4 ms·

On desktop benchmarks no Apple CPU in top 20 (AMD beats the most powerful apple chip by 2.5 times): https://www.cpubenchmark.net/desktop.html https://www.cpube

On desktop benchmarks no Apple CPU in top 20 (AMD beats the most powerful apple chip by 2.5 times): https://www.cpubenchmark.net/desktop.html https://www.cpubenchmark.net/desktop.html On laptops it looks the same (AMD is 2.x times faster): https://www.cpubenchmark.net/laptop.html https://www.cpubenchmark.net/laptop.html So where is the astonishing performance Apple claims all the time? Yes, I know it consumes less power, but my laptop is docked 95% of the time, so I'd rather have more power than more battery life.

kitsunesoba 3y ago
Good perf per watt means more than just good battery life. It also means the laptop's fans won't be screaming like the machine is being burned alive when being pushed, that the chassis won't get uncomfortably hot easily, and and performance doesn't fall off of a cliff when you're unplugged. In short, it's about balance, which I think is just as important as raw performance for most people.
zamadatix 3y ago
The M2 Ultra will definitely crack the top 20 but it's no surprise a 24 core option is beat by 64 core competition, which tops that chart. Usually their performance claims are about previously available x86 Mac workstations, not x86 in general. You could even go beyond that chart and look at dual socket AMD to get 192 cores, you just won't be able to run a supported macOS on it. Their more traditional "vs the rest of the industry" performance claims on laptops aim towards the single thread performance of M2. While maybe not a chart topper a year later, it's quite a far cry from the 2.5x difference you've found there. In the end though, if all you want is the fastest, loudest, most plugged in laptop you can just go out and buy a laptop chassis which accepts desktop CPUs from the likes of Clevo et al.
alpaca128 3y ago
I don't get paid to run benchmarks, so I couldn't care less what the high scores are. My M2 laptop is 3x as fast as my previous laptop while being completely silent, making it very enjoyable to work with.
- madeofpalk 3y ago
  How do you know it’s 3x faster? At what?
  - phamilton 3y ago
    Not OP, but at compiling code (Rust). Some anecdata: time cargo test # old machine (2018 16" Macbook Pro 2.2 GHz 6-Core i7) cargo test 2077.73s user 221.03s system 596% cpu 6:25.67 total # linux dev machine (AMD Ryzen 5 3550H - 4-Core 2.1Ghz, using mold linker) real 4m48.771s user 25m0.121s system 2m9.922s # new machine (2023 14" Macbook Pro M2 Pro - 10 cores 6perf/4efficiency) cargo test 389.91s user 39.56s system 478% cpu 1:29.66 total

kitsunesoba 3y ago

Good perf per watt means more than just good battery life. It also means the laptop's fans won't be screaming like the machine is being burned alive when being pushed, that the chassis won't get uncomfortably hot easily, and and performance doesn't fall off of a cliff when you're unplugged. In short, it's about balance, which I think is just as important as raw performance for most people.

zamadatix 3y ago

The M2 Ultra will definitely crack the top 20 but it's no surprise a 24 core option is beat by 64 core competition, which tops that chart. Usually their performance claims are about previously available x86 Mac workstations, not x86 in general. You could even go beyond that chart and look at dual socket AMD to get 192 cores, you just won't be able to run a supported macOS on it. Their more traditional "vs the rest of the industry" performance claims on laptops aim towards the single thread performance of M2. While maybe not a chart topper a year later, it's quite a far cry from the 2.5x difference you've found there. In the end though, if all you want is the fastest, loudest, most plugged in laptop you can just go out and buy a laptop chassis which accepts desktop CPUs from the likes of Clevo et al.

alpaca128 3y ago

I don't get paid to run benchmarks, so I couldn't care less what the high scores are. My M2 laptop is 3x as fast as my previous laptop while being completely silent, making it very enjoyable to work with.

madeofpalk 3y ago
How do you know it’s 3x faster? At what?
- phamilton 3y ago
  Not OP, but at compiling code (Rust). Some anecdata: time cargo test # old machine (2018 16" Macbook Pro 2.2 GHz 6-Core i7) cargo test 2077.73s user 221.03s system 596% cpu 6:25.67 total # linux dev machine (AMD Ryzen 5 3550H - 4-Core 2.1Ghz, using mold linker) real 4m48.771s user 25m0.121s system 2m9.922s # new machine (2023 14" Macbook Pro M2 Pro - 10 cores 6perf/4efficiency) cargo test 389.91s user 39.56s system 478% cpu 1:29.66 total