News Ryzen 9 7945HX Games 10% Faster Than Core i9-13950HX, According to AMD Benchmarks

I kind of want to see what that note under the graph says, LOL.

Marketing is evil, never forget that. Those foot notes are very important.

That being said, and the initial reviews I've managed to read, all point to good things around the Zen4-based mobile SKUs.

Regards 😀
 
Another fake review from AMD. Instead of comparing it with the fastest 13980HX, they've compared it with the slower 13950HX . :joycat:
 
how do you extrapolate a 4% performance improvement from a 1.8% clock speed difference between the 13950hx to the 13980hx?

Then why not compare it with Intel's best instead of comparing it with something slower? Typical AMD.

For starters, the 13980HX is faster than 13950HX by almost 7.5% in some benchmarks.

Comparing Ryzen 9 7945HX with the slower Core i9-13950HX is pretty much stupid. It should be compared with the faster Core i9-13980HX. Then the performance delta will be only around 6%
 
Thing about laptops is that due to different thermal capacities and their cTDPs, we know that different laptops even with the same chips perform differently. so I wouldn't hold my breath that it will be 10% faster outside some high end models.
 
  • Like
Reactions: KraakBal
It is faster. No doubt. But not by 10%. But by a meager 6%. Thats all.
Huh? That's only the 7zip benchmark (also, the delta is 6.8% because you're looking at the reciprocal rather than the complement). That ignores power-scaling on Cinebench and it ignores the Blender results.

The Cinebench results are pretty shocking, because I mean who really wants a laptop using 100 W or even 80 W? I have a "mobile workstation"-class laptop for my job, and I turn down the power settings even when it's plugged in. The fans are too obnoxious, otherwise.

BTW, those Blender results are sorted in the wrong order. It's measuring time, so the fastest CPU has the shortest time. If you look at the reported times, the 7945HX is 7.5% faster than the i9-13950HX.

With all that being said, the main thing to keep in mind is that laptops vary a lot in power parameters and cooling efficiency. So, we would ideally compare the performance several implementations of each CPU. It's error-prone to read too much into the performance measured from a single unit.