Xeon Phi

TObject · 2013-05-23, 02:58

60 LL tests running side-by-side, anyone?

http://www.tomshardware.com/news/Int...CPU,22700.html

firejuggler · 2013-05-23, 04:21

I don't know if P95 can handle more than 32 core. I recall a thread that stated that even with 2 instance of P95 running, it wouldn't use the 48 core of the machine.

ixfd64 · 2013-05-23, 04:57

I may be wrong, but I believe the 64-bit version of Prime95 could handle 64 threads.

LaurV · 2013-05-23, 05:14

that would not be a problem, virtual box (or equivalent), ten copies of the guest OS running with 32 cores allocated to each, P95 running there... just give me the wheelbarrow with the 320 cores... For the phi, the real problem will be its common caches, in fact.. because no matter how fast your dog can eat, you still must be able to feed him so fast...

So, in spite of having 7-8 times more cores then a "top" 8-cores Xeon E5-2670, it can only get 3-times more DP performance. That would be about 5- to 6-times faster compared with a i7-2600k (as everybody know what a 2600k is, but not everybody knows what a e5-2xxx is

).

edit: and this is for clock per clock, but don't forget the 2600k has a faster clock, and runs much overclocked too, in water-cooled rigs. The pessimist in me says we won't see a more than 5 times performance increase, and that is already reachable with cudaLucas.

Next big hit would be a FFT library for Radeons...

ewmayer · 2013-05-23, 19:22

Quote:

Originally Posted by firejuggler

I don't know if P95 can handle more than 32 core. I recall a thread that stated that even with 2 instance of P95 running, it wouldn't use the 48 core of the machine.

TObject has it right as far as maximizing overall throughput - 1 LL test per core, or perhaps one LL test per 2-4 physically 'clustered' cores. Would be interesting to experiment and see what the optimal test/core matching strategy is here.

TheJudger · 2013-05-23, 20:01

Quote:

Originally Posted by TObject

60 LL tests running side-by-side, anyone?

http://www.tomshardware.com/news/Int...CPU,22700.html

Yepp, I've tried in Feb 2013...

Quote:

Originally Posted by ixfd64

I may be wrong, but I believe the 64-bit version of Prime95 could handle 64 threads.

Yes, but Prime95/mprime wouldn't run on Xeon Phi...
Xeon Phi

has no MMX
has no SSE
has no AVX
is x86 compatible
is not binary compatible, you must recompile your code for Xeon Phi

I've tried ewmeyers mlucas. I did compile, It runs singlethreaded (only SSE code is multithread which I can't use), and it is horrible slow: mlucas singlethreaded on a E5-2670 was ~50 times faster than a single core of Xeon Phi in my test.
I failed to run glucas on Xeon Phi.

Oliver

P.S. My feeling tells me that you'll need to code an LL test explicit for Xeon Phi to achive good performance... but how many users would run it? There are no consumer cards for Xeon Phi.

TObject · 2013-06-17, 19:04

The Xeon + Xeon Phi powered supercomputer takes the top spot. Titan moves to the second.

http://www.anandtech.com/show/7075/j...takes-top-spot

ewmayer · 2013-09-30, 02:24

Quote:

Originally Posted by TheJudger

Xeon Phi

has no MMX
has no SSE
has no AVX
is x86 compatible
is not binary compatible, you must recompile your code for Xeon Phi

The above ma be true for current releases, but the roadmap here looks very promising: Within 2 years Xeon Phi will feature next-gen AVX with 512-bit SIMD registers [versus 256-bit for current AVX], twice as many of these [32 versus the current 16], in a GPU-like manycore compute fabric. It sounds frickin' awesome.

Oliver, I have just finished work on a beta version of Mlucas with pthreading-also-for-non-SIMD-builds ... PM me if you're interested in giving it a try. It can make use of up to 32 threads in LL-test mode [and up to 64 in Fermat-mod mode].

frmky · 2013-09-30, 05:23

I've played with the Xeon Phi. As previously mentioned, you need to recompile the code for the Phi without ASM, and to sufficiently hide memory latency you need to run 240 threads.

ixfd64 · 2013-10-04, 00:06

Does anyone know if Xeon Phi would be useful for trial factoring?

CRGreathouse · 2013-10-04, 01:11

Quote:

Originally Posted by ixfd64

Does anyone know if Xeon Phi would be useful for trial factoring?

I imagine it would be much worse than a comparably-priced GPU -- the extra power of these cores wouldn't help much for sieving.

2013-05-23, 02:58	#1
TObject Feb 2012 3⁴×5 Posts	Xeon Phi 60 LL tests running side-by-side, anyone? http://www.tomshardware.com/news/Int...CPU,22700.html

2013-05-23, 05:14	#4
LaurV Romulan Interpreter Jun 2011 Thailand 23×419 Posts	that would not be a problem, virtual box (or equivalent), ten copies of the guest OS running with 32 cores allocated to each, P95 running there... just give me the wheelbarrow with the 320 cores... For the phi, the real problem will be its common caches, in fact.. because no matter how fast your dog can eat, you still must be able to feed him so fast... So, in spite of having 7-8 times more cores then a "top" 8-cores Xeon E5-2670, it can only get 3-times more DP performance. That would be about 5- to 6-times faster compared with a i7-2600k (as everybody know what a 2600k is, but not everybody knows what a e5-2xxx is ). edit: and this is for clock per clock, but don't forget the 2600k has a faster clock, and runs much overclocked too, in water-cooled rigs. The pessimist in me says we won't see a more than 5 times performance increase, and that is already reachable with cudaLucas. Next big hit would be a FFT library for Radeons... Last fiddled with by LaurV on 2013-05-23 at 05:40

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Xeon vs. Quad CPU (775)	EdH	Hardware	19	2017-06-08 22:06
Motherboard for Xeon	ATH	Hardware	7	2015-10-10 02:13
Intel® Xeon Phi	pinhodecarlos	Hardware	2	2015-02-10 18:42
New Xeon	firejuggler	Hardware	8	2014-09-10 06:37
Dual Xeon Help	euphrus	Software	12	2005-07-21 14:47

2013-05-23, 04:21	#2
firejuggler Apr 2010 Over the rainbow 19·137 Posts	I don't know if P95 can handle more than 32 core. I recall a thread that stated that even with 2 instance of P95 running, it wouldn't use the 48 core of the machine.

2013-05-23, 04:57	#3
ixfd64 Bemusing Prompter "Danny" Dec 2002 California 2×3²×7×19 Posts	I may be wrong, but I believe the 64-bit version of Prime95 could handle 64 threads.

2013-06-17, 19:04	#7
TObject Feb 2012 3⁴·5 Posts	The Xeon + Xeon Phi powered supercomputer takes the top spot. Titan moves to the second. http://www.anandtech.com/show/7075/j...takes-top-spot

2013-09-30, 05:23	#9
frmky Jul 2003 So Cal 2111₁₀ Posts	I've played with the Xeon Phi. As previously mentioned, you need to recompile the code for the Phi without ASM, and to sufficiently hide memory latency you need to run 240 threads.

2013-10-04, 00:06	#10
ixfd64 Bemusing Prompter "Danny" Dec 2002 California 2·3²·7·19 Posts	Does anyone know if Xeon Phi would be useful for trial factoring?