A topic on Variable Frequency & Power:
Bear with me here long post incoming. It will be worth a read. And hopefully, it'll be able to clear some misunderstanding regarding the variable frequency of PS5. I'll do my best to explain with the limited info we have available to the best of my ability. And pardon my English as it's not my native language.
Instead of asking "Why didn't MS implement variable frequency?"
A better question is why did Sony choose to NOT implement fixed frequency?
It's simply because the fixed frequency strategy didn't allow Sony to reach where they wanted to be in terms of frequency. What does that mean? Their goal from the get-go for PS5 was to set the GPU frequency as high as possible.
As Cerny mentions:
But with the fixed frequency strategy, he couldn't go as high as he wanted to, more on this below.
So what's this "old fixed frequency strategy"?
It is to run the GPU at a constant frequency at all times and let
power vary based on the GPU workload. Btw, GPU power draw varies A LOT from game to game and even scene to scene in a game. Anyway, this is the exact same strategy that PS4, PS4 Pro, and Xbox consoles from XB1 to Series X/S use.
And why was 2 GHz unreachable?
Cerny's above "2 GHz was looking unreachable" comment doesn't mean the chip was having thermal issues nor was it incapable of reaching 2 GHz. In fact, the chip was capable and had the potential to go even further than 2 GHz comfortably. It was simply due to this old strategy of
power being variable, they were unable to achieve their goal of crossing 2 GHz, in other words the power being supplied to the chip was likely insufficient. Now with the
variable frequency strategy, "a completely different paradigm" as Cerny calls it, they're able to fully tap into the full potential of the GPU in terms of its frequency. More on this below.
So what is this different paradigm and how are they achieving higher frequency with this
variable frequency strategy?
To this Cerny answers:
So what was previously an unreachable 2 GHz is now way over 2+GHz.
So the goal of 2.2+ GHz is now achieved and this was done so, in Cerny's words, by "supplying a generous amount of electrical power" i.e. constant power. And they had to cap it at 2.23 GHz to ensure the logic operated properly. And SmartShift was used as a supplement to this strategy for the SoC to further boost performance/pixels. SmartShift isn't the solution to variable frequency, btw.
Why does Cerny like running the GPU at a higher frequency?
Suppose Sony had gone with the same 56 CUs as MS with 4 disabled (52 CUs active) for PS5 and with the goal of 10.3 TF. They would have to set the frequency at 1544 MHz to achieve that. They would surely achieve their goal of 10.3 TF but the performance would be noticeably different between 36 CU at 2230 MHz vs 52 CU at 1544 MHz. Cerny even gave this e.g. but with 36 vs 48 CU.
At Hotchips MS revealed their "GPU Evolution" slide for Series X comparing its GPU all the way from the original Xbox One GPU.
For the GPU evolution slide, they focus on 4 metrics here to show the evolution.
- Computational power
- Memory bandwidth
- Rasterization rate and
- Pixel fillrate.
Here's how a notional PS5's GPU perf with 52 CU @1544 MHz config would look like:
10.3 TFLOPS, 448 GB/sec, 6.18 Gtri/sec, 98.8 Gpix/sec
Here's what the actual PS5's GPU perf looks like with the current 36 CU @2230 config:
10.3 TFLOPS, 448 GB/sec, 8.92 Gtri/sec, 142.7 Gpix/sec
You reach your TF goal but look at rasterization and pixel fillrate. They take a massive hit. With 36 CUs those other units are now 44% higher than they were with 52 CUs. Not only does rasterization go up, but pixel fillrate also goes up along with the processing of the command buffer. And L2 and other caches now get 44% more bandwidth. Not to mention it would cost more money to go with a larger GPU. Please note I'm not downplaying MS here. Just pointing out the strategy that Sony has taken based on the information that's out there. Whatever strategy MS followed for their machine works the best for them and I'm not downplaying that one bit.
Hope this post was able to better explain Cerny's variable frequency topic just a bit more and why they went with it. There is obviously more to talk about it and there's still not much detail and info like what's the power consumption like, etc.