By using this site, you agree to our Privacy Policy and our Terms of Use. Close
sc94597 said:
haxxiy said:

The . For actual GPUs in real-world scenarios, you can see the improvement was more like ~ 3 times between a GeForce 10 and a GeForce 40:

The Switch 1's chip was Maxwell, not Pascal. You'd want to compare the 4080 to a 980. That is more of a 5 times difference. 

The X1 in the Switch was also originally 20nm. The die-shrink to 16nm happened with later revisions, with a battery life improvement. 

The 980 used 28 nm. 20 nm and 16 nm are both a full node ahead of the 980 and other GPU Maxwells (their physical features are about the same size) and are in theory twice as power efficient. Key word: in theory.

The thing is that 20 nm had massive leakage issues for anything consuming more than a few W. So, TSMC branded their 20 nm with FinFets as 16 nm since they considered it a half-node worth of improvement.

The fact that the Switch uses Maxwell in its SoC is not anywhere as relevant as the fact that it uses 16 nm instead of 28 nm. The architectures are very similar, the nodes are very different.

Though if Nintendo aims for a battery life more like the OG Switch, then yeah, they have more leeway than I implied before, I'll give you that.