7 ms·

Like a few other comments have mentioned, AMD's Strix Halo / AI Max 380 and above is the chip family that is closest to what Apple has done with the M series. I

Like a few other comments have mentioned, AMD's Strix Halo / AI Max 380 and above is the chip family that is closest to what Apple has done with the M series. It has integrated memory and decent GPU. A few iterations of this should be comparable to the M series (and should make local LLMs very feasible, if that is your jam.)

aurareturn 1y ago
On Cinebench 2025 single threaded, M4 is roughly 4x more efficient and 50% faster than Strix Halo. These numbers can be verified by googling Notebookcheck. How many iterations to match Apple?
- kangs 1y ago
  yes and no. i have macbook pro m4 and a zbook g1a (ai max 395+ ie strix halo) In day to day usage the strix halo is significantly faster, and especially when large context LLM and games are used - but also typical stuff like Lightroom (gpu heavy) etc. on the flip side the m4 battery life is significantly longer (but also the mpb is approx 1/4 heavier) for what its worth i also have a t14 with a snapdragon X elite and while its battery is closer to a mbp, its just kinda slow and clunky. so my best machine right now is the x86 actually!
  - nicolaslem 1y ago
    I also have a Strix Halo zbook G1A and I am quite disappointed in the idle power consumption as it hovers around 8W. Adding to that, it is very picky about which power brick it accepts (not every 140W PD compliant works) and the one that comes with the laptop is bulky and heavy. I am used to plugging my laptop into whatever USB-C PD adapter is around, down to 20W phone chargers. Having the zbook refuse to charge on them is a big downgrade for me.
    - omnimus 1y ago
      I am also on search of good portable brick to replace 140w. I found 100w Anker Prime was working well. And surprisingly there is almost identical 3 port Baseus 100w GaN but half the price. For some reason it is hard to come by (they have few other 100w bricks that are not so portable) i think it might be discontinued.
    - yaro330 1y ago
      > Adding to that, it is very picky about which power brick it accepts (not every 140W PD compliant works) and the one that comes with the laptop is bulky and heavy. I am used to plugging my laptop into whatever USB-C PD adapter is around, down to 20W phone chargers. Having the zbook refuse to charge on them is a big downgrade for It's Dell, they are probably not actually using PD3.1 to achieve the 140w mark, instead they are prolly using PD3.0 extension and shove 20v7a into the laptop. I can't find any info, but you can check on the charger. If it lists 28V then it's 3.1, else 3.0. If it's 3.1 you can get a Baseus PowerMega 140W PD3.1, seems like a reeeeally solid charger from my limited use.
      - nicolaslem 1y ago
        It is HP, and the output of the provided adapter does 28V 5A, so in spec. With some of the other 28V 5A adapters I have, it charges until triggering a compute heavy task and then stops. I have seen reports online of people seeing this behavior with the official adapter. My theory is that the laptop itself does not accept any ripple at all.
        yaro330 1y ago
        Ah my bad. Are you sure your cable can do 140w? That was the source of most of my pains trying to push 100w to my work laptop. Baseus and Anker have some good PD3.1 chipped cables that worked for me. What kind of charger are you using?
  - aurareturn 1y ago
    yes and no. i have macbook pro m4 and a zbook g1a (ai max 395+ ie strix halo) You're comparing the base M4 to a full fat Strix Halo that costs nearly $4,000. You can buy the base M4 chip in a Mac Mini for $500 on sale. A better comparison would be the M4 Max at that price. Here's a comparison I did between Strix Halo, M4 Pro, M4 Max: https://imgur.com/a/yvpEpKF https://imgur.com/a/yvpEpKF As you can see, Strix Halo is behind M4 Pro in performance and severely behind in efficiency. In ST, M4 Pro is 3.6x more efficient and 50% faster. It's not even close to the M4 Max. (but also the mpb is approx 1/4 heavier) Because it uses a metal enclosure.
    - KingOfCoders 1y ago
      Someone has these two machines, and claims the x86 feels faster in his work. You don't own any of the machines but have "made" a comparison by copying data from the internet I assume. This is like explaining to someone who eats a sweet apple that the internet says the apple isn't sweet. MacBook Pro, 2TB, 32gb, 3200 EUR HP G1a, 2TB, 128gb, 3700 EUR If we don't compare laptops but mini-PCs, Evo X2, 2TB, 128gb, 2000 EUR, Mac Mini, 2TB, 32gb, 2200 EUR
