The 12th Annual Folding@Home Holiday Season Race

[StefanR5R]

I just looked at the FAH team stats pages for our two teams and they are credited with 600 cpu's and us with 7............ not sure how to interpret that.

These numbers must be garbage. Tom's Hardware and Barnacules Nerdgasm each have exactly 600 "active CPUs within 50 days" too.

 

[crashtech]

Losing gracefully means doing our best. And most certainly we will do that.
I am considering to add my CPUs to the fray - it means to move them from WCG to F@H, but OTOH: this is the 12th Annual Folding@Home Holiday Season Race.

Indeed, despite my initial shock at the magnitude of their gain, rest assured I'll be giving F@H everything I have for the duration.

 

[VirtualLarry]

Well, I'm trying to figure out how to deal with my primary Ryzen R5 1600 rig rebooting on it's own, whether I need to RMA the CPU with AMD, or what. (Probably will just replace it when, when Ryzen 2 series comes out in Feb.) Until then, I've got this race to worry about, and my RX 570 is in this box too. I am unsure whether the issue is actually CPU, or GPU-related, as it seems mostly stable running the WCG on CPU on 10 threads, but when I add in the F@H on the RX 570, GPU temp hits 89C (NOT in Crossfire), even with a dual-fan card. I've got one 120mm Intake, and one 120mm exhaust blowhole, and then the rear exhaust is a 120mm through a radiator. I guess I need better case cooling. I've got a better case here waiting to be upgraded, I may do that this week. Until further notice, though, that hardware is down.

 

[StefanR5R]

Don't you have the side panel open? (Plus, could turn the case on its side, so that less heat from the GPU strikes the CPU/ VRM/ RAM area.)

 

[StefanR5R]

I am receiving CPU work again.

Also, some explanations:

Re: No credit from 155.247.166.220 and 155.247.166.219

by vvoelz » Tue Dec 12, 2017 22:28 UTC

A follow-up: We've been working hard to fix this firewall issue -- sorry it took so long! From everyone's comments, it looks like the stat server is now able to communicate with our Folding@home work servers at Temple University, and participants are indeed receiving credit. The delay was a result of an ongoing security review by Temple's computer services; only after successfully passing internal security tests could we petition to install the firewall settings to allow the stats server to connect (whew!).

Thanks again to all those who reported and helped diagnose this issue. We sincerely appreciate all your efforts, and especially your patience!

Vincent Voelz
Temple University

 

[Pokey]

I am receiving CPU work again.

Also, some explanations:

Wonder what Temple would take to remove the fix and put things back the way they were? ..........

 

[crashtech]

Wonder what Temple would take to remove the fix and put things back the way they were? ..........

That's like closing the barn door after all the ponies have escaped!

 

[tiwake]

That's like closing the barn door after all the ponies have escaped!

You can't escape the pony. Challenging somebody better than us brings out the best in us.

 

[VirtualLarry]

Hmm, I might have my GPU issues partially sorted. It appears that my RX 570 is simply overheating. I guess that they draw more power, and are a bit more aggressive with the factory voltage, so they run hotter. I've got my RX 570 on Power Limit -30%, and Fan Speed on Manual, set to 2900RPM (up from 2200). Seems to have stabilized temps at 75C.

I wish I knew what the two BIOS switches did. (I think I tossed the manual when I got it.) Maybe one of them is a "cool and quiet" (lower voltage) setting, and one is an "all out overclocked for gaming" setting, which is probably what I'm on.

 

[hiigaran]

Just do what I do. Set all fans permanently to 100% and wear earplugs for the rest of your life!

 

[StefanR5R]

No bunkers in F@H (coming in 8 days)

Exceptions prove the rule.

 

[Ken g6]

Exceptions prove the rule.

You realize that the later you return results the less credit you get, right?

 

[StefanR5R]

I referred to the "bunkering" that Temple Uni performed for us.

 

[crashtech]

I'm getting enough CPU work now.

 

[StefanR5R]

~46.3M points last 24 hours! (and climbing!)

