techpowerup.com
G-Jow Forumsoyim, I...
O-oi vey, S-SHUT IT DOWN
Brandon Garcia
Other urls found in this thread:
archive.is
techpowerup.com
twitter.com
Samuel Young
>using intelbench r20 from wangblows (((store)))
Dylan Ortiz
>intelbench r20
based
Angel Taylor
yup, embree.github.io is used by intelbench r20, what does this mean i do not know.
James Thomas
>Overpriced i9 and Xeon housefires lose heavily to much cheaper and way more efficient Zens
>in Intel biased bench
>SHUT IT DOWN
Easton Morales
according to their license, if you get kikebench from their site, you cannot modify it and redistribute it.
so, if you get it from the windows repository, that restriction does not apply.
archive.is
Robert Campbell
yeah Intel win cause native 256bit instructions. you'll add CB R20 DOESN'T MATTER! sooner than later when Zen2 can do it too
Joseph Wilson
>native 256 bit instructions
Zen does AVX 2 very well already, you dumb coping Intbecile.
Grayson Sanchez
Zen does 256bit AVX2 by fusing 2 128bit units. Zen 2 (Ryzen 3000) will have native 256bit units that will basically double AVX2 performance.
Landon James
It's SHIT benchmark, so noone should care...
Dominic Powell
>2x128bit
calm down man. 4Ghz 9900k performs better than a 2700X at 4.1Ghz by like ~250 points. nothing to be ashamed of when consindering this. now calm your company wars since no one got the best intentions just for consumers sake but to make money, AMD seems like the least slimiest of the bunch right now
Isaac Butler
Wrong. Zen does AVX 2 by brute forcing through it, while Zen 2 will be brute forcing through AVX 512.
Parker Bailey
intel will have to move to avx1024
and inteltards will have to defend 1024bit precision in floating point.
how's that protein folding simulation going, inteltards?
Charles Sanders
>protein folding
Ironically enough, Mellanox and nGreedia's CUDA are the best at mining folders @ home.
Cameron Stewart
so can someone fix me a link now that's it's gone from both guru3d and techpowerup?
Levi Diaz
>what is internet
Juan Russell
No it doesn't. Zen 2 has no support for AVX512. en.wikichip.org/wiki/amd/microarchitectures/zen_2
What does "brute forcing" even mean? Guessing the correct result? Cite a better source or be silent.
Thomas Green
>Zen 2 has no support for AVX512
Learn reading comprehension, you dumb fuck.
Asher Perez
I dont think you understand what brute forcing mean
Charles Lewis
You can run AVX256 on Zen 1, it does it by using 2x128. You CAN NOT run AVX512 on Zen 2 even tho it has 256bit units. Can't put it any simpler, sorry.
Bentley Collins
>You CAN NOT run AVX512 on Zen 2
Again, Intbecile: reading comprehension. Learn it.
Ryan Williams
ironically enough, novidia gpus were shit as accelerators outside the cuda sandbox, until novidia spent 2 years ((((((contirbuting)))))) code to those "oss" libraries and made everything cuda dependent.
Aaron Foster
Colton Jones
Sorry, can't comprehend nonsense. And I'm not running intel.
Nathan Jones
You can't comprehend that Zen doesn't have native AVX 512 support, but merely brute forces through amount of calculations it does (thus being better at AVX 512 than Inturd, while not having actual AVX512 on it's board)? What a fucking retard you are. Learn to count 2+2.
Oliver Flores
I doubt you know what AVX512 is. Hint: it's not 2x AVX256 nor is it AVX256 on 512bit.
Henry Reyes
Again, you DUMB DENSE idiot: Zen doesn't have AVX512, but AMD simply BRUTE FORCES through the amount of calculations EQUAL TO AVX 512's by increasing it's AVX 2 pipeline roughly by twice. It's NOT AVX 512 that AMD does, but TWICE THE AVX2.
Carson Peterson
enjoy your latency
Luis Rodriguez
>BRUTE FORCES
you keep using that word, care to explain what do you mean?
Jose Perry
>simd
It's a CPU, not GPU, who fucking cares.
Grayson Green
Again, AVX512 is not only about "calculations", try again. This time after reading what AVX512 is.
Cameron Perry
|
|>
|
|
|
Caleb Adams