      - aurareturn 1y ago
        You don't own any of the machines but have "made" a comparison by copying data from the internet I assume. This is like explaining to someone who eats a sweet apple that the internet says the apple isn't sweet. Yea, I never said he is wrong in his own experience. I was pointing out that the comparison is made between a base M4 and maxed out Ryzen. If we want to compare products in the same class, then use M4 Max. MacBook Pro, 2TB, 32gb, 3200 EUR A little disingenuous to max out on the SSD to make the Apple product look worse. SSD prices are bad value on Apple products. No one is denying that.
        KingOfCoders 1y ago
        I didn't "max out" the SSD, I chose an SSD to match the machine of the user. You: "You're comparing the base M4 to a full fat Strix Halo that costs nearly $4,000." Then You: "A little disingenuous to max out on the SSD to make the Apple product look worse."
- FirmwareBurner 1y ago
  >How many iterations to match Apple? Until AMD can built a tailor made OS for their chips and build their own laptops.
  - aurareturn 1y ago
    Here's an M4 Max running macOS running Parallels running Windows compared to AMD's very best laptop chip: https://browser.geekbench.com/v6/cpu/compare/13494385?baseline=13494130 https://browser.geekbench.com/v6/cpu/compare/13494385?baseli... M4 Max is still faster. Note that the M4 Max is only given 14 out of 16 cores, likely reserving 2 of them for macOS. How do you explain this when Windows has zero Apple Silicon optimizations?
    - FirmwareBurner 1y ago
      Maybe Geek bench is not a good benchmark?
      - aurareturn 1y ago
        Maybe it is? Cinebench favors Apple even more. GB correlates highly with SPEC. AMD also uses GB in their official marketing slides.
      - TiredOfLife 1y ago
        Geekbench is the closest thing to a good benchmark that's usable across generations and architectures.
- noelwelsh 1y ago
  > How many iterations to match Apple? Why are you asking me? I'm not in charge of AMD. Yes the Strix Halo is not as fast on the benchmarks as the M4 Max, its bandwidth is lower, and the max config has less memory. However, it is available in a lot of different configurations and some are much cheaper than comparable M4 systems (e.g. the maxed out Framework desktop is $2000.) It's a tradeoff, as everything in life is. No need to act like such an Apple fanboi.
  - aurareturn 1y ago
    Why are you asking me? I'm not in charge of AMD. Because you claimed this so I thought you knew: A few iterations of this should be comparable to the M series
- KingOfCoders 1y ago
  On one of the few workloads where massive parallelism makes sense, why quote a single threaded number? I'm curious.
  - aurareturn 1y ago
    To show in real numbers why people always say a Macbook always feels miles ahead of AMD and Intel in actual real world experience. The primary reason is the ST speed (snappy feeling) and the efficiency (no noise, cool, long battery life). It just so happens that Cinebench 2025 is the only power measurement metric I have available via Notebookcheck. If Notebookcheck did power measurements for GB6, I'd rather use that as it's a better CPU benchmark overall. Cinebench 2025 is a decent benchmark but not perfect. It does a good enough job of demonstrating why the experience of using Apple Silicon is so much better. If we truly want to measure the CPU architecture like a professional, we would use SPEC and the measure power from the wall.
- alt227 1y ago
  The important part of this is 'Single threaded'. Whereas if you are actually using Cinebench to do real rendering you would always want multi core performance. Which pretty much makes Apples single Core benchmark results pointless.
TiredOfLife 1y ago
> It has integrated memory And has had for many years. Even Apple had that with Apple II
- noelwelsh 1y ago
  What point are you trying to make? Strix Halo was released this year. How is the architecture of the Apple II relevant?
whynotminot 1y ago
> A few iterations of this should be comparable to the M series This assumes Apple's M series performance is a static target. It is not. Apple is iterating too.

