Why do ppl disrespect Raja? Read the Anandtech article on his return to AMD. Mark Papermaster asked him to develop a post-GCN, brand new architecture. Very similar goals to Jim Keller being hired back to AMD, post-FX CPU new architecture, Zen.
Raja delivered them Navi/RDNA. Despite AMD going through very tough financial problems during the 2014 - 2017 period. Ppl can go back and look at forums then, lots of ppl actually expected AMD to go bankrupt. RTG didn't get good funding until the end of 2017 when Zen success returned AMD to profitable levels.
Shoestring budget, while iterating GCN and trying to create a new architecture is going to cause issues and delays. They even tried to back-port Primitive Shaders of RDNA into the last GCN, Vega, without success. But clearly you can see attempts were made to improve GCN in graphics performance.
RDNA2 is the result of AMD being in a great position again, money rolling in 2018 and onwards. Heard the tech press talk about the Zen team of Suzane Plummer heading over to RTG to help them develop RDNA 2? Extra R&D, more funding, and its going to show in RDNA2.
I'm not sure I agree about his leadership skills. Vega was a terrible gaming card, HBM was a terrible bet in general for gaming cards and 4GB HBM1 was a death blow for Fury (not sure how much this was his fault, as this was ~2 years after he returned, but he still kept talking about HBM all the time) - but he's a well respected engineer, and maybe with unlimited funds at Intel he'll do better. However, I mostly dislike him because of marketing and his huge mouth, and he couldn't deliver on his marketing. Not saying it's all his fault, it's difficult combating both Intel and Nvidia with lower R&D funding, but marketing under his leadership was terrible.
AMD's Raja Koduri talks moving past CrossFire, smaller GPU dies, HBM2 and more. After hosting the AMD Capsaicin event at GDC tonight, the SVP and Chief
pcper.com
Raja promised me and our readers that we “would be really really pleased.” We expect to see Polaris-based GPUs across the entire performance stack.
Never Happened.
Raja further expands on it, telling me that in order to make multi-GPU useful and productive for the next generation of APIs, getting multi-GPU hardware solutions in the hands of developers is crucial. He admitted that CrossFire in the past has had performance scaling concerns and compatibility issues, and that getting multi-GPU correct from the ground floor here is crucial.
Never Happened.
The naming scheme of Polaris (10, 11…) has no equation, it’s just “a sequence of numbers” and we should only expect it to increase going forward. The next Polaris chip will be bigger than 11, that’s the secret he gave us.
Never happened (Polaris 20/30 don't count, they were just Polaris 10 rehashes). IIRC there was actually a Polaris 12 that was smaller.
VB: Does the Polaris brand supplant the Radeon brand? Koduri: It’s an architecture codename. It’ll still be Radeon something something on the box. But we didn’t have a consistent architecture name like our competitors do. It was hard, because for people, including yourselves and some of the...
venturebeat.com
This is Polaris 10 and that’s Polaris 11. In terms of what we’ve done at the high level, it’s our most revolutionary jump in performance so far.
[...]
This is very early silicon, by the way. We have much more performance optimization to do in the coming months. But even in this early silicon, We’re seeing numbers versus the best class on the competition running at a heavy workload, like Star Wars—The competing system consumes 140 watts. This is 86 watts. We believe we’re several months ahead of this transition, especially for the notebook and the mainstream market. The competition is talking about chips for cars and stuff, but not the mainstream market.
Spoiler: RX 480 was less efficient than most of the Maxwell cards (significantly less efficient than the higher-end Maxwell cards). It was also not a revolutionary jump in performance. Note that "The competition" released Pascal which was a really excellent release, and technically to this day AMD doesn't really have a card the outright beats the 1080ti (just one more week though). In addition, he would post stuff about Vega months ahead of its launch. For example,
this picture, more than a year before Vega 64 was released.
Raja went to Intel and ppl laughed it off, that Intel has no chance to be competitive in graphics since they were so far behind. Yet Tiger Lake with Xe iGPU is trading blows with Zen 2 APUs in performance. In a single generation Intel manage to come back from what, 3 gens behind on graphics performance?
We'll see, but Zen2 APUs are essentially are ~2 gens behind AMD's latest RDNA2 cores. Either way, the problem IMO isn't the hardware, but the software. The drivers will probably make or break Xe.
Also, still doing his marketing things:
www.anandtech.com