• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

nVidia working on their own multi GPU die solution

Arachnotronic said:
How is it "clear" that NVIDIA will lag AMD and Intel here?
Click to expand...

Seems like a strange interpretation of the original post.

Technically AMD has achieved the multi die solution with Ryzen so the lagging started with it's launch if thats how you wish to describe it.

Every time the multi die solution has been brought up in the past on this forum it's been said that it's impossible to make it work as it wouldn't function as 1. Things like no way it'll work as multi gpu is dead, lack of sli/crossfire profiles, etc. Not going to dig for it but it's in a couple of the many AMD bashing threads in VC&G forum.
 
Makes sense; the same manufacturing constraints are going to effect all manufacturers of large dies.
 
Kenmitch said:
Seems like a strange interpretation of the original post.

Technically AMD has achieved the multi die solution with Ryzen so the lagging started with it's launch if thats how you wish to describe it.

Every time the multi die solution has been brought up in the past on this forum it's been said that it's impossible to make it work as it wouldn't function as 1. Things like no way it'll work as multi gpu is dead, lack of sli/crossfire profiles, etc. Not going to dig for it but it's in a couple of the many AMD bashing threads in VC&G forum.
Click to expand...

- Might not be the right place to discuss this, but multi die solutions have existed previously in the form of various Multi-Chip Modules, including Athlon/Pentium4/Core 2 products. I'm nearly positive my old Q9550 was two dual core dies on the same package with a small kingdom's worth of L2 Cache to limit trips across the bus. AMD created an iteration on this old technique with Ryzen and the real nugget there is that each die is actually a fundamentally capable chip. The MCM (Multi Chip Module) is nothing new.

Getting a bunch of GPU dies to play nicely as a MCM is a wholly different beast, and AMD (and to a greater extent RTG) are fighting an arguably losing battle on two opposing fronts with a fraction of the resources of their competitors. The (seemingly) never ending saga of the RX Vega launch is instructive in this case.

Intel and NV's comparatively endless coffers of R&D money is not to be underestimated. Both have shown that they are very capable of tight course corrections, pinpoint execution and engineering their way out of a bad spot.

I would not say that AMD beating their competitors to the punch is a forgone conclusion at all.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top