Alder Lake Core i9 processor is faster than the M1 Max

I still have a 2011 era MacBook that I can use, and was very pleasant to use until an OS update a couple of years ago.

But that was because I was able to upgrade the RAM, the SSD, and replace the battery.

None of those things are possible with macs from the last few years, so their longevity is much more likely to resemble phones than older macs.

It sounds really weird to talk about 2500 EUR laptop as "entry level".

> They want to offer computers at every price point

They don't offer anything under 1000 EUR.

Not that I'm trying to blame Apple, they've done a great job and it would be irrational for a company to not take the free money (I'd do the same and it's not that their competitors are better, high end PC laptops have similar margins it's just that less people want them so Dell, Lenovo etc. tend to sell them at massive ~50% discounts eventually (same with Android phones) which Apple never does, since their prefer to sell at relatively low volumes but with huge margins). However their prices for ram/storage are objectively predatory and the only reason they can get away with it is because there is no viable competition, it has nothing to do with "subsidizing" entry levels Macs. If they wanted to they could easily cut their prices by 500-1500 across the entire 14/16" mbp range and still remain more profitable than all the other laptop manufacturers.

Binning comes in with the lower-core-count variants of each chip.

it's a very strange feeling, they actually burn so much that they itch

It seems like most of the high performance numerical computing software today is GPU-accelerated. If performance for your job is super important, why isn't your software GPU-accelerated?

I might just be naive but it's honestly hard for me to imagine why anybody would need a good CPU and not also want a good GPU.

Are you implying that's not something to be impressed about just because lithography is not a property of the architecture?

But as we all know what Apple is/was shipping and the best you can get in a PC laptop are very different things.

My M1 Air absolutely DESTROYS my RTX 3090 in h265/422 video editing, which is what my camera (Sony A7S iii) shoots.

My 3090 would shutter and play my 60fps footage at 11fps using 350W while my passively cooled M1 doesn't miss a frame.

I expect them to come up with something much crazier with a newer design, perhaps in the M2 generation. Doesn't mean much, but the IRQ controller architecture and driver in these machines claim to scale to 8 dies, even though the existing implementation in the M1 Max is sized for 2. That's no guarantee they're looking at a product like that, but they might be.

Edit: Since the A6-era for CPU cores and and A10/11 for the GPU cores.

The only thing the Apple M1 has in common with ARM is the instruction set. The actual core design? Completely Apple. No Cortex X1s or A73s in there. Apple does use the stock ARM M3 in a few places IIRC but that's a very-low-power ~100mhz part for managing some internals.

Also, as for who can push the limits of ARM, Apple appears to know ARM better than ARM knows themselves. Both based on Apple's higher-performing core designs than ARM stock, and also because Apple cofounded ARM with VLSI and Acorn, and an Apple VP was the first CEO.

Edit: On that note, it's kind of ironic how every Android phone has the DNA of a company Apple cofounded and initially led.

Edit: thinking about this for a few more seconds, I realise this is a completely separate discussion, but I was just trying to think how Apple could accomplish the Quad branding even if internally its separate CPUs.

I expect Apple has an extreme focus on power efficiency and especially idle / leakage power, much more than Intel considering the core basically the same as they use in their phones. They also have a different approach to turbo / dvfs. So I would expect M1 to actually be a lot tighter than Intel and not have so much OC headroom.

Obviously you can buy timing with voltage to some degree, so there would be something there probably. Modern nodes are running into more problems with voltage induced breakdown though so the OC limit looks very different to what you can ship in a product. Has anyone measured M1's VDD?

> AMD chips are doing similar clocks to Intel on TSMC N7, so Apple could (but won't) have a chip running way higher than the clocks they are currently shipping with.

Not their existing microarchitecture though. They do nearly 2x the work per clock as AMD chips which necessitates more logic per stage. Getting a microarchitectural edge means making less logic do more work and it's very possible Apple have some edge there, it just wouldn't be near 2x IMO.

The silicon technology of course plays into it, but when you look at how fast individual transistors and the shortest poly to connect them can switch, speeds over 100GHz have been possible on 90nm. Today's cutting edge is probably over 200GHz (e.g., search ring oscillator). So it's not a fundamental switching speed limit of the tech that gets you.