aurareturn 1y ago

On Cinebench 2025 single threaded, M4 is roughly 4x more efficient and 50% faster than Strix Halo. These numbers can be verified by googling Notebookcheck. How many iterations to match Apple?

kangs 1y ago
yes and no. i have macbook pro m4 and a zbook g1a (ai max 395+ ie strix halo) In day to day usage the strix halo is significantly faster, and especially when large context LLM and games are used - but also typical stuff like Lightroom (gpu heavy) etc. on the flip side the m4 battery life is significantly longer (but also the mpb is approx 1/4 heavier) for what its worth i also have a t14 with a snapdragon X elite and while its battery is closer to a mbp, its just kinda slow and clunky. so my best machine right now is the x86 actually!
- nicolaslem 1y ago
  I also have a Strix Halo zbook G1A and I am quite disappointed in the idle power consumption as it hovers around 8W. Adding to that, it is very picky about which power brick it accepts (not every 140W PD compliant works) and the one that comes with the laptop is bulky and heavy. I am used to plugging my laptop into whatever USB-C PD adapter is around, down to 20W phone chargers. Having the zbook refuse to charge on them is a big downgrade for me.
  - omnimus 1y ago
    I am also on search of good portable brick to replace 140w. I found 100w Anker Prime was working well. And surprisingly there is almost identical 3 port Baseus 100w GaN but half the price. For some reason it is hard to come by (they have few other 100w bricks that are not so portable) i think it might be discontinued.
  - yaro330 1y ago
    > Adding to that, it is very picky about which power brick it accepts (not every 140W PD compliant works) and the one that comes with the laptop is bulky and heavy. I am used to plugging my laptop into whatever USB-C PD adapter is around, down to 20W phone chargers. Having the zbook refuse to charge on them is a big downgrade for It's Dell, they are probably not actually using PD3.1 to achieve the 140w mark, instead they are prolly using PD3.0 extension and shove 20v7a into the laptop. I can't find any info, but you can check on the charger. If it lists 28V then it's 3.1, else 3.0. If it's 3.1 you can get a Baseus PowerMega 140W PD3.1, seems like a reeeeally solid charger from my limited use.
    - nicolaslem 1y ago
      It is HP, and the output of the provided adapter does 28V 5A, so in spec. With some of the other 28V 5A adapters I have, it charges until triggering a compute heavy task and then stops. I have seen reports online of people seeing this behavior with the official adapter. My theory is that the laptop itself does not accept any ripple at all.
      - yaro330 1y ago
        Ah my bad. Are you sure your cable can do 140w? That was the source of most of my pains trying to push 100w to my work laptop. Baseus and Anker have some good PD3.1 chipped cables that worked for me. What kind of charger are you using?
- aurareturn 1y ago
  yes and no. i have macbook pro m4 and a zbook g1a (ai max 395+ ie strix halo) You're comparing the base M4 to a full fat Strix Halo that costs nearly $4,000. You can buy the base M4 chip in a Mac Mini for $500 on sale. A better comparison would be the M4 Max at that price. Here's a comparison I did between Strix Halo, M4 Pro, M4 Max: https://imgur.com/a/yvpEpKF https://imgur.com/a/yvpEpKF As you can see, Strix Halo is behind M4 Pro in performance and severely behind in efficiency. In ST, M4 Pro is 3.6x more efficient and 50% faster. It's not even close to the M4 Max. (but also the mpb is approx 1/4 heavier) Because it uses a metal enclosure.
  - KingOfCoders 1y ago
    Someone has these two machines, and claims the x86 feels faster in his work. You don't own any of the machines but have "made" a comparison by copying data from the internet I assume. This is like explaining to someone who eats a sweet apple that the internet says the apple isn't sweet. MacBook Pro, 2TB, 32gb, 3200 EUR HP G1a, 2TB, 128gb, 3700 EUR If we don't compare laptops but mini-PCs, Evo X2, 2TB, 128gb, 2000 EUR, Mac Mini, 2TB, 32gb, 2200 EUR
    - aurareturn 1y ago
