3 ms·
This was explained well by early reviewers: the benchmarks are too different to compare at this point and a bit immature for Apples new architecture. Some of t
by invalidname 4y ago
This was explained well by early reviewers: the benchmarks are too different to compare at this point and a bit immature for Apples new architecture.
Some of the benchmarks only ran in Intel mode which puts Apple at an unfair disadvantage and might bias the results. Other benchmarks weren't designed for machines like the M1 Max and leave it well below 100% utilization even when natively compiled. A completely new cross platform benchmark is needed to do a proper comparison.
There are still some benchmarks on Ars comparing both architectures when applicable e.g when Intel launched the new CPUs and claimed they were faster than the M1 Ars ran the comparison to test that claim.
- danaris 4y agoIs there any reason to think that the benchmarks, once improved to more accurately capture what the Apple Silicon chips do, will show it performing worse than they do now? 'Cause if they're only likely to give better results later on, that just supports TFA's thesis even more...
- invalidname 4y agoI don't think showing misleading numbers supports anything. It's not like there were no benchmarks comparing Intel and M1. There were plenty of those and it's pretty easy to find them. So we should have a sense of what these "broken benchmarks" indicate. Showing a flawed misleading comparison for every laptop that comes out is redundant and misleading regardless of the direction.