I would say Apple could probably redo the physical design and synthesis work and minimal logic changes to target a faster and leakier device that's not suitable for phones but might be a little fairer comparison. It wouldn't put it at a 5-6GHz frequency, but could easily be enough to re-take these benchmarks and still be ahead on efficiency.

This causes heat and load spikes, and while you get no errors, the increased frequency doesn't give you any real world performance gains. Returning a to a slower configuration actually gives a much snappier and performant system.

Since this is HN, many of us like… compiling, you know, stuff. It is a regular workload for many of us on here.

The main beneficiary of the M1 Max wide memory architecture appears to be the Haskell GHC compiler, which is very, very voracious when it comes to the memory consumption and the data bus width. Regular Haskell builds shift the data around at 35-40 Gbps on average, with the Haskell optimiser hovering around 50 Gpbs and easily peaking out at 65 Gbps on medium size projects. Since cabal (the Haskell build toolchain) for some reason can't fully utilise the «-j10» parameter, roughly a 1/2 of CPU cores are not utilised. Which means we can kick off another large Haskell build in parallel without affecting memory transfer speeds of the first build.

Rust release builds are a runner-up with 30-35 Gbps transfer speeds on average. Static type analysis and type inference in Haskell and Rust is hard, is complex and requires a lot of fast memory and benefits vastly from large CPU caches.

I have not attemped a Rust LTO build yet, but I would expect the LTO builds to like the wide memory bus, like, A LOT. In general, a LTO build (e.g. large C++ builds) is heavily memory bound and shifts data blocks, small and large, around extensively.

But an application is not very useful with a dataset, so many of us then like training our data models. Data model training can easily saturate 200 Gbps per a M1 Max CPU cluster.

Then the application needs to load the data in and usually do something with it. For example, issue complex queries to a graph database. Large CPU caches coupled with a wide data bus is a winner here.

All of the above are examples of real, practical and regular workloads.

Apple's foiree into ARM has been going better than it could have gone, but recommending an ARM chip still comes with several asterisks, which make it kinda hard to recommend for people who don't make a living out of reading/writing words on a screen and highly specific creative work that may-or-may-not benefit from the architecture of the chip. Regardless, they've got their work cut out for them, and it should be interesting to see how Apple responds.

[0] https://youtu.be/VHUF8A2vpos?t=423

We all know GPUs burn watts when being heavily used. But when a laptop GPU is just showing a video shouldn’t it be able to scale down pretty far?

If I’m just surfing the web or working in Word I wouldn’t want my laptop GPU using 15-20 watts.

Surely it can get down to like 2 watts right? Or less?

In my experience if the dGPU is on at all it won’t get 6 hours of anything.

Typically if I plugged a monitor in I also had the power cord in. And with the Power plugged in, I had the laptop set to highest performance mode in Windows.

There is a balanced mode which is more conservative with power consumption. I wonder if the GPU idle draw would have been lower.

The laptop I used also had the external TB4/USB-C and HDMI ports wired to the NVIDIA GPU. This means that external monitor = GPU on. I wonder if there are some laptops that can use the iGPU for basic tasks, and only switch to GPU as needed.

There are also laptops with less powerful discrete GPUs that might draw less power at idle.

Chrome has separate processes and fewer tabs on my machine. Still summing them up seems to reach 3gb.

IntellJ takes 6gb on my machine so it's the biggest individual user of RAM.

I think parent post might be exaggerating but I have the 64gb M1 Max.

I love the Air but excited for my 64gb macbook pro arriving in a month.

It's not fantastic by any means, but it's a disservice to call it _horrible_.

Why in the world would you want a 1080p display just to get an extra 120 FPS? Even if you can perceive that difference (which I kind of doubt) you’re not going to get rates that high with your settings up high anyway. In The Verge’s tests only the old CS:GO got above 150 (it got over 400!).

“ In conclusion, putting a 1080p screen on this system is like sticking an MLB player in a Little League game. Sure, he’s gonna look impressive, but what is the point? These are 4K chips, and I’m not just referring to the price of this unit.”