>Mid June 2022
>Still no good Intel consumer GPUs
So, did he con Intel too?
Mid June 2022
Levi Clark
Jeremiah Thomas
Dear Sir kindly please revert me
Thomas Jackson
i would say that it's pretty cool that you can now have i5-2500k power in a $200 11" netbook.
Jace Foster
It's a shame he wasn't able to redeem intel
Caleb Taylor
he did the needful
Zachary Turner
I still have hope. I want to build a Linux-based 3D workstation for Blender (as well as other 3D software like Substance and Houdini) before the end of this year and finally get away from Windows, so I hope Intel Arc has good rendering performance for a cheaper price than the other two. Nvidia is pretty much a no-go on Linux, and while there is AMD, I don't want to pull the trigger too soon if Intel GPUs are going to have the better value-to-performance ratio.
Robert Anderson
bloody bastard bich
Adam Rivera
Landon Young
Vega was the worst thing that ever happened to Radeon, and I blame that Pajeet for that.
He's a Nvidia spy infiltrated to destroy Intel and AMD from inside
Jacob Cook
whata fak man
Cooper Martin
It's not his fault intel fabs are up to shit.
Remember 14nm+++++++++
Michael Fisher
Vega high end was bad, but it was decent enough to be in iGPUs for a long time.
Jayden Long
He finished Vega but his main work at AMD was RDNA.
Isaac Cooper
>his main work at AMD was RDNA.
rdna is mark cerny's brainchild
Mason Jones
And Zen is Mike Clark's brainchild. Raja was still supervising his division as was Keller.
Grayson Price
Sony said they worked closely with AMD on RDNA2, not RDNA. And that wasn't even fully true. RDNA2 was just a vanilla architecture that any of AMD's customers including MS and Samsung had access to for integration.
Zachary Davis
Oh yea, and then the Principle Graphic Engineer at Sony said some dumb shit like "we're not using pure RDNA2" and then when people asked what he meant he said "RDNA2 but actually RDNA cores". As it turned out, the PS5 just had a normal RDNA2 GPU according to AMD. One of Mark Cerny's lead engineers didn't even know what he was talking about. Sony just licensed an off the shelf design that was coming down the pipeline, the only input they had the configuration of their APU
Parker Rodriguez
It's not RDNA2 in the proper sense of what was released out to desktop, that's why they can't say it's RDNA2 and can only say it's based on it.
>// GFX1013: "target-features"="+16-bit-insts,+ci-insts,+dl-insts,+dpp,+flat-address-space,+gfx10-insts,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst"
>// GFX1030: "target-features"="+16-bit-insts,+ci-insts,+dl-insts,+dot1-insts,+dot2-insts,+dot5-insts,+dot6-insts,+dot7-insts,+dpp,+flat-address-space,+gfx10-3-insts,+gfx10-insts,+gfx8-insts,+gfx9-insts,+s-memrealtime,+s-memtime-inst"
PS5 is gfx1013, which we know from Linux support for the AMD Ryzen 4700s which are repurposed defective PS5 dies. You can see it's primarily missing instructions around DP4a or the dot instructions for matrix math most important for machine learning. Partially is probably why AMD didn't use AI for FSR 2.0.
Justin Ward
It's too late for Arc. It won't be able to compete with the 4000 series or RDNA3 mid-low end cards at least not without selling at a loss. If they were released in April they might have made some money but who are we kidding, their Xe drivers are hot garbage.
Mason Jenkins
Vega was actually great. The yields and Globalfoundries were the problem.
Chase Thompson
I'm talking out of my ass right now but you have to consider how good the drivers are going to be from Intel right? I've seen "rumor" videos of people saying intel is delaying arc because of bad drivers in windows, so i wonder how good the drivers are gonna be in linux