IC: With the FP units now capable of doing 256-bit on their own, is there a frequency drop when 256-bit code is run, similar to when Intel runs AVX2?
MP: No, we don’t anticipate any frequency decrease. We leveraged 7nm. One of the things that 7nm enables us is scale in terms of cores and FP execution. It is a true doubling because we didn’t only double the pipeline with, but we also doubled the load-store and the data pipe into it.
IC: Now the Zen 2 core has two 256-bit FP pipes, can users perform AVX512-esque calculations?
MP: At the full launch we’ll share with you exact configurations and what customers want to deploy around that.anandtech.com /show /13578 /naples-rome-milan-zen-4-an-interview-with-amd-cto-mark-papermaster
AHAHHAHAHAHAHAHAHAHHAAHAH intel is trully going to get assraped without even lubricant this time
Attached: 052.png (672x794, 422K)
November 13, 2018 - 09:38
NOOOOOOOO THIS CANT BE HAPPENING
Attached: 1539344654359.jpg (1035x1000, 189K)
November 13, 2018 - 09:41
Eh, but the roundabout answer to the second question reads pretty much as "no" to me
November 13, 2018 - 09:50
if the data pipe is trully doubled then it doesnt matter
November 13, 2018 - 09:52
It's funny how Rome is completely destroying every single one of these damage control slides.
Attached: Intel-AMD-Naples-Reply-7-1080.179689716.png (1500x844, 69K)
November 13, 2018 - 10:03
How it doesn't matter? You either can do it or you can't. If you can, why didn't ge just say you can?
November 13, 2018 - 10:09
IT'S NOT FAIR INTELBROS WE MUST FIGHT FOR WHITE GENOCIDE
Attached: 1542027477262.png (807x745, 205K)
November 13, 2018 - 10:10
because if the data throughput can remain the same but on a lower freq hence matching intel
it really does not matter
November 13, 2018 - 10:10
Dude asked if the two 256bit units can be used to perform 512bit operations. AMD guy chose not to say "yes". How does it NOT matter?
November 13, 2018 - 10:13
LISA SU IS LITERALLY HITLER ITS ANUDDA SHOAH OY GEVALT
Attached: 1542070594742.png (1295x1180, 1.16M)
November 13, 2018 - 10:18
November 13, 2018 - 11:21
he literally said that we will reveal the specific configuration about this when its released
so its yes but its a more complex YES
November 13, 2018 - 11:45
It's probably the same as Ryzen on 14nm and AVX-256
November 13, 2018 - 11:52
Attached: 1541892583390.jpg (818x693, 91K)
November 13, 2018 - 12:37
Attached: 1541542283130.jpg (3638x960, 382K)
November 13, 2018 - 13:03
DELETE THIS ANTI-SEMITIC SHIT RIGHT NOW!!!
Attached: IntelDelidThis.png (882x758, 192K)
November 13, 2018 - 13:26
OH NO NO NO NO NO NO NO
Attached: POO IN GPU.png (760x594, 211K)
November 13, 2018 - 13:27
Those are like trade cards, love em.
Attached: 1541545242451.jpg (1536x2048, 470K)
November 13, 2018 - 13:30
POO IN JOO >POO IN JOO POO IN JOO >POO IN JOO
November 13, 2018 - 15:30
G-POO-U
Attached: CANT NO COMPETE.png (714x748, 313K)
November 13, 2018 - 17:03
just delid and delap it bro
November 13, 2018 - 17:04
and buy a industrial cooler
November 13, 2018 - 17:19
>IF link between CPU and GPU is feasible but not announced this time
you know what is coming niggas
November 13, 2018 - 17:40
November 13, 2018 - 18:54
>needing ovens to set intel CPUs on fire
November 13, 2018 - 18:58
GPU chiplets built into CPUs IO communication latency and speed between chiplets is extremely good.
November 13, 2018 - 19:40