Cheater or stats fail?

[jcool]

Just stumpled upon this machine...

http://boincstats.com/stats/host_gra...wcg&id=1275071

Hes doing twice the average PPD of my SR-2 machine (Dual Hex @ 4Ghz) with a lowly old Dual Gainestown (according to the stats).
Obviously that kind of PPD is impossible with the stated hardware. Is this some kind of cheat, tweak, or a simple stats fail?

[D_A]

I have no idea how he'd be cheating, unless it's a really badly hacked client but even then the quorum should be catching that.

[retro77]

Did you see that there was 16 cores? Math add up?

[jcool]

Did you see that there was 16 cores? Math add up?

No, it doesn't add up, not even close. It's 8 cores / 16 threads, my Dual Hexas show as 12 (24) there. Unless it's some quad socket rig showing up as a Dual E5540 I'd have to say it's some kind of hack.

Here's my SR-2 for comparison. http://boincstats.com/stats/host_gra...wcg&id=1142087

[stoneageman]

Here's the owner/operator and he even has more

Detailed host stats on Free-dc

[sierra_bound]

That machine may be running more than 16 cores. Possibly 32 on quad socket board? Sometimes BOINC doesn't correctly read the number of cores on a cruncher.

[trn]

He post on the wcg forums; I think i've seen him seen him say vaguely he has dual quads :dunno: at 24,500 boinc RAC that would be more RAC than 4 x 5540's could do anyways, it would take about 6 x 5540's for that output.

[jcool]

That machine may be running more than 16 cores. Possibly 32 on quad socket board? Sometimes BOINC doesn't correctly read the number of cores on a cruncher.

Ok, but then why on earth does it recognize the CPUs as E5540 if they are something else entirely? I don't get it, but I believe something is off here...
I've never seen BOINC report completely wrong CPU info (core info yes).