Oxford/Princeton AVX benchmarks
 

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

David

[url]http://mersenne.org/report_benchmarks/?exover=1&exbad=1&specific_cpu=4373214[/url]

is an Ivy Bridge,

[url]http://mersenne.org/report_benchmarks/?exover=1&exbad=1&specific_cpu=2150153[/url]

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.

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

D

[QUOTE=davieddy;300693]THX Tom.
But does AVX alter the size of FFT needed for a given exponent?

D[/QUOTE]

Not significantly.

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

(Sorry for any Oxymorons)

D

[QUOTE=davieddy;300822]To avoid comparing apples with oranges,
how much speedup has AVX made to a given Ivy Bridge?
[/QUOTE]

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

[QUOTE=axn;300824]due to memory bandwidth bottleneck.[/QUOTE]
Has "memory contention" any relation to AVX?

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

[QUOTE=davieddy;300871]Has "memory contention" any relation to AVX?[/quote]
Don't know what you mean by "memory contention". Googling doesn't help.
[QUOTE=davieddy;300871]BTW Don't you agree "bandwidth" is an overworked term?[/QUOTE]
Dunno. However, in this context, it is not.

[QUOTE=axn;300910]Don't know what you mean by "memory contention". Googling doesn't help.

Dunno. However, in this context, it is not.[/QUOTE]

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

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

[QUOTE=davieddy;300935]Different cores (or whatever) contending (competing) for memory access.

"Bandwidth" is used in just about ALL contexts these days!
[/QUOTE]

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.

[QUOTE=Dubslow;300976]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.[/QUOTE]
Don't really want to get too deep into this
(a few years ago I would have done)
But one of the "speed" tricks in my day was doing as much
as possible in "registers" and accessing memory as infrequently as possible.
Memory access has since got faster, so possibly is used too much these days?

D

[QUOTE=davieddy;301007]Don't really want to get too deep into this
(a few years ago I would have done)
But one of the "speed" tricks in my day was doing as much
as possible in "registers" and accessing memory as infrequently as possible.
Memory access has since got faster, so possibly is used too much these days?

D[/QUOTE]
Ask George. I'm pretty sure his assembly already uses the on die cache (and registers) as much as possible. I've certainly seen him describe it as "cache-friendly".

[QUOTE=Dubslow;301008]Ask George. I'm pretty sure his assembly already uses the on die cache (and registers) as much as possible. I've certainly seen him describe it as "cache-friendly".[/QUOTE]
George is 7 years younger than me.
As I intimated, a few years ago I would love to have talked to him about shit hot programming.

D

To whom whoever "corrected" Oxbridge to Oxford, would he/she kindly replace the original.
Any humour in the title would otherwise be lost.
You may as well replace [url=http://www.youtube.com/watch?v=rmTrKaDrpbw]"Ivy League"[/url] with Princeton.

Snob

[QUOTE=davieddy;301031]To whom whoever "corrected" Oxbridge to Oxford, would he/she kindly replace the original.
Any humour in the title would otherwise be lost.
You may as well replace [url=http://www.youtube.com/watch?v=rmTrKaDrpbw]"Ivy League"[/url] with Princeton.

Snob[/QUOTE]

Mike perhaps? (Did he also perchance [strike]remove an "f" from[/strike] replace the title of a recent thread in S&T forum?)

[QUOTE=Dubslow;301032]Mike perhaps? (Did he also perchance [strike]remove an "f" from[/strike] replace the title of a recent thread in S&T forum?)[/QUOTE]

"f"? Don't beat about the bush.

Anyway I'm really Dark Blue through and through.
(You may need to consult Paul or Tom on that one)

D

[QUOTE=davieddy;301037]"f"? Don't beat about the bush.

Anyway I'm really Dark Blue through and through.
(You may need to consult Paul or Tom on that one)

D[/QUOTE]

No really, I could have sworn it was spelled correctly when originally posted, but later it was misspelled.

[QUOTE=Dubslow;301038]No really, I could have sworn it was spelled correctly when originally posted, but later it was misspelled.[/QUOTE]
This is getting spooky.
Bedtime!

xxx

[url=http://www.youtube.com/watch?v=WQ4ZGGt-1rs]I put a spell on you[/url]

Ok, I have access to a 3.1GHz Ivy Bridge.

AVX enabled, P-1 tests. Getting 25GHz-days/day on 4 cores, 21.5GHz on 3 cores. Slight memory starvation @1600MHz ram when 4 cores run stage2 at same time.

Setting highmemworkers=3, but very hard to get benchmarks (as work is paused/restarted etc..). Rough estimate 26GHz/days per day.

-- Craig

[QUOTE=nucleon;301173]Ok, I have access to a 3.1GHz Ivy Bridge.
-- Craig[/QUOTE][URL="http://www.youtube.com/watch?v=eK9E6-Eu3-Y"]Key to the Highway?[/URL]

D