AMD CDNA Discussion Thread

This is the one used by Frontier not the one in page 8. (only Description is in page 8). The CPU connection uses the external IF links.
One link goes to Slingshot 200Gbps

 

Right, it seems the CPU and NIC connections ride the IF links. So GPU P2P can't use the full 800GB/s in practice.

Another interesting thing is that IF tops out at 25Gbps per pin over an interposer while NVLink is at 50GBps per pin over a PCB. Shouldn't IF be much, much faster on the interposer?

 

And NVidia calls its stuff open.

LOL

 

IF is couple of types. IFOP and IFIS. IFIS is designed for long range. IFOP has higher BW but designed for low energy inter die. IFOP have more BW than IFIS.
Both are designed for low latency instead of BW because of CPU roots.

 

NICs are behind MI200's extra PCIe links, not IF.

 

The SERDES'es used can do both, IF (IFIS) and PCIe, but only one at the same time. So it's running on PCIe protocol, but blocking one of the eight IFs.

 

Nah, only one link contains the actual PCIe root per GCD.

 

I didn't say otherwise ("The SERDES'es used"). Please refer to context.

 

Well, I didn't mention any names, but if you ask… As for your original posting about bandwidth:

For CDNA you choosed to use lower / directional value (without mentioning it) and negative connotation:

For A100 you choosed to use higher / bi-directional value (again, without mentioning it) and positive connotation:

Hard to keep technological discussion, when like half of the posts have to be dedicated to deciphering of biased and misleading content, mixing apples and oranges etc.

 

This isn't positive or negative, it was in the context of explaining power budget. And it wasn't definitely in the context of comparing inter die link speeds, which has to be fast enough to handle communication between two GPUs.

Maybe if you stopped digging through years old posts and started discussing technical points based on their merits alone.

 

Me wants!


Also, me cannot afford or need one

 

there:

https://www.hpcwire.com/off-the-wir...-healthcare-research-with-exascale-computing/

 

>meme flops
where's the real deal

 

Looks like 2 Fury X glued together.

 

The glue transfers 400 GByte/s (both directions combined).

 

That's good glue I'll give you that.

 

Not sure how many people watched this video with results from @AMD MI200 (MI250x) prepared by AMD/ORNL and others. Quite good results. Video: https://t.co/xFS12CwNEz #HPC https://t.co/IoDzyZdPc4

 

Just link the tweet.