Carzy Zarx’s PC Gaming Emporium - Catch Up on All the Latest PC Gaming Related News

[JEMC]

Bofferbrauer2 said:

JEMC said:

The worst part is that they chose the worst time to make the split from GCN to RDNA+CDNA. GCN was a very capable compute architecture, which was kind of useless back then and took die space for more shaders, so they went and removed it... just when Nvidia launched added their compute heavy Tensor cores in their gaming GPUs, paving the way for DLSS and Ray Tracing.

If AMD had waited one extra gen, they would have had something to match or one-up Nvidia. But they didn't, and it has hurt them.

We'll see how long it takes for the merge, likely a couple of generations given how long they start planning new architectures before they become an actual product, and what does this merge bring to the table.

From what I understood, RDNA4 will be the last RDNA before UDNA. So they're certainly already working hard on it.

My comment came from the Tom's Hardware article:

Tom's Hardware [TH], Paul Alcorn: So, with UDNA bringing those architectures back together, will all of that still be backward compatible with the RDNA and the CDNA split?

JH: So, one of the things we want to do is ...we made some mistakes with the RDNA side; each time we change the memory hierarchy, the subsystem, it has to reset the matrix on the optimizations. I don't want to do that.

So, going forward, we’re thinking about not just RDNA 5, RDNA 6, RDNA 7, but UDNA 6 and UDNA 7. We plan the next three generations because once we get the optimizations, I don't want to have to change the memory hierarchy, and then we lose a lot of optimizations. So, we're kind of forcing that issue about full forward and backward compatibility. We do that on Xbox today; it’s very doable but requires advanced planning. It’s a lot more work to do, but that’s the direction we’re going.

But I hope haxxiy is right as that unification would take way too much time.