Apple M4 Max chip outperforms Core Ultra 9 285K and Ryzen 9 9950X in Geekbench 6

[DragonSlayer101]

Something to look forward to: Apple just launched new MacBook Pro laptops powered by the M4 Pro and M4 Max chips, and initial reports suggest that the new SoCs not only beat their predecessors by a healthy margin, but are also faster than flagship desktop-grade CPUs from Intel and AMD.

Apple released a spate of new Mac models powered by the M4 chips last month. They include the upgraded iMac, the redesigned Mac Mini, and the all-new MacBook Pro models with Thunderbolt 5. While the M4 Pro features a 14-core CPU and 20-core GPU, the M4 Max comes with a 16-core CPU and a 40-core GPU.

The company had earlier launched the 7th-generation iPad Pro with the vanilla M4 chip that has a 10-core CPU and 10-core GPU.

In Geekbench 6, the M4 Pro scored 3,925 points in the single-core benchmark, which is higher than the 3,450 points notched up by the Intel Core Ultra 9 285K and 3,359 points managed by the AMD Ryzen 9 9950X. This is even more impressive when you consider that the M4 Pro is clocked at 4.5GHz, while the Intel and AMD chips have boost clocks of up to 5.7GHz.

In the multi-core benchmark, the M4 Pro scored 22,669 points, which is slightly lower than the 23,024 points racked up by the 285K. However, it is still significantly higher than the 20,183 points scored by the 9950X. It is worth noting here that the M4 Pro has only 14 cores, while the Core Ultra 9 285K features 24 cores. The Ryzen 9 9950X has 16 cores and 32 threads.

Although the M4 Pro's scores are impressive, the M4 Max is even faster. The chip notched up 4,060 points in the single-core test and a whopping 26,675 in the multi-core benchmark, both of which are significantly higher than the scores put up by the 285K and 9950X.

While preliminary scores for the M4 Pro and M4 Max look impressive, it is worth remembering that Geekbench 6 is only a synthetic benchmark and does not necessarily reveal a processor's true performance potential. We will likely have to see how the M4 twins perform in applications like Cinebench or Handbrake to gauge their real-world performance.

Permalink to story:

Apple M4 Max chip outperforms Core Ultra 9 285K and Ryzen 9 9950X in Geekbench 6

 

[DukeJukem]

None of these benchmarks matter unless the Mac has been forced to run for long enough to force itself to throttle into Oblivion due to it's terrible cooling. Prove your cooling can keep up because it never has or these benchmarks aren't real.

 

[Rudyduu]

Geekbench scores don't represent any real world software used to do anything, please stop using them to prove anything. Cinebench and handbrake don't measure a CPU's relative performance in every program only those specific ones.

Beating the newest CPU from Intel is something everyone is doing including Intel's previous generation of CPU's, it's not an achievement.

 

[opckieran]

And the starting price for any device equipped with the M4 Max is currently $3199...

 

[bviktor]

Oh I love it. When the brand that you don't like performs great in benchmarks, then the benchmark is always flawed

It is quite well known and established that the Apple Silicons are actually GREAT, they've been around for like what, 5 years now? Yet people still have doubts about it. I don't understand.

And I'm saying this as a person that truly HATES Macbooks because macOS and the keyboard layout are driving me nuts within minutes, even after 10 years of trying to get to like them.

But performance and battery life leave everything else in the dust, period. I don't know why other manufacturers can't make comparable laptops, but they just can't.

 

[Squid Surprise]

But performance and battery life leave everything else in the dust, period. I don't know why other manufacturers can't make comparable laptops, but they just can't.

Because Apple has the money and influence to leverage the best node processing from TSMC whereas other companies can't...

Apple, at least for the foreseeable future, will always have a huge advantage due to having the best node.

I just wish they cared about gamers...

 

[HardReset]

It is quite well known and established that the Apple Silicons are actually GREAT, they've been around for like what, 5 years now? Yet people still have doubts about it. I don't understand

How about memory upgrade? Oh, Apple chips doesn't support that

 

[Squid Surprise]

How about memory upgrade? Oh, Apple chips doesn't support that

The chips support it... APPLE doesn't... big difference

 

[TheBigT42]

The problem is they are only good for running MAC OS.

 

[Lew Zealand]

Because Apple has the money and influence to leverage the best node processing from TSMC whereas other companies can't...

Apple, at least for the foreseeable future, will always have a huge advantage due to having the best node.

Intel's and AMD's current CPUs all use a 3nm TSMC mode like Apple so this tired excuse is dead now.

 

[HardReset]

The chips support it... APPLE doesn't... big difference

My point is that integrated memory gives more speed but you cannot expand it. For obvious reasons.

Intel's and AMD's current CPUs all use a 3nm TSMC mode like Apple so this tired excuse is dead now.

For AMD, only some server CPUs.

 

[WhoCaresAnymore]

If AMD and INTEL could charge what Apple charges, we would have monster CPU’s. Apple knows its base.

 

[Liranan]

Geekbench on Apple silicon is only good for benching that version of Geekbench on Apple silicon.

 

[Mr Majestyk]

Intel's and AMD's current CPUs all use a 3nm TSMC mode like Apple so this tired excuse is dead now.

Ah no, AMD's desktop Zen 5 are all on N4P, only Epyc Turin uses N3E.

