Intel Core i9-12900HK Review: Can Intel's Fastest CPU Beat Apple's M1 Pro?

Status
Not open for further replies.

QuantumPhysics

Posts: 6,308   +7,247
If I were just making content then I'd choose the Apple CPU for high performance and low energy use.

But I create content and play games (which noticeably isn't on this comparison).

As long as gaming is a primary concern, Apple can't get my money.
 

Mister_K

Posts: 2,163   +862
I am no fan of Apple PCs but you have to say, Apples architecture is amazing. If OSX wasn't so awful I would jump ship for my editing rig but in the desktop environment, full size GPUs are still the king.

That said with the best M1 chips + SSD NAS, it's a close battle for non-heavy GPU intensive workloads. Still I am sure the laptops will screech when you throw some noise reduction to 4K/8K clips (not that desktops wouldn't either).
 

Sausagemeat

Posts: 1,597   +1,422
The M1 pro is emulating half of these benchmarks through Rosetta and it’s still comparable. But for me the real telling number is the idle power consumption. It’s a fraction of what the X86 parts are doing and this is going to give you far better battery life as it’s not as if you run your system on battery constantly encoding or rendering etc. Intel and AMD are comparable at efficiency when at full load but that’s it. They get wiped off the floor for general system power consumption. And on top of this M1 equipped parts are far thinner, quieter and lighter than X86 based laptops.

When it comes to windows the AMD stuff appeals to me more than Intels stuff for laptop use. But I don’t use Windows for anything but desktop gaming. For personal computing I use an iPad Pro which has an M1. There’s no tablet around anywhere as fast as this iPad and no laptop anywhere near as portable.
 

Jon Tseng

Posts: 93   +71
These benchmarks are meaningless without SpectreV2 mitigations.
According to Phoronix there are losses of up to 35% with patches on Linux with Alder Spectre-Lake
Are we 100% sure M1 will not require mitigations and take a performance hit?

The coverage I saw suggested the standard ARM implementations were vulnerable to Spectre V2. Is there any reason we can be certain the Apple ISA implementation might not require a fix?
 

sreams

Posts: 288   +411
What’s telling is that AMD is barely mentioned despite just releasing the 6000 series... While I’m sure their shills will post eventually, this shows yet another loss for Ryzen in the high performance mobile market.

The 6000 series isn't an updated architecture (like the 7000 series will be). Nobody expected it to have a big performance improvement.
 

Loadedaxe

Posts: 72   +106
While the Apple silicon is been a huge uplift in technology using ARM, they still are not there unless your locked in the macOS eco system. I have a M1 Mac Mini I use for work because I have to. I also have 10900K for gaming. It would be nice to be able to use bootcamp. Especially on the new Mac Studio, if I could I would buy one, but I wont spend any more than I need to for work.

If only they could get Windows 11 to run outside a 3rd party app so you could game on them, until then they will be reserved for Apple users.

Close Apple, but not close enough.
 

Squid Surprise

Posts: 5,335   +4,980
The 6000 series isn't an updated architecture (like the 7000 series will be). Nobody expected it to have a big performance improvement.
Because no one expected they would have to… the 5000 series smoked the 11000 series and I suspect that AMD assumed the same would happen with Alder Lake… but they were wrong…
 

Mr Majestyk

Posts: 1,222   +1,119
Honestly looking at this, the 12900HK is pointless, just for bragging rights and IMO the 6900HS is the most balanced by far. Beat the M1 in most benchmarks, gets fantastic battery life, the iGPU trashes the Iris Xe and in the normal 45W range is so close whop cares. I'd only get intel for code compiling. Roll on Phoenix and Arrow Lake.
 

DonquixoteIII

Posts: 110   +65
The 12900 used in this test (?) is a 'k' variant, which means it is unlocked. Of course you knew that, as did all readers of this site, right? K processors are never worth their price UNLESS you OC them. But when you do, then they can be worth their weight in flops. OC the 12 9, THEN compare the 12 7....

Throwing the M1 PRO into this mix is hardly "apples to apples" There is only ONE apple chip in this whole roundup, so it is truly in a class of its own. OK, tongue out of cheek. Compare all of the above chips in render times on Photoshop. Or any of a number of platforms written FOR OSX. You really cannot compare Apples and x86's with the same benchies.
 

GregonMaui

Posts: 330   +120
Gotta call some BS here. 1) cinebench is not optimized specifically for Apple silicon using the best toolset available, so it's impressive that M1 does as well as it does, but otherwise meaningless. 2) handbrake likewise, the video encoders/decoders were obviously not employed, yet they give the best results in quality as well as in performance, 3) Blender is not using the correct version for apple silicon, the version truly optimized for Apple silicon using Metal is 3.1. That is like saying a certain CPU performs better on OpenCL which has been deprecated for years on Apple, so maybe it does, but meh.

Let's have a truthful "benchmark" comparison actually comparing times to produce best results. That takes nothing away from alder Lake, they are impressive no doubt,
 

HardReset

Posts: 1,665   +1,322
Because no one expected they would have to… the 5000 series smoked the 11000 series and I suspect that AMD assumed the same would happen with Alder Lake… but they were wrong…
AMD CPU's smoke Intel ones on laptops. Like article states, this chip is pretty useless when running battery. And if laptop is plugged on most times anyway, Ryzen 5950X is OK and smokes Intel's "mobile" CPU's. When running battery, AMD 6000 series is much better. Overall, Alder Lake mobiles are useless. Battery life is very poor and for desktop replacement, AMD's non-mobile CPU's are faster.
 

