davis.anthony
Veteran
I wonder what the costs are between running a large memory bus with GDDR6/X and a high amount of internal cache vs just using HBM3.
Cache capacity is just one factor that drives effective hit rate. If anything, all hit rate graphs are bound to the specific context of how one particular architecture makes use of that one level of cache hierarchy. It is useful to predict how different SKUs of the same architecture performs, but never a good portable performance predictor across generations of architectures, where you probably expect significant changes everywhere over one another.I mean, AMD has reduced the Infinity Cache on Navi 31 from 128MB to 96MB. Doing this backstep in specs would be fine if 96MB is sufficient, but if it's not(as in the situation being talked about here), then such a backstep feels an awful lot like they're just not producing a great product like they could, out of cheapness.
For cost efficiency one wants to reuse the MCD die across all models. An increase of the MCD integrated cache that appears to be justifiable if one looks solely at N31, may not be looking very appealing for smaller models. But on the other hand, the price AMD can achieve for the fastest version of N31 might justify the additional cost of stacking some additional cache dies. At the high end, the price/performance ratio is far from linear. Even relatively small performance increases at the top justify relatively larger cost increases for the manufacturer. That is not the case at the low end. Furthermore, stacking cache increases the cache size by a multiple, not just a small increment, so it ends up with a way bigger return (in terms of performance) than just a moderately larger cache on all MCDs.The Vcache chip for the 5800X3D represents a very significant percentage increase in silicon usage.
The alternative - adding a much larger L3 cache on the base chiplet - would be even less space efficient.
So it makes sense here. The performance potential unlocked is also very obvious.
For Navi 31, it's quite different. An extra 8MB of L3 per MCD would not represent a huge percentage increase for silicon needed. And again, it's all on the cheaper process compared to the compute die.
Upcoming RDNA 3 GPUs are presumably RX 7000 desktop GPUs. RX 6000 is RDNA 2."RX 6000 series and upcoming RDNA 3 GPUs"
What other RDNA3 GPUs is he referring to?
AMD Rage Fury with RDNA3?
I wonder what the costs are between running a large memory bus with GDDR6/X and a high amount of internal cache vs just using HBM3.
According to the patent application:Yeah, I've wondered along those lines too. The expected design for the high end is a big GPU surrounded by 6 chiplets which house L3 cache and memory controllers and PHYs, all sitting on an interposer.
Ah that's for CDNA3 and onwards.Notice there is no "MCD" in that document
A lot actually.Do we expect any changes to the shader core over RDNA 1 and 2?
Do we expect any changes to the shader core over RDNA 1 and 2?
I think it's going to be largely irrelevant overall, especially considering the issues are likely a very small number of people who buy 4090's, which in itself will be a very tiny portion of PC enthusiasts.I have a feeling this is gonna turn in to a huge advantage for AMD this cycle.
I think it'll overshadow the whole 4000 series and will loom large over nVidia for this generation, but it'll be interesting to see either way.I think it's going to be largely irrelevant overall, especially considering the issues are likely a very small number of people who buy 4090's, which in itself will be a very tiny portion of PC enthusiasts.
I think it'll overshadow the whole 4000 series and will loom large over nVidia for this generation, but it'll be interesting to see either way.