| sc94597 said: By the way, this is why extrapolating TFLOPs from a Steel Nomad Light benchmark is a bad idea. Here are the TFLOPs and average Steel Nomad Light benchmark scores for various GPU's in that range. Notice that an R9 285 has a similar score to a GTX 1050ti, despite the R9 285 being a 3.25 TFLOPs card and the GTX 1050ti being 2.2 TFLOPs.
In fact if you do a simple linear regression on these 12 data points (with Steel Nomad being the dependent variable), you get an R^2 of only about .287, and an insignificant p-value (assuming .05 alpha.) I hope this table also shows why TFLOPs aren't a good measure of performance, even for synthetic benchmarks like Steel Nomad Light. Even still, using the equation from that simple linear regression you get something like a predicted 2.825 TFLOPs +/- 0.4 TFLOP given a Steel Nomad score of 2205. But really we don't need to do this to know what SW2's TFLOPs are. Again, TFLOPs are exactly a function of max clock rate and core count. We know both of these. Docked FP32 TFLOPs = 1536 Cores * 1.007 Ghz x 2 ~ 3.09 TFLOPs. Handheld FP32 TFLOPs = 1536 Cores * 561 Mhz x 2 ~ 1.72 TFLOPs. |
Well I've got a RTX 2050 mobile laptop which is basically Switch 2 GPU architecture but over twice the power and compared to the RX 580 (Xbox One X) it is still weaker in real world tests as you would expect. The Switch 2 is under half that performance when docked.
https://pc-builds.com/compare/gpu/0Yg1aB/radeon-rx-580/geforce-rtx-2050
Lets not forget Nintendo does not give out full specs and development hardware is not retail hardware. The Geekerwan analysis is the best we have at the current time. Lets also not forget the Switch 2 is based on a dated mainly 10Nm fabrication process which is very power hungry and it only comes with a 19Wh battery well below other mobile gaming devices. This idea that somehow Nintendo have clocked the system to high levels despite always clocking low previously and simply not having that level of power. Even the Geekerwan analysis only gave the theoretical peak figures for GPU performance when portable, likely below 1 Teraflop in real terms a lot of the time. This is nothing new we know Nintendo always clocks lower, they always have, it means greater reliability, longer battery life and less returns for them, a cooler running system is a longer life unit. It's what Nintendo does to maximise profits. Why do we keep getting people pretending Nintendo have clocked to the maximum clocks. Development hardware is always clocked higher than retail hardware especially where Nintendo is concerned. I don't get why we have to pretend the Switch 2 is more powerful. The Geekerwan analysis was brilliant it gave us a decent and a REALISTIC viewpoint of the hardware without fanboy nonsense. Nothing we are seeing in how the Switch 2 performs is contradicting it. Ultimately again we are in a situation where people just want to believe what they want to believe which is pure nonsense. The Switch 2 is a decent portable handheld but its based on a fabrication process from 2020 and has a tiny battery and yet people keep pretending its clocked to high limits. The reason the Switch 2 can perform well is the fantastic DLSS upscaling and of course its a fixed platform so can be fully optimised to work around weaknesses. Some games do this and some do not, quick and dirty ports to Switch 2 show a very low performance level.