 

[redgarl]

Geekbench... phhh...

Why not Userbench at this point...

Once again, it will fall short in real compute benchmarks like cinebench, PovRay or blender...

 

[redgarl]

Oh I love it. When the brand that you don't like performs great in benchmarks, then the benchmark is always flawed

It is quite well known and established that the Apple Silicons are actually GREAT, they've been around for like what, 5 years now? Yet people still have doubts about it. I don't understand.

And I'm saying this as a person that truly HATES Macbooks because macOS and the keyboard layout are driving me nuts within minutes, even after 10 years of trying to get to like them.

But performance and battery life leave everything else in the dust, period. I don't know why other manufacturers can't make comparable laptops, but they just can't.

CPU-Z, Geekbench and Userbench scores doesn`t mean anything. Every time that we reference these pre-release scores, they end up delivering underwhelming gains.

Remember SnapDragon X Elite? Yeah, what an underwhelming joke.

 

[Squid Surprise]

CPU-Z, Geekbench and Userbench scores doesn`t mean anything. Every time that we reference these pre-release scores, they end up delivering underwhelming gains.

Remember SnapDragon X Elite? Yeah, what an underwhelming joke.

Remember when the M3 scored great on geekbench - and it WAS great?

Remember when the M2 scored great in geekbench… and it was also great?

Remember when the M1 scored great in geekbench - and… wait for it… it was also great?

 

[airbornex]

I’m curious to see how this translates to real-world applications. Geekbench shows promising results, but the real test will be how the M4 Pro and Max handle sustained workloads in creative applications like video rendering, where thermal management and throttling have historically been challenges.

 

[Xenocea]

Being faster in Geekbench doesn't mean anything. In real world performance, it gets trounced by those processors from both AMD & Intel.

 

[toooooot]

I’m curious to see how this translates to real-world applications. Geekbench shows promising results, but the real test will be how the M4 Pro and Max handle sustained workloads in creative applications like video rendering, where thermal management and throttling have historically been challenges.

I just wanna see top 10 app total score compared to Intel or AMD. I watched how Apple's chips do in games with their latest AAA ports. They were still noticeably slower.
Could this latest CPU close the gap? If so, Apple could become even more attractive with its computers. And phones, there are already too many.

 

[Rudyduu]

Oh I love it. When the brand that you don't like performs great in benchmarks, then the benchmark is always flawed

It is quite well known and established that the Apple Silicons are actually GREAT, they've been around for like what, 5 years now? Yet people still have doubts about it. I don't understand.

And I'm saying this as a person that truly HATES Macbooks because macOS and the keyboard layout are driving me nuts within minutes, even after 10 years of trying to get to like them.

But performance and battery life leave everything else in the dust, period. I don't know why other manufacturers can't make comparable laptops, but they just can't.

Read the posts again and tell me where anyone said the M4 Max wasn't a good SoC. My phone, my music workstation, my tablets all Apple. The issue is just like we said Geekbench scores don't line up with the performance of any real world software and any performance gain in Geekbench does not match any performance gain in actual use. It's a terrible benchmark and is meaningless.

There are no doubts about it, the main issue is the prices especially for upgrading to decent specs for what the computer is. The Mac Mini should have more storage especially at a time when storage is very cheap.

The reason others can't make comparable laptops is because everyone else is using an open ecosystem where software and hardware is designed by many companies, and all has to work together.

 

[dipique]

None of these benchmarks matter unless the Mac has been forced to run for long enough to force itself to throttle into Oblivion due to it's terrible cooling. Prove your cooling can keep up because it never has or these benchmarks aren't real.

You can expect significant throttling on fanless systems (e.g. MacBook air). Devices like the iMac and Mac mini have some active cooling but will still throttle. The MacBook Pro and Mac studio have beefier cooling and effectively minimize throttling.

This is a good match for real world use. Most users do not max out their CPU for sustained periods of time. Those who do would be reasonably expected to purchase a system with better thermal dissipation.

 

[dipique]

Geekbench scores don't represent any real world software used to do anything, please stop using them to prove anything. Cinebench and handbrake don't measure a CPU's relative performance in every program only those specific ones.

Beating the newest CPU from Intel is something everyone is doing including Intel's previous generation of CPU's, it's not an achievement.

The M4 is the leading CPU in this benchmark, bar none. You can decide for yourself how important or representative of your use case that is.

 

[HardReset]

The M4 is the leading CPU in this benchmark, bar none. You can decide for yourself how important or representative of your use case that is.

Against selected CPUs. How about Threadripper? Apple loses by wide margin. Then Apple fanboys would cry because comparison is "unfair". But it already is, Apple SOC vs AMD/Intel CPU.

 

[Lew Zealand]

You can expect significant throttling on fanless systems (e.g. MacBook air). Devices like the iMac and Mac mini have some active cooling but will still throttle. The MacBook Pro and Mac studio have beefier cooling and effectively minimize throttling.

Got any links to the iMac or Mini thermal throttling? Apple loves to run their machines hot but silent, not from a shortage of cooling but instead from not turning up the fan speed. With normal fan speeds (Macs Fan Control is great for this) the Mini and especially the iMac shouldn't throttle.

The Air should be used for office use and content consumption only.