The M1 scores that using 15.5Ws. So the AMD still takes close to 2x the power. Slowing it down more to score 6% less will not suddenly halve the power on the 5800H.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
M1 chip fast my ass...
- Thread starter Kung gu
- Start date
- Sort by reaction score
BS. I have logs from my MBA M1 pulling ~20W at full CPU load in text terminal without iGPU.
Found it. From powermetrics.
CPU Power: 19647 mW
Last edited:
That’s BS.
I’ll just point you to an unbiased source with powermetrics outputs. It matches what I see on my M1 MBP
CPU Power: 14962 mW
Ehhh … from what I can tell reading about it from the outside a lot of developer hesitancy around the Mac isn’t writing the engine in Metal vs DX vs Vulkan - heck they can use compatibility layers if they want - or the lack of Macs with beefy GPUs (even a lot of AAA games are built to scale across different chipsets). It’s closer to @xWhiplash ‘s experience that testing and support are too expensive or it’s ancillary requirements (not technological) with Apple and the dev process that just pissed them off or it’s an ingrained cultural mindset of ignoring the Mac (which to be fair Apple had a big part to play in - see second or in this list). Unfortunately this last point especially will lag any advances in the GPU, market share, and API. (Unless Apple makes a concerted push to court game developers back to Apple.)
He knows. Many in this forum have had this discussion with him many, many times. He doesn’t care. He’ll post something else and then eventually circle back to this exact argument again with the slightest of variations. This at least the third or fourth time I’ve seen him make these types of claims despite that pretty much everyone has already told him exactly what you posted with the exact same screen shots from Andrei to support it.
Last edited:
Got it.. Time to disengage from pointless discussions with that person.
Agreed. Minimum specs for Back 4 Blood (just recently released) is a NVIDIA GeForce GTX 1050 Ti OR AMD Radeon RX 570. I think a mac with a 5700 XT beats this no? And in fact the M1 itself competes with the 1050 Ti. Recommended specs are NVIDIA GeForce GTX 970 OR AMD Radeon RX 590. Again, I think a 5700 XT beats this no? Certainly beats my old GTX 1080.
Why don't you run the tests for yourself and understand the behavior instead of copying and pasting?
Peak is ~20W for CPU only but when it throttles it's lower but along with that is lower performance. Here's a 34% drop in performance from throttling.
He did. And he used M1 MBP which does not suffer from throttling.(At least not for Cinebench)
If a $2500 dollar laptop beats my m1 MacBook Air that's well built, great keyboard, awesome trackpad, awesome battery life, nice screen speakers, and great built quality, all for $850 than that's fine with me. Oh and my portable power bank can fully charge my laptop BAM!!!
Good numbers
Doubt the next Gen of Apple Silicon will play such games as they wont be out on the platform or remotely optimised. I like gaming myself, equally they don't pay the bills, so not a big consideration for the work machines. My point is that you can have this performance and battery longevity with the M1 in a thin & light 13" chassis for productivity. The M1 can do for me what this Asus does which on the go makes for a far better proposition.Next Windows notebook will be AMD based as Intel remain to be far too power hungry, hot and too many security issues...
Q-6
I don't generally worry on cost as my notebooks pay for themselves, however the M1 Air for $850 has to be one of the best if not the very best bargains in the portable computing market. Overall I think the pricing of the M1 Mac's is decent. My only complaint with my own M1 MBP is only having two ports other than that it's the best Mac I've owned.
Interested to see the next generation, equally I wont consider one until 2nd or 3rd Gen as Apple has been known to get things wrong from time to time
Although I strongly suspect after the lacklustre 2016 MBP redesign & Butterfly Keyboard debacle the message from the very top of Apple was to not screw up the launch of Apple Silicon and not repeat the same embarrassment to the company...Q-6
I was looking for a gaming laptop for work and play and let me tell you for 1500 to 2000 price range the track pad and keyboard were not on par with my m1 Air. Hell I prefer my M1 air over my 16 in mbp. Lighter and more comfortable moving around.I don't generally worry on cost as my notebooks pay for themselves, however the M1 Air for $850 has to be one of the best if not the very best bargains in the portable computing market. Overall I think the pricing of the M1 Mac's is decent. My only complaint with my own M1 MBP is only having two ports other than that it's the best Mac I've owned.
Interested to see the next generation, equally I wont consider one until 2nd or 3rd Gen as Apple has been known to get things wrong from time to time
Q-6
Eeryody else is getting 15W on R23 multicore (Andrei from Anandtech, other posters on this forum, myself). You are the only one who claims a different number, so you have to excuse us if we are skeptical.
Your Zen3 results shows that Cezanne is around 10% faster while consuming 100% more power which makes perfect sense given what we know about these CPUs. Interestingly enough, your constrained Cezanne score is actually lower than a 15W 5800U, which again confirms that the 15W AMD chips do not run on 15W at all.
Overall, Cezanne laptops are great if you want strong sustained multicore performance while still retaining decent mobility. Intel does not have an answer in this department yet (but Alder Lake is coming). M1 does not focus on sustained multicore performance — it's a product designed for low power consumption in the first place, so the fact that it can almost match a sustain specialist while consuming significantly less power (and having 1/2 of computational resources to begin with) again speaks volumes about Apple's superiority when it comes to CPU design.
