Oxford/Princeton AVX benchmarks

davieddy · 2012-05-30, 09:50

How much speedup over a Core2 has George (et al) achieved
in LL testing?

David

fivemack · 2012-05-30, 10:59

http://mersenne.org/report_benchmark...ic_cpu=4373214

is an Ivy Bridge,

http://mersenne.org/report_benchmark...ic_cpu=2150153

is a core2.

Overclocking causes a bit of confusion, but it looks as if a 4096k FFT takes 90ms on the Core2 and 25ms or so on the Ivy Bridge. So a bit over a factor three.

davieddy · 2012-05-30, 11:21

THX Tom.
But does AVX alter the size of FFT needed for a given exponent?

D

Dubslow · 2012-05-30, 19:55

Quote:

Originally Posted by davieddy

THX Tom.
But does AVX alter the size of FFT needed for a given exponent?

D

Not significantly.

davieddy · 2012-05-31, 11:24

To avoid comparing apples with oranges,
how much speedup has AVX made to a given Ivy Bridge?

(Sorry for any Oxymorons)

D

axn · 2012-05-31, 11:54

Quote:

Originally Posted by davieddy

To avoid comparing apples with oranges,
how much speedup has AVX made to a given Ivy Bridge?

About 30% (+/-). However, the speed gains are less, when run on multiple cores, due to memory bandwidth bottleneck.

davieddy · 2012-05-31, 19:36

Quote:

Originally Posted by axn

due to memory bandwidth bottleneck.

Has "memory contention" any relation to AVX?

BTW Don't you agree "bandwidth" is an overworked term?

axn · 2012-06-01, 03:11

Quote:

Originally Posted by davieddy

Has "memory contention" any relation to AVX?

Don't know what you mean by "memory contention". Googling doesn't help.

Quote:

Originally Posted by davieddy

BTW Don't you agree "bandwidth" is an overworked term?

Dunno. However, in this context, it is not.

davieddy · 2012-06-01, 10:14

Quote:

Originally Posted by axn

Don't know what you mean by "memory contention". Googling doesn't help.

Dunno. However, in this context, it is not.

Different cores (or whatever) contending (competing) for memory access.

"Bandwidth" is used in just about ALL contexts these days!

Try Googling "bone of contention",
or "I used to be a contender".

David

davieddy · 2012-06-01, 11:15

When Shannon (dubbed by Turing a poser) asked
von Neumann (he of "no hidden variables" notoriety)
what he should call his new term, the reply was:
"Call it Entropy. After all nobody understands what that means".

I wonder decreasingly as to why Boltzmann topped himself.

David

Dubslow · 2012-06-01, 19:47

Quote:

Originally Posted by davieddy

Different cores (or whatever) contending (competing) for memory access.

"Bandwidth" is used in just about ALL contexts these days!

Yes; they're too darned fast. With more bandwidth, there would be enough room for all four/whatever cores to talk to mem as necessary, so there wouldn't be any contention.

2012-05-30, 09:50	#1
davieddy "Lucan" Dec 2006 England 2·3·13·83 Posts	Oxford/Princeton AVX benchmarks How much speedup over a Core2 has George (et al) achieved in LL testing? David

2012-05-30, 11:21	#3
davieddy "Lucan" Dec 2006 England 6474₁₀ Posts	THX Tom. But does AVX alter the size of FFT needed for a given exponent? D Last fiddled with by davieddy on 2012-05-30 at 11:24

2012-06-01, 11:15	#10
davieddy "Lucan" Dec 2006 England 2×3×13×83 Posts	Terminology When Shannon (dubbed by Turing a poser) asked von Neumann (he of "no hidden variables" notoriety) what he should call his new term, the reply was: "Call it Entropy. After all nobody understands what that means". I wonder decreasingly as to why Boltzmann topped himself. David

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Orienteering in Oxford	davieddy	Lounge	20	2011-08-21 11:46
Benchmarks	MurrayInfoSys	Information & Answers	3	2011-04-14 17:10
LLR benchmarks	Oddball	No Prime Left Behind	11	2010-08-06 21:39
benchmarks	Unregistered	Information & Answers	15	2009-08-18 16:44
Benchmarks for i7 965	lavalamp	Hardware	21	2009-01-06 04:32

2012-05-30, 10:59	#2
fivemack (loop (#_fork)) Feb 2006 Cambridge, England 2³×11×73 Posts	http://mersenne.org/report_benchmark...ic_cpu=4373214 is an Ivy Bridge, http://mersenne.org/report_benchmark...ic_cpu=2150153 is a core2. Overclocking causes a bit of confusion, but it looks as if a 4096k FFT takes 90ms on the Core2 and 25ms or so on the Ivy Bridge. So a bit over a factor three.

2012-05-31, 11:24	#5
davieddy "Lucan" Dec 2006 England 194A₁₆ Posts	To avoid comparing apples with oranges, how much speedup has AVX made to a given Ivy Bridge? (Sorry for any Oxymorons) D