Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

SoyCapitanSoyCapitan

Suspended
Original poster
Jul 4, 2015
4,487
2,551
Paris
Interesting. It uses an API to seamlessly make multiple GPUs appear as one GPU to the OS and they will scale it up. It can turbo boost to higher clock speeds with an AIO cooler. Power hungry though.

The 'e' denotes how many GPUs. X2, X4, X8 etc


Process 7nm (Intel)
Architecture Xe
Stream Processors 12288 (6144 x2)
Core Clock 1600 MHz
Boost Clock 2718 MHz
Memory 32 GB 4D XPoint
Bandwidth 8 TB/s
Peak Performance 66.8 TFLOPs
Die Size 600mm2 (x2)
GPU Type MCM
Interconnect Unknown (TSV based)
Feature Level Direct3D 14_2
TDP 350W
Power Connectors 3x 8-pin
Launch 6/31 2020

 
Did they get me? The bastards :p

1er-avril-poisson-davril[1].jpg

Be alert...
 
Interesting.
....
Memory 32 GB 4D XPoint
Bandwidth 8 TB/s
....
Interconnect Unknown (TSV based)
.....

Those three make no sense together.

No direct VRAM ( only Optane/XPoint ). TSV and Optane/XPoint together makes about zero sense. TSV between packages even less so.

The bandwidth is Area 51 like if talking bi-section bandwidth. (as opposed to just globbing some number together. )


Some non April Fool's ( at least if Intel doesn't bork the development ) Xe context:

https://www.nextplatform.com/2019/0...openpower-for-exascale-dominance-with-aurora/
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.