in vitro tests that they are doing now, any computer simulation project is slow:
Computer simulations are a better choice to solve long-term problems like cancer or more advanced viruses like HIV.
The in vitro tests will "solve" Corona rather quickly, by making better and better vaccines just like for the flu.


So I prefer to use my PCs against cancer and HIV for example.
It is just my choice.

I do not have nearly as much computing power as you have. :-)
I recently purchased an AMD Ryzen 9 3900 (non -x) for the Mapping Cancer Markers project at WCG.
That is a 12-core, 24 thread CPU with a TDP of only 65 Watts, so a quite efficient CPU ( OEM only, so a little difficult to find though. )
For the rest, I only have 1 Core i5 8400 6-core/6-thread, and older pcs with 2 AMD A8-9600 quad cores and then an even older Core-2 Duo and a Celeron dual core too.
So compared to your CPU computing power, that is almost nothing. :-)

I recently purchased an AMD Ryzen 9 3900 (non -x) for the Mapping Cancer Markers project at WCG.
That is a 12-core, 24 thread CPU with a TDP of only 65 Watts, so a quite efficient CPU ( OEM only, so a little difficult to find though. )

I didn't know there was a non -x version. That looks like good efficiency to me.
There is supposed to be a 5900 non -x version too, but also OEM only. I would love to find one later in the year. I am full for now.

Hmmm... I hate to be a party pooper, but:
My personal opinion is that crunching for COVID-19 is not very useful.
The whole world is looking for drugs to fight that disease, and they are doing thousands of in vitro tests a day, which works a lot quicker than computer simulations.

[snip]

You're ignoring the rather high expense of getting new drug candidates made if they have never been produced before.

The computer simulations are much faster and less expensive than getting those new drug candidates made, and are useful in determining which new drug candidates are the most likely to be worth the expense and time required to have them made so that in vitro tests for them become possible.

Also, while there are many in vitro studies and successes, they often fail at the next stage of animal testing. That is because they kill the virus in the lab, but also kill the animal because they are toxic.

Can the computer studies improve on that success rate? I don't know. We should be finding out before long I would think.

The lab testing on WCG/OPN should begin shortly.
https://www.worldcommunitygrid.org/about_us/viewNewsArticle.do?articleId=674

I recently purchased an AMD Ryzen 9 3900 (non -x) for the Mapping Cancer Markers project at WCG.
That is a 12-core, 24 thread CPU with a TDP of only 65 Watts, so a quite efficient CPU ( OEM only, so a little difficult to find though. )

I didn't know there was a non -x version. That looks like good efficiency to me.
There is supposed to be a 5900 non -x version too, but also OEM only. I would love to find one later in the year. I am full for now.

fyi:
I found my Ryzen 9 3900 non-x at a site called AWD-IT in the UK.
As it is an OEM chip, they are not allowed to sell it like that,
so they sell it in combination with at least a motherboard.
Apparently for AMD that is "OEM enough". :-)
And you can choose the motherboard.

I was lucky enough to buy mine just before Brexit (I live in Brussels).
But so if you live outside the UK now, I guess customs will be applied.

And if they run out of chips, someone else found one
at a Russian shop called X-Com in Moscow. I saw that if you let your browser translate their site to English , it is quite readable.

We have vaccines for COVID-19 now

Those vaccines
1. don't cure those who are actually sick with COVID-19 (ok, some of them does)
2. are hard to produce, store, transport, inject
3. are expensive
So these vaccines are useless for those who live in poverty (for the majority of the world's population).

The aim of the COVID-19 moonshot project (FAH is a main contributor of that with its computing power) is to design a drug which
1. will cure those who are actually sick, or will get sick with COVID-19 (and perhaps many similar corona viruses in the future)
2. are easy to produce, store, transport, take
3. are inexpensive

Don't forget that this corona virus is actually the third (obviously the most successful) one in the last 20 years which threatens our civilization (1st one: SARS; 2nd one: MERS). So brace yourselves, these corona viruses won't let humanity alone in the future. Therefore it is vital that we could develop much better drugs against them much faster than this time if we want to get our everyday lives back, and don't let the world economy collapse (we couldn't support research against cancer and AIDS in this case).