Mobility was a big factor as is battery life for me. Once on an engineering project, things can get hectic fast and from time to time I need the muscle the likes of this 17" Asus provides. With the 13" MBP I have all that in a single package. Initially I was sceptical myself, however once I saw the raw potential first hand and the battery runtime validated the M1 MBP was an instant purchase.
I've always run portable Mac's in one capacity or another. Once squeezing just over 10 hours of runtime on a 2014 13" MBP, in one click it literally paid for itself. Had it been a Windows notebook likely wouldn't have been the same ending...
Q-6
Your Cezanne iGPU can offer playable FPS in games like Cyberpunk 2077? That's news to me...
You are really a funny one. You clam to have logs, yet you don't show them. You claim drop in performance yet you post a screenshot of a running benchmark with no power readings or scores. How do you expect to be taken seriously? The only hard data you have provided so far is that an 8-core Cezanne at 30W is barely faster than an 4-core M1 at 15W... (according to the results of a well-regarded hardware review publication)
I would assume he/she has a dGPU in the system, hence the comment nor being remotely equal.
Q-6
Oh, I know, I was just being silly. It's just with mi7chy is't all about moving goalposts. They are mixing facts and fiction freely, while constantly changing the context. Since this make it very difficult for me to take them seriously, you will have to excuse an occasional joke or two
I can somehow understand his point. M1 does have lower raw performance comparing to everything he mentioned, and when raw performance is the only thing matters, he does have a point. He was just using some bad examples to describe his point. For someone who does not care about lower power usage, a lower power chip is an under-powered chip, and the improved power efficiency is an unacceptable trade off by sacrificing raw performance.Oh, I know, I was just being silly. It's just with mi7chy is't all about moving goalposts. They are mixing facts and fiction freely, while constantly changing the context. Since this make it very difficult for me to take them seriously, you will have to excuse an occasional joke or two
That’s @bobcomer ’s point in this thread, which as you say is reasonable from a certain POV.
With mi7chy, his central thesis rarely extends beyond “Apple sux” and he moves the goalpost wherever they must be to make that claim and as @leman says mixes fact and fiction freely. That’s why the examples he provided were so bad, it’s why they’ve always been bad every time he does this. One of the previous times he tried claiming the M1 was terrible he ran a benchmark through an x86 VM in WINE while saying this is simply the performance of the chip. That’s right: he ran it through two levels of indirection. When called out on this, he next try to claim that he did so because the only M1 performance metrics anyone should care about is running the processor on such nested emulation because there aren’t any native programs for the M1 because it’s terrible. This, despite the fact that the very benchmark he was running was based on an application which already had a native M1 version. *That’s* mi7chy.
Last edited:
If looking at sheer raw performance you'd be looking to compute in the Cloud, a potent desktop or at very least a desktop replacement such as a portable workstation or a fast gaming notebook which also make for good workstation's as long as you don't require Xeon CPU's & ECC RAM, non of which present a good portable scenario.
The M1 notebooks are entry level systems that punch well above their weight computationally, run at full speed on battery and can push up to 20 hours runtime (M1 MBP). Compare them directly to their Windows counterparts and they are simply far more performant. Suggesting that the M1 Mac's are slow is just an argument for the sake of. I could rattle on all day how this Asus is far more potent, equally it has multiple drives, 32GB RAM and a dedicated dGPU and weighs in at around 3Kg with a corresponding battery runtime...
The only real pertinent question is can Apple Silicon run the software for your needs, if not then an Intel Mac or Intel/AMD PC is the solution.
Q-6
Of course this point is valid, nobody has ever contested that! Zen3 chips have higher multicore sustained performance in many workloads, this is obvious. M1 and Cezanne are different products designed for different niches (the second one is a sustained performance specialist, the first one is a all-rounder with the focus on low power consumption). What is problematic however is his continued attempts to portray M1 as some sort of inferior technology that is not usable outside of casual internet and email use.
I think the heart of these conflicts is that there is a product and that there is a technology. We can all agree that Cezanne is a better product for some users. No question. However, I would maintain that M1 is overall a superior technology. Its performance and flexibility is unprecedented at this power consumption levels, it’s cache sizes and out of order execution capabilities dwarf that of similar chips and its compartmentalized, security oriented design is a novel, very advanced take on addressing modern security challenges. And I don’t see why I should withhold my comments if someone claims that this advanced technology - which IMO is the most exiting thing to happen to mainstream processor design since the Intel Core architecture - is „crap“.
All this is moot anyway. Come Monday we will see what Apple intends to do in the professional laptop market and I for one will be very surprised if their new mobile chips are any slower than x86 enthusiast level desktop stuff. If already a tiny M1 can challenge performance specialists in many workflows (e.g. M1 is faster at compiling code than even 8-core x86 mobile CPUs: https://build2.org/blog/apple-m1-compilation.xhtml), what do you expect from a system that is supposed to double the computational resources?
Anything mass produced is prone to failure. Just a matter of percentage that makes for a good or bad product...
Q-6