Folding Forum

The article got published on Saturday 6/11! https://www.telegraph.co.uk/technology/ ... deadliest/ (unfortunately behind a paywall, but I was lucky enough to buy the actual newspaper!).

I just tried the link and it wasn't protected by a pay wall. So that may have been changed.

Due to the ice storm here in Oregon, my power was off for 6 days thus no folding at all. So I was doing my part in not overloading the servers.

My daily points have been fluctuating a lot in recent weeks, with big dips around Christmas and New Years. I guessed that there was a shortage of work units due to people being off for the holidays.

Does M/S not realize that there are some people who actually need to get work done on their computers, instead of playing the literal game of "whack-a-code" and re-installing the GPU manufacturers WHQL tested drivers that come with the library files that may be needed by other software???...

It could be that a Windows 10 update decided to modify or replace the Nvidia drivers that you had just installed.

I have an Epson V700 scanner that works all the time from Windows 7. But from Windows 10 several times a year a Windows 10 update breaks it and I have to reinstall the Epson drivers.

I don't know for sure that the original error was from failing hardware but it sure looks that way. After years of service, and more recently working OK with CUDA work units, it started giving intermittent errors. And then it totally failed. No fans or anything. At 5 years old, it isn't ancient but ...

An update to say that my old GeForce GTX 980 Ti failed last week. Totally dead.

I was hoping it would last long enough for me to find a new RTX 3070 in stock someplace. As that didn't happen, I replaced it with another RTX 2070 which has been using CUDA and has run for a week now with no errors.

Nope, no overclocking of either the CPU or the GPU.

PantherX wrote:Once FahCore_22 version 0.0.14 (or newer) is released, I would suggest that you try CUDA to see what happens.

Thanks. I'll watch for that.

I did update the NVidia driver to 456.71 but that didn't make a difference. With OpenCL the gpu is running at about 80 degrees C, and 83 with CUDA. The max temp. according to NVidia is 92. So it doesn't look like heat is an issue. My other computer has a GEForce RTX 2070 and has never had a FAH erro...

Just had this error a couple of times. The Windows error info: Problem signature: Problem Event Name: APPCRASH Application Name: FahCore_22.exe Application Version: 0.0.0.0 Application Timestamp: 5f656e96 Fault Module Name: OpenMMCUDA.dll Fault Module Version: 0.0.0.0 Fault Module Timestamp: 5f6518b...

Performance Ratio is based on Price/performance (est); Real life numbers will differ slightly. 2080 Ti: 4352 Cuda Cores, 352 bit GDDR6, $1200 ~3.6-4.1M PPD ; RATIO: 3.2 3070: 5888 Cuda Cores, 256 bit GDDR6, $499 ~3.5-4.0M PPD (EST) ; RATIO: 7.5 3080: 8704 Cuda Cores, 320 bit GDDR6X, $699 ~6.4M-7.2M...

Postby ajm » Sun Jun 28, 2020 10:28 am

It should be automatic. You can check in your log whether it's been done, a bit after that line

Thanks. My log shows Core22 so it did indeed update with no intervention from me.

Interesting that the Version: is a mix of 0.0.10 and 0.0.11.

JohnChodera wrote:We're excited to roll out core22 0.0.10 to full FAH after extensive testing of updates since 0.0.5.

~ John Chodera and the core22 team

Is there something users need to do to get this or is it a behind the scenes automatic update?

I only included one of the crashes, thinking that there was no point in showing the same thing several times.

The computer's up time is 9 days. So it wasn't a shutdown. It is back to normal operation so I guess it can be ignored now.