Though Folding@home is now focusing on SARS-COVID-19, it's not only about COVID-19, it has ongoing research about myosins, GPCR, cancer (CPU only), Halorubrum lacusprofundi, Parkinson's, G-proteins.

My main issue with FAH (besides that it's not within BOINC) was that their client (core in their terminology) was openCL (not CUDA). But now it is CUDA, so that's great news for NVidia users like me. My main concern is their set-and-forget approach, so the FAH user interface and cache strategy and web infrastructure is not made to match my tastes. I cannot control my farm through "venues", I can't set cache size, there's no webpage for my hosts to track their workunits etc.

My main concern is their set-and-forget approach, so the FAH user interface and cache strategy and web infrastructure is not made to match my tastes. I cannot control my farm through "venues", I can't set cache size, there's no webpage for my hosts to track their workunits etc.

If you use HFM.net, they have a "Tools/Work Unit History Viewer". I don't use it myself and don't know how it compares to BOINC, but it might help. (I have HFM.net installed on a Win10 machine to monitor all my Linux machines over the LAN).

They are stuck with their cache strategy by the type of work they do. They have to start up the next one right away, so can't keep much of a cache. In fact, that is the main reason they don't use BOINC I believe. I have learned to live with their limitations.

We have vaccines for COVID-19 now

Those vaccines
1. don't cure those who are actually sick with COVID-19 (ok, some of them does)
2. are hard to produce, store, transport, inject
3. are expensive
So these vaccines are useless for those who live in poverty (for the majority of the world's population).

The aim of the COVID-19 moonshot project (FAH is a main contributor of that with its computing power) is to design a drug which
1. will cure those who are actually sick, or will get sick with COVID-19 (and perhaps many similar corona viruses in the future)
2. are easy to produce, store, transport, take
3. are inexpensive

Don't forget that this corona virus is actually the third (obviously the most successful) one in the last 20 years which threatens our civilization.

You are correct.
Virologists have told us that - had we invested say a billion Dollars in antivirals in the past - we would not be in this situation now.
But we haven't done that because politicians don't want to think about the long term future: They get elected for just 4/5 years, not for 20 years.

But vaccines are not by definition more expensive to make:
Antivirals also need to get an update every year or every x years.
The RNA vaccines like Pfizer's are indeed expensive, hard to store and transport etcetera.
So for Africa for example this vaccine is not a realistic solution.
But Johnson&Johnson has made a vaccine with the developing countries in mind:
It is a lot easier and cheaper to make, and it can be stored for weeks/months in a simple refrigerator. A bit like the flu vaccines actually.
In the EU it will get approval in the coming weeks.
It is less effective in that it has a +-65% chance to keep you from getting sick, but it is equally successful in making sure you don't need to go to a hospital if you do get sick, and that is more important than keeping you from getting sick:
Keep Corona patients out of the hospitals, and keep them from dying.
And this Johnson&Johnson vaccine only needs 1 injection, so that makes it 50% cheaper to transport, store and administer.

I hope we will find antivirals, but will we succeed?
There are many viruses for which we don't have (good) antivirals, because they do also mutate.
HIV is an exception, but as you know they need to keep making new anti-HIV drugs because the virus mutates.
And that makes antivirals also expensive in the long run I guess.

But vaccines might be more expensive because you need to give them to everybody, not just to those who get sick. That is true.

But vaccines might be more expensive because you need to give them to everybody, not just to those who get sick. That is true.

The expense should not even be a factor. Whether it is $15 per dose or $150, it is trivial compared to the economic loss. The only factors are speed and efficiency.

By the way, I wonder about the new drug they gave to Mr. Donald Trump when he got Corona.
It was called "Regeneron" I think.

I wonder how it works:
Is it a simple antiviral molecule, or does it work more like immune therapy, or is it a new generation drug with a patient specific compound...

But vaccines might be more expensive because you need to give them to everybody, not just to those who get sick. That is true.

The expense should not even be a factor. Whether it is $15 per dose or $150, it is trivial compared to the economic loss. The only factors are speed and efficiency.

I totally agree:
Once a pandemic has become so big, money should not be the first thing to think about.
That is why I said from the beginning that thanks to the first generation of vaccines, this problem will soon be solved, at least temporarily.
Finding antivirals will take a lot more time, if we find them.