      You don't own any of the machines but have "made" a comparison by copying data from the internet I assume. This is like explaining to someone who eats a sweet apple that the internet says the apple isn't sweet. Yea, I never said he is wrong in his own experience. I was pointing out that the comparison is made between a base M4 and maxed out Ryzen. If we want to compare products in the same class, then use M4 Max. MacBook Pro, 2TB, 32gb, 3200 EUR A little disingenuous to max out on the SSD to make the Apple product look worse. SSD prices are bad value on Apple products. No one is denying that.
      - KingOfCoders 1y ago
        I didn't "max out" the SSD, I chose an SSD to match the machine of the user. You: "You're comparing the base M4 to a full fat Strix Halo that costs nearly $4,000." Then You: "A little disingenuous to max out on the SSD to make the Apple product look worse."
FirmwareBurner 1y ago
>How many iterations to match Apple? Until AMD can built a tailor made OS for their chips and build their own laptops.
- aurareturn 1y ago
  Here's an M4 Max running macOS running Parallels running Windows compared to AMD's very best laptop chip: https://browser.geekbench.com/v6/cpu/compare/13494385?baseline=13494130 https://browser.geekbench.com/v6/cpu/compare/13494385?baseli... M4 Max is still faster. Note that the M4 Max is only given 14 out of 16 cores, likely reserving 2 of them for macOS. How do you explain this when Windows has zero Apple Silicon optimizations?
  - FirmwareBurner 1y ago
    Maybe Geek bench is not a good benchmark?
    - aurareturn 1y ago
      Maybe it is? Cinebench favors Apple even more. GB correlates highly with SPEC. AMD also uses GB in their official marketing slides.
    - TiredOfLife 1y ago
      Geekbench is the closest thing to a good benchmark that's usable across generations and architectures.
noelwelsh 1y ago
> How many iterations to match Apple? Why are you asking me? I'm not in charge of AMD. Yes the Strix Halo is not as fast on the benchmarks as the M4 Max, its bandwidth is lower, and the max config has less memory. However, it is available in a lot of different configurations and some are much cheaper than comparable M4 systems (e.g. the maxed out Framework desktop is $2000.) It's a tradeoff, as everything in life is. No need to act like such an Apple fanboi.
- aurareturn 1y ago
  Why are you asking me? I'm not in charge of AMD. Because you claimed this so I thought you knew: A few iterations of this should be comparable to the M series
KingOfCoders 1y ago
On one of the few workloads where massive parallelism makes sense, why quote a single threaded number? I'm curious.
- aurareturn 1y ago
  To show in real numbers why people always say a Macbook always feels miles ahead of AMD and Intel in actual real world experience. The primary reason is the ST speed (snappy feeling) and the efficiency (no noise, cool, long battery life). It just so happens that Cinebench 2025 is the only power measurement metric I have available via Notebookcheck. If Notebookcheck did power measurements for GB6, I'd rather use that as it's a better CPU benchmark overall. Cinebench 2025 is a decent benchmark but not perfect. It does a good enough job of demonstrating why the experience of using Apple Silicon is so much better. If we truly want to measure the CPU architecture like a professional, we would use SPEC and the measure power from the wall.
alt227 1y ago
The important part of this is 'Single threaded'. Whereas if you are actually using Cinebench to do real rendering you would always want multi core performance. Which pretty much makes Apples single Core benchmark results pointless.

TiredOfLife 1y ago

> It has integrated memory And has had for many years. Even Apple had that with Apple II

noelwelsh 1y ago
What point are you trying to make? Strix Halo was released this year. How is the architecture of the Apple II relevant?

whynotminot 1y ago

> A few iterations of this should be comparable to the M series This assumes Apple's M series performance is a static target. It is not. Apple is iterating too.