Last year's peak(?) when racing Tom's Hardware:

The peak was 53.3 M according to petrusbroder's final stats of the 1-week race:

At this time, EOC has six 3-hour updates which appear no longer affected by the Temple stats hickup.
Extrapolating from these last 18 h to 24 h:

TeAm = 51.7 M PPD, Brony = 48.9 M PPD​

Extrapolating from only the last 9 h to 24 h:

TeAm = 52.9 M PPD, Brony = 50.4 M PPD
​

 

[Pokey]

I already resigned myself to wait a few days for the stats averages to settle down. We will know more when this big blip fades from view.

We do know that Brony has twice as many folders as we do, so that.

 

[petrusbroder]

Here are the stats for the 13th day of the 12th Folding@Home Holiday Season Race
Stats are as of December 13, 2017, 20:30 UTC (approx).
The inter-team race stats are from the Folding@Home stats page.
The intra-team race stats are from EOC.com.

The inter-team race:

The intra-team race:

The intra-team race daily production:

The intra-team race individual daily production:

Well folded, folks! We need new folders ...

 

[Orange Kid]

Thanks for all the great stats you keep churning out.

 

[TennesseeTony]

Indeed, thank you petrusbroder!

Lost some points today. For the second time, I've caught a F@H client stuck downloading. I should have remembered to check before leaving for work this morning, but it looks as though one 1080 was down for 400,000k points, or 12 hours.

 

[petrusbroder]

Here are the stats for the 14th day of the 12th Folding@Home Holiday Season Race
Stats are as of December 14, 2017, 20:50 UTC (approx).
The inter-team race stats are from the Folding@Home stats page.
The intra-team race stats are from EOC.com.

The inter-team race:

The intra-team race:

The intra-team race daily production:

The intra-team race individual daily production:

I am considering some slight restructering of the teams to make the intra-team race more ... enjoyable?, interesting? Well, whatever ...
Please let me know what you think ...

 

[StefanR5R]

So last year we peaked at 53.3 M per day on the 6th day of the race against Tom's.

This year it took us 13 (or maybe 12) days to reach the same output. Though on the bright side, this is possibly still not our peak this year.

 

[ao_ika_red]

So last year we peaked at 53.3 M per day on the 6th day of the race against Tom's.

This year it took us 13 (or maybe 12) days to reach the same output. Though on the bright side, this is possibly still not our peak this year.

To be fair, there was an article about F@H competition in both Tom's and Anand's front page that helped bringing additional heavyweights to the party. As for this year, unless you actively check DC subforum, you don't even realise the race existence.

 

[StefanR5R]

By the way:

Brony@Home are now at 670 M points after 13 days. This includes some points which were earned before the race, but credited during the race, due to Pande group's site-to-site traffic being broken until December 12.

I estimate that the real average production of Brony@Home within Dec 2...Dec 13 14 was 49...50 M PPD. But let's assume conservatively that they did 52 M PPD at average over these 12 13 days. Then they would be at 620 or 630 M 676 M by now.

(We too probably got a few points credited during the big stats jump which were really earned before the race. But it really must have been just a few. I for one assure you that I did not run F@H on CPUs before the race, and I was the one in our team with by far the biggest points update at the December 12 stats jump.)

I don't want to create bad vibes, just putting this out here as food for thought. I don't know myself yet what to make of this. Generally, project servers acting up are among the common difficulties of DC competitions. Messed-up project credit databases are arguably a less common occurrence though.

Edit:
Sorry, I was off by one day. 13 days * 49..50 M/day = 640...650 M. So, 30...20 M from before the race by this less conservative estimation; 0 M from before the race by the conservative upper-bound estimation with 52 M PPD average.

 

[IEC]

To be fair, there was an article about F@H competition in both Tom's and Anand's front page that helped bringing additional heavyweights to the party. As for this year, unless you actively check DC subforum, you don't even realise the race existence.

Yes, it would be nice to have some publicity to recruit some TeAm members, even if only for a month or two.

 

[petrusbroder]

Here are the stats for the 15th day of the 12th Folding@Home Holiday Season Race
Stats are as of December 15, 2017, 20:15 UTC (approx).
The inter-team race stats are from the Folding@Home stats page.
The intra-team race stats are from EOC.com.

The inter-team race:

The intra-team race:

The intra-team race daily production:

The intra-team race individual daily production:

I'll probably restructure the teams tomorrow ... we will see.