Folding Forum

Team,

From yours stats reports, I noticed that "Active CPUs" is approximately half of what it was a year ago.

http://fah-web.stanford.edu/cgi-bin/mai ... pe=osstats CPUs
http://fah-web.stanford.edu/cgi-bin/mai ... e=osstats2 Cores

1) I wonder how much of this drop was due to Sony discontinuing support for folding-at-home on PS3
2) I wonder how much of this drop was due to AMD/ATI dropping support for OpenCL on Windows-XP
3) I wonder how much of this drop is due to people moving from PCs to pads

Any thoughts on this?

p.s. are the Macintosh stats correct? 437 active CPUs seem somewhat low.

Several persons were tracking the stats in a Google doc, that is discussed in the topic Overall F@H Stats Graph. The stats put out by PG changed last Fall - blog post, that may be part of the change in active CPU's reported. A separate Google doc was created to track the new stats.

From those old stats, there were 16,000+ PS3 CPU's listed shortly before the Sony announcement. That would partly explain the drop. The drop of OpenCL support by AMD might be part of the drop, but I think a bigger loss was all of the pre-5000 series cards that stopped getting work last year. No idea how much is due to people switching to tablets.

One thing to remember when looking at the stats a year back, they were inflated a bit by the HPCS burn-in. Between clients run under the HPCS team and those by persons using the free test accounts, that did briefly increase the total CPU count for a while.

neilrieck wrote:p.s. are the Macintosh stats correct? 437 active CPUs seem somewhat low.

The 437 number is for PPC CPU's. The supply of WU's for those was turned off in early December, so that is the count of CPU's active in the last 50 days. It will continue to drop, looking at the log for the WS involved the number of WU's turned in was down to one or two a day last week. As for the OS X Intel count, that is definitely inaccurate. At 8 active CPU's and 26 cores it is definitely off as I have 3 and 18 respectively active that are currently contributing.

If this statistic is correct, visible noticeable decline in the number of active members. From my observations about previous years (November/December/January), the trend of the number of active members is or (+) or a constant for these months.

According to: http://kakaostats.com/

Members Active:

-59 600;(17 000 -Team Default)-15.05.2012
-57 600;(16 400 -Team Default)-15.06.2012
-54 200;(15 130 -Team Default)-15.07.2012
-50 100;(13 300 -Team Default)-15.08.2012
-48 200;(12 700 -Team Default)-15.09.2012
-48 800;(12 870 -Team Default)-15.10.2012
-49 000;(12 880 -Team Default)-15.11.2012
-43 100;( 9 880 -Team Default)-15.12.2012
-37 850;( 7 700 -Team Default)-15.01.2013