Recent Boost in F@H Output?

[Kurtis200200]

I gathered some numbers from DC-stats, and it appears that, since May 11th, F@H output has increased ~30%. See PDF for spreadsheet.

Could this apparent boost be due to the launch of CureCoin around May 11th, or are the statistics too fuzzy to make that kind of conclusion?

edit: fixed broken pdf link

[bruce]

The cause-and-effect conclusions are always a bit fuzzy, but compare the stats for the CureCoin team with the totals and draw your own conclusions.

[Jesse_V]

The cause-and-effect conclusions are always a bit fuzzy, but compare the stats for the CureCoin team with the totals and draw your own conclusions.

Curecoin PPD: http://folding.extremeoverclocking.com/ ... =&t=224497

Could be coincidence, but I suspect that there's a correlation/causation.

[PantherX]

Before the CureCoin was launched, this was the data from the Client statistics by OS page (04 May IIRC):

Code: Select all

OS Type 	Native TFLOPS*	x86 TFLOPS* 	Active CPUs	Active Cores	Total CPUs
Windows 	742	742	170529	296855	5420657
Mac OS X 	20	20	7421	49142	160929
Linux 	35 	35 	11433	19938	764867
ATI GPU 	842 	1777 	5933	5933	369038
NVIDIA GPU 	2410 	5085 	12753	12753	313536
NVIDIA Fermi GPU 	17209 	36311 	50763	147926	390862
Total 	21258	43970	258832	532547	7419889

This is what it currently reports:

Code: Select all

OS Type 	Native TFLOPS*	x86 TFLOPS* 	Active CPUs	Active Cores	Total CPUs
Windows 	603	603	138661	230997	5440504
Mac OS X 	22	22	8036	61685	162035
Linux 	36 	36 	11694	21481	766733
ATI GPU 	1658 	3498 	11676	11676	375783
NVIDIA GPU 	2299 	4851 	12166	12166	315431
NVIDIA Fermi GPU 	17091 	36062 	50417	158629	399068
Total 	21709	45072	232650	496634	7459554

Sorry that the data isn't formatted into a pretty table since time is against me.

[7im]

The cause-and-effect conclusions are always a bit fuzzy, but compare the stats for the CureCoin team with the totals and draw your own conclusions.

Curecoin PPD: http://folding.extremeoverclocking.com/ ... =&t=224497

Could be coincidence, but I suspect that there's a correlation/causation.

No way to tell how much is purely new CureCoin donors, and how many donors were pulled from other teams.

[SodaAnt]

The cause-and-effect conclusions are always a bit fuzzy, but compare the stats for the CureCoin team with the totals and draw your own conclusions.

Curecoin PPD: http://folding.extremeoverclocking.com/ ... =&t=224497

Could be coincidence, but I suspect that there's a correlation/causation.

No way to tell how much is purely new CureCoin donors, and how many donors were pulled from other teams.

It would require a closer look at the data, but I'd suspect most people that moved to curecoin are using a previous name used for folding, versus new people are using a name that hasn't appeared or hasn't folded much before.

[7im]

Unless you looked at each donor name on the CureCoin team to see if they had 2 fah teams listed, that data is not exposed at a level that anyone but PG could "take a closer look" at with any level of ease.

So while there IS a way that is very manually intensive, nobody is likely to do that. End result is the same.

Might be interesting to look at a few of the top donors on that team though.

[Jesse_V]

Many of the top donors hang out in #curecoin, and I've heard a good number of reports of people switching over to F@h from mining to folding, or firing up high-end machines or massive cloud instances specifically for Curecoin, bringing them over a million PPD in some cases. 7im is right though, without a deeper inspection it's all speculation, but based on what I've read there's a chunk of that performance that actually is new to F@h.

[jrweiss]

What difference does it make? Obviously there has been an upsurge in Folding.

It doesn't make ANY difference at all to the project what team any particular Folder belongs to. The data posted by PantherX shows a significant boost in the number of cores Folding, so there has likely been a combination of people switching teams and people devoting new machines to Folding.

In either case, it is good for the project!

[7im]

What PX data are you reading? The active clients and the active cores went down. Flops went up, only a little.

[Kurtis200200]

Before the CureCoin was launched, this was the data from the Client statistics by OS page (04 May IIRC):
…
Sorry that the data isn't formatted into a pretty table since time is against me.

Unfortunately, snapshot core data isn't as robust as multiple core polling points over time, but it's a meaningful addition and I didn't think of checking the OS Stats page , thanks PantherX

[bollix47]

Does the information here help?

Information prior to 11/28/2012 can be found here.

[k1wi]

If PPD has increased while TFLOPS have stayed relatively constant, then one would assume that many less powerful clients leaving have been offset by the addition of fewer more powerful clients...