3 ms·

I see a lot of reviews that say things like this but seem to be written by people who aren't testing against commonly available mini PCs that are built on effic

by dangus 2y ago

I see a lot of reviews that say things like this but seem to be written by people who aren't testing against commonly available mini PCs that are built on efficiency architectures. How different is the efficiency of this compared to something like an Intel N100/200/300 or a Ryzen 7 7735HS that you can get in cheap mini PCs from manufacturers like Beelink? I am not doubting that Apple's processors are class-leading but at the same time it seems like I see a lot of people impressed that a mini PC can idle under 10 watts. That's been common for a long time now.

geerlingguy 2y ago
I have an N100 in that listing on the linked GitHub project, it's the best Intel system I've tested, and it gets around 2.5 Gflops/W (which is a little less than a Raspberry Pi 5, which is not known for being the most efficient Arm system).
lispm 2y ago
https://nanoreview.net/en/cpu-compare/apple-m4-vs-amd-ryzen-7-7735hs https://nanoreview.net/en/cpu-compare/apple-m4-vs-amd-ryzen-...
- dangus 2y ago
  These auto-generated comparison pages are not useful.
  - lispm 2y ago
    why not? Having the specs and benchmark results for both CPUs in a table gives a good overview. What advantage would it bring to manually write such comparison tables?
    - dangus 2y ago
      Specifically, I don’t like them because they show really vague percentages or “scores” where it’s very unclear what specific benchmarks are being used. The benchmark bars for things like single and multi-core performance don’t break down different applications and use cases or even say which benchmarks they are referring to. This link in particular never goes deeper than Geekbench and has a bunch of comparisons to stuff that shouldn’t guide purchase decisions like: Newer - released 1 year and 4 months later More modern manufacturing process – 3 versus 6 nanometers I much prefer the more in-depth tests that outlets like GamersNexus do, where CPUs are tested using multiple workflows and solutions. For example, you can test specific games and applications, test file compression and video encoding, test simulation speed on CPU-dependent simulation games, test clock speed behavior and throttling (does the CPU have jittery boost clocks or does it settle in to a consistent frequency).
      - lispm 2y ago
        > Specifically, I don’t like them because they show really vague percentages or “scores” where it’s very unclear what specific benchmarks are being used. It think it's a good overview over various technical data and benchmark results. Benchmarks are linked. > that shouldn’t guide purchase decisions like ... More modern manufacturing process – 3 versus 6 nanometers Actually that's influencing buying decisions and gives an idea what kind of efficiency a CPU might be able to deliver. I have various M processor machines from Apple and a CPU on the recent 3 nanometer TSMC process has advantages over a 6 nm process. For example the newer M CPUs are more power efficient & smaller and some of that enables the M4 series of chips to be ahead of the rest in terms of efficiency. Over time other CPUs will move to such 3nm processes. I've seen that with the M4 in the iPad Pro and now with the M4 Pro in the new Macs. > I much prefer the more in-depth tests That serves a very different purpose. I like both in-depth tests and tabular comparisons between CPUs.
rstuart4133 2y ago
All his tests are published here: https://github.com/geerlingguy/top500-benchmark?tab=readme-ov-file https://github.com/geerlingguy/top500-benchmark?tab=readme-o... He does test some N100 machines. TL;DR: The M4 blows everything away, but isn't a general purpose machine. The N100 is slightly better than 50% of the power efficiency of the remainder. It's also slightly faster despite having less cores. Single core speed is looks to be twice the speed of ARM. An N100 box with 16GB of RAM and 1TB of NVME is around the USD$200 mark, which is far cheaper than the Mac or Ampere, but in line with the other ARM options. It comes in more form factors with more customisability options than you can poke a pointed stick at. All in all, it doesn't fair too badly. The low price, fast single core speed, and compatibility with everything makes up for a lot of sins.
- rty32 2y ago
  A few x86 CPUs used are 4-7 generations old. Not sure what to make of that.
- rstuart4133 2y ago
  I'm pondering the down votes. My current theory is they are over this: > The M4 blows everything away, but isn't a general purpose machine. Perhaps that upset Apple fans. My definition of "general purpose machine" is one that I can realistically write software for. I written operating systems (with a custom network and GUI stack) in the past. Doing that requires hardware that isn't locked down, and somewhat less obviously is well documented. As in "Intel Architecture Manual" style documented, or failing that at least open source drivers like the Linux i915 driver. Apple doesn't come close to meeting that criteria. That's a shame, because their consumer hardware is stellar. If it was open, I would chose it over anything else.
  - dangus 2y ago
    I’m sorry but in what universe is writing operating systems “general purpose?” That’s about the most niche computing task I can think of.
    - 2y ago

by dangus 2y ago

geerlingguy 2y ago

I have an N100 in that listing on the linked GitHub project, it's the best Intel system I've tested, and it gets around 2.5 Gflops/W (which is a little less than a Raspberry Pi 5, which is not known for being the most efficient Arm system).