Squid Surprise

Posts: 5,335   +4,980
AMD CPU's smoke Intel ones on laptops. Like article states, this chip is pretty useless when running battery. And if laptop is plugged on most times anyway, Ryzen 5950X is OK and smokes Intel's "mobile" CPU's. When running battery, AMD 6000 series is much better. Overall, Alder Lake mobiles are useless. Battery life is very poor and for desktop replacement, AMD's non-mobile CPU's are faster.
When plugged in, the Ryzen gets smoked... and even at similar power levels, the Alder Lake tends to win... they only win on power efficiency...
 

HardReset

Posts: 1,665   +1,322
When plugged in, the Ryzen gets smoked... and even at similar power levels, the Alder Lake tends to win... they only win on power efficiency...
Ryzen Desktops won't get smoked. Similarily Alder Lake desktops are cheaper And faster than mobile versions.

What's the point using mobile CPU if laptop is plugged in all the time? And why both AMD and Intel waste silicon for huge iGPU when most laptops have external GPU? Laptop market is so stupid...
 

waclark

Posts: 352   +240
If I were just making content then I'd choose the Apple CPU for high performance and low energy use.

But I create content and play games (which noticeably isn't on this comparison).

As long as gaming is a primary concern, Apple can't get my money.

I tend to agree, though I do have an Apple MacBook Air. It's a great travel companion over my ASUS ROG Strix 17 beast. Apple does have some games that I enjoy playing, but isn't as deep in the online, competitive apps like Overwatch or any of the recent MMO games coming out for PC. Streaming game services may change that, so we'll see.
 

waclark

Posts: 352   +240
Ryzen Desktops won't get smoked. Similarily Alder Lake desktops are cheaper And faster than mobile versions.

What's the point using mobile CPU if laptop is plugged in all the time? And why both AMD and Intel waste silicon for huge iGPU when most laptops have external GPU? Laptop market is so stupid...

The point to a mobile device is that I can take it with me. No, I don't sit around airports working on my laptop and I would rarely game on battery for the simple reason that I like to have a mouse and full sized keyboard when gaming. Therefore, I am likely to be sitting at a desk near a power outlet.

The reason to include an onboard GPU is that when I don't need that external GPU, I can save battery power running the on-board GPU. Not all laptops have external GPU and for people who don't need them the on-board GPU is plenty sufficient. Plus I don't want that external GPU creating heat when I'm sitting at my desk reading Techspot or other web surfing activities, not to mention the fan noise.
 

Sausagemeat

Posts: 1,597   +1,422
Gotta call some BS here. 1) cinebench is not optimized specifically for Apple silicon using the best toolset available, so it's impressive that M1 does as well as it does, but otherwise meaningless. 2) handbrake likewise, the video encoders/decoders were obviously not employed, yet they give the best results in quality as well as in performance, 3) Blender is not using the correct version for apple silicon, the version truly optimized for Apple silicon using Metal is 3.1. That is like saying a certain CPU performs better on OpenCL which has been deprecated for years on Apple, so maybe it does, but meh.

Let's have a truthful "benchmark" comparison actually comparing times to produce best results. That takes nothing away from alder Lake, they are impressive no doubt,
Yeah I agree, this test was extremely biased towards Intel. But there aren’t many tools for comparison. In most use cases you would use these CPUs with different software. In this case they’ve just put the M1 through their windows test suite as best as they can but a lot of it is effectively being emulated on the M1 side.

Also these claims Intel and AMD have about power efficiency matching the M1 is only under load and in specific scenarios. The overall power efficiency is much greater on an M1. I’d like to see a battery life test between a MacBook and a Windows notebook with a similar battery size to find out the real world difference.
 

HardReset

Posts: 1,665   +1,322
The point to a mobile device is that I can take it with me. No, I don't sit around airports working on my laptop and I would rarely game on battery for the simple reason that I like to have a mouse and full sized keyboard when gaming. Therefore, I am likely to be sitting at a desk near a power outlet.
OK and because you are using it as desktop replacement, it makes no sense to use mobile parts instead desktop ones.
The reason to include an onboard GPU is that when I don't need that external GPU, I can save battery power running the on-board GPU. Not all laptops have external GPU and for people who don't need them the on-board GPU is plenty sufficient. Plus I don't want that external GPU creating heat when I'm sitting at my desk reading Techspot or other web surfing activities, not to mention the fan noise.
I said huge iGPU. 12900K (desktop model) has iGPU too. It's much slower and takes less die space. Basically 12900K has more CPU cores instead larger GPU. For casual use, 12900K has good enough GPU and for gaming 12900HK GPU is too slow. 12900HK simply doesn't make sense at all. 12900K is cheaper and faster on everything expect when using integrated GPU that won't be used on heavier loads. 12900HK has better battery life but who cares if battery performance is poor or very poor?
 

Squid Surprise

Posts: 5,335   +4,980
OK and because you are using it as desktop replacement, it makes no sense to use mobile parts instead desktop ones.

I said huge iGPU. 12900K (desktop model) has iGPU too. It's much slower and takes less die space. Basically 12900K has more CPU cores instead larger GPU. For casual use, 12900K has good enough GPU and for gaming 12900HK GPU is too slow. 12900HK simply doesn't make sense at all. 12900K is cheaper and faster on everything expect when using integrated GPU that won't be used on heavier loads. 12900HK has better battery life but who cares if battery performance is poor or very poor?
I understand you feel you must defend AMD on every thread... but you might just want to give this up...
 
Status
Not open for further replies.