Live data from Hacker News

Apple M5 chip

apple.com

161–170 of 1001 posts

Re: Apple M5 chip

#161

The modern Apple feels like their hardware teams way outperforming the software teams.

I pretty much see the Macbook as some fancy toys with mediocre software. Maybe the kernel is solid but other software are very meh, even comparing to Windows. But I'm definitely biased as a Windows/Linux user, and my hobby is system programming so naturally a Linux box is more suitable.

Biggest grief with MacOS software:

- Finder is very mediocre comparing to even File explorer in Windows

- Scrollbar and other UI issues

Unfortunately I don't think Asahi is going to catch up, and Macbook is so expensive, so I'll probably keep buying second hand Dell/Lenovo laptop and dump a Linux on top of it.

Re: Apple M5 chip

#162
post #2

This is only the base model, no upgrades yet for the Pro/Max version. The memory bandwidth is 153GB/s which is not enough to run viable open source LLM models properly.

You don’t want to be bandwidth-bound, sure. But it all depends on how much compute power you have to begin with. 153GB/s is probably not enough bandwidth for an Rtx5090. But for the entry laptop/tablet chip M5? It’s likely plenty.

Re: Apple M5 chip

#164

The modern Apple feels like their hardware teams way outperforming the software teams.

> The modern Apple feels like their hardware teams way outperforming the software teams. There aren't a lot of tangible gains left to be made by the software teams. The OS is fine, the office suite is fine, the entertainment apps are fine. If "performance" is shoving AI crap into software that was already doing what I wanted it to do, I'd rather the devs take a vacation.

There were a few things on that page that made me excited for the future of where computing is going, but I do think we're going to hit a "lull" in terms of exciting new features until some of the really futuristic stuff comes to pass.

Who knows, maybe the era of "exciting computing" is over, and iteration will be a more pleasant and subtle gradient curve of improvements, over the earth-shattering announcements of yore (such as the advent of popular cellular phones).

Re: Apple M5 chip

#165
So how many hardware systems does Apple silicon have for doing matrix multiplies now?

1. CPU, via SIMD/NEON instructions (just dot products)

2. CPU, via AMX coprocessor (entire matrix multiplies, M1-M3)

3. CPU, via SME (M4)

4. GPU, via Metal (compute shaders + simdgroup-matrix + mps matrix kernels)

4. Neural Engine via CoreML (advisory)

Apple also appears to be adding a “Neural Accelerator” to each core on the M5?

Re: Apple M5 chip

#166

Earlier quoted context omitted.

[flagged]

What I have seen with iPhones is that the ram has gone from 4gb to 12gb very quickly compared to how it went from 1gb to 3gb. Apps used to use less ram but over the years apps have become big and more complicated. This is probably why iPhones feel sluggish because new iPhones have more memory and apps snap back faster as newer iPhones which also have faster storage and memory bandwidth to reduce latency of reading mo…

Apple has been proven to intentionally slow down older devices, but it's definitely not to inflate their profits. It's just a way to kindly preserve your old battery for you. And they try to keep it a secret from you so you don't get confused.

Re: Apple M5 chip

#167
It’s unfortunate that this announcement is still unspecific about what they improved in the Neural Engine. Since all we know about the Neural Engine comes from Apple papers or reverse engineering efforts (https://github.com/hollance/neural-engine), it’s plausible that they addressed some quirks to enable better transformer performance. They have written quite interesting papers on transformers on the Neural Engine:

- https://machinelearning.apple.com/research/neural-engine-tra...

- https://machinelearning.apple.com/research/vision-transforme...

Things have definitely gotten better with MLX on the software side, though it still seems they could do more in that area (let’s see what the M5 Max brings). But even if they made big strides here, it won’t help previous generations, and the main thing limiting Apple Intelligence (in my opinion) will continue to be the 8 GB of unified memory they still insist on.

Re: Apple M5 chip

#168
post #40
post #26

Earlier quoted context omitted.

Not really. Back in the day you wouldn't buy a MacBook because it was powerful. Most likely it had a very shitty Intel CPU with not a lot of cores and with thermal challenges, and the reason you bought it was because macOS.

> very shitty Intel CPU with not a lot of cores and with thermal challenges Very often the intel chips in macbooks were stellar, they were just seriously inhibited by Apples terrible cooling designs and so were permanently throttled. They could never provide decent cooling for the chips coupled with their desire to make paper thin devices.

They made things even worse with fan curves tuned for silence until the CPU was practically at TjMax.

Re: Apple M5 chip

#169
post #153

Wonder how it compares with the M4 Max that I've just bought haha

Same I just bought an M4 Max 2 weeks ago and had a bit of anxiety for a moment. I'm going to justify it because they haven't released M5 Max yet

It's going to be fine, what's important is what we do with the thingy :)

Logos is King

Re: Apple M5 chip

#170
post #103

The M5 MacBook Pro still gets the Broadcom WiFi chip but the M5 iPad Pros get the N1 and C1X (Sweet). All in all, apple is doing some incredible things with hardware. Software teams at apple really need to get their act together. The M1 itself is so powerful that nobody really needs to upgrade that for most things most people do on their computers. Tahoe however makes my M1 Air feel sluggish doing the exact same task…

The Broadcom WiFi support 320Mhz while N1 is stuck with 160Mhz. There were report of N1 not supporting 4096 QAM as well but I didn't check.

Does it? If it’s the same WiFi chip used in other M4 Mac’s then it’s still limited to 160MHz:

https://support.apple.com/en-gb/guide/deployment/dep268652e6...

Post reply on HN