lispm 2y ago

https://nanoreview.net/en/cpu-compare/apple-m4-vs-amd-ryzen-7-7735hs https://nanoreview.net/en/cpu-compare/apple-m4-vs-amd-ryzen-...

dangus 2y ago
These auto-generated comparison pages are not useful.
- lispm 2y ago
  why not? Having the specs and benchmark results for both CPUs in a table gives a good overview. What advantage would it bring to manually write such comparison tables?
  - dangus 2y ago
    Specifically, I don’t like them because they show really vague percentages or “scores” where it’s very unclear what specific benchmarks are being used. The benchmark bars for things like single and multi-core performance don’t break down different applications and use cases or even say which benchmarks they are referring to. This link in particular never goes deeper than Geekbench and has a bunch of comparisons to stuff that shouldn’t guide purchase decisions like: Newer - released 1 year and 4 months later More modern manufacturing process – 3 versus 6 nanometers I much prefer the more in-depth tests that outlets like GamersNexus do, where CPUs are tested using multiple workflows and solutions. For example, you can test specific games and applications, test file compression and video encoding, test simulation speed on CPU-dependent simulation games, test clock speed behavior and throttling (does the CPU have jittery boost clocks or does it settle in to a consistent frequency).
    - lispm 2y ago
      > Specifically, I don’t like them because they show really vague percentages or “scores” where it’s very unclear what specific benchmarks are being used. It think it's a good overview over various technical data and benchmark results. Benchmarks are linked. > that shouldn’t guide purchase decisions like ... More modern manufacturing process – 3 versus 6 nanometers Actually that's influencing buying decisions and gives an idea what kind of efficiency a CPU might be able to deliver. I have various M processor machines from Apple and a CPU on the recent 3 nanometer TSMC process has advantages over a 6 nm process. For example the newer M CPUs are more power efficient & smaller and some of that enables the M4 series of chips to be ahead of the rest in terms of efficiency. Over time other CPUs will move to such 3nm processes. I've seen that with the M4 in the iPad Pro and now with the M4 Pro in the new Macs. > I much prefer the more in-depth tests That serves a very different purpose. I like both in-depth tests and tabular comparisons between CPUs.

rstuart4133 2y ago

All his tests are published here: https://github.com/geerlingguy/top500-benchmark?tab=readme-ov-file https://github.com/geerlingguy/top500-benchmark?tab=readme-o... He does test some N100 machines. TL;DR: The M4 blows everything away, but isn't a general purpose machine. The N100 is slightly better than 50% of the power efficiency of the remainder. It's also slightly faster despite having less cores. Single core speed is looks to be twice the speed of ARM. An N100 box with 16GB of RAM and 1TB of NVME is around the USD$200 mark, which is far cheaper than the Mac or Ampere, but in line with the other ARM options. It comes in more form factors with more customisability options than you can poke a pointed stick at. All in all, it doesn't fair too badly. The low price, fast single core speed, and compatibility with everything makes up for a lot of sins.

rty32 2y ago
A few x86 CPUs used are 4-7 generations old. Not sure what to make of that.
rstuart4133 2y ago
I'm pondering the down votes. My current theory is they are over this: > The M4 blows everything away, but isn't a general purpose machine. Perhaps that upset Apple fans. My definition of "general purpose machine" is one that I can realistically write software for. I written operating systems (with a custom network and GUI stack) in the past. Doing that requires hardware that isn't locked down, and somewhat less obviously is well documented. As in "Intel Architecture Manual" style documented, or failing that at least open source drivers like the Linux i915 driver. Apple doesn't come close to meeting that criteria. That's a shame, because their consumer hardware is stellar. If it was open, I would chose it over anything else.
- dangus 2y ago
  I’m sorry but in what universe is writing operating systems “general purpose?” That’s about the most niche computing task I can think of.
  - 2y ago