Quote from: Dont_Look_Up on Yesterday at 11:06:12I have a question regarding Geekbench. It seems that it is giving 1.5 times higher score to Apple M CPUs in comparison to any other Intel or AMD CPU than any other benchmark (like CB 23, CB 2024, Blender etc) gives. It creates such a false impression about the Apple M series and many articles use only Geekbench to compare CPUs...

For example in this article both Ryzen halo 395 and 390 are faster than M4 14 core in CB23, but in Geekbench, M4 is faster than the Ryzen 390 by 44% !

It is simply a biased and unreliable benchmark and I don't understand why reviewers use it at all.

JUST THE USUAL R23 CLOWN trying to justify x86 superiority with a bullshit x86 biased benchmark that was compiled for M1(even that was poorly done).
CB2024 M4P 1780
M4M 2150
STRIX HALO isnt winning s***

Typical x86 fanboy BS comment, Geekbench6 is fair comparison, it adopts all the x86 intrinsics, AVX2/512, VNNI, and lots of other hardware accelerators like DLboost. All you c l o w n s talking cheap BULLCRAP like "SME2 is cheating" KYS!! IF SME2 is cheating than AVX2, AVX512, VNNI and DLboost is cheating too!!!

M4MAX 16incher gets 187/2150 on CINEBENCH2024, for comparison 160w++ PL1 285HX gets 135/2200.
In comparison LLVM 16.0 Ninja Bulid takes 233s for M4MAX and 286s for DESKTOP 285K with 8000CUDIMM.

Accept once and forever that ARM64 is the superior platform, oh and not to mention SVE2.2 is just about to be adopted for M5 and NVIDIA DIGITS already has SVE2.1, even in fulll SIMD loads like Raytracing the M4MAX with a 4x128b /core NEON+FMLA(1clock cycle/NEON and FMLA efficiency) can almost catch up with 275HX.
SVE2 ARM CPU's will wipe the floor with crap x86 in SIMD loads. They already win in NOSIMD and SCALAR.

Quote from: AlexZ on Yesterday at 18:21:16Geekbench is crap.  If anything, it's only suitable for comparing x86 CPUs with each other, but not between x86 and Apple M.

Correct, but explain that to people who read articles online comparing Apple to other CPUs based only on Geekbench, The situation is so ridiculous that if you believe Geekbench, a 10W tablet with Apple M CPU has similar score to a top of the line 150W Ryzen Desktop CPU!

Go figure!

Quote from: Dont_Look_Up on Yesterday at 11:06:12I have a question regarding Geekbench. It seems that it is giving 1.5 times higher score to Apple M CPUs in comparison to any other Intel or AMD CPU

Geekbench is crap.  If anything, it's only suitable for comparing x86 CPUs with each other, but not between x86 and Apple M.

The UK does have some import tariffs (a few percent at most), whereas the USA has 20% on kit like this, soon to rise to ???%

UK imports and UK goods are all subject to VAT (like a sales tax as in NY), however businesses can reclaim this against VAT on items or services they sell [even if they keep it to use].

That pricing can't be right. I was looking in the UK on HP's own website and that's about the same price as the 64GB 395 with 1TB of storage which is £2,291.99

I have a question regarding Geekbench. It seems that it is giving 1.5 times higher score to Apple M CPUs in comparison to any other Intel or AMD CPU than any other benchmark (like CB 23, CB 2024, Blender etc) gives. It creates such a false impression about the Apple M series and many articles use only Geekbench to compare CPUs...

For example in this article both Ryzen halo 395 and 390 are faster than M4 14 core in CB23, but in Geekbench, M4 is faster than the Ryzen 390 by 44% !

It is simply a biased and unreliable benchmark and I don't understand why reviewers use it at all.