Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: Why hasn't x86 caught up with Apple M series?

news.ycombinator.com

Post reply on HN
Print view View original

121–130 of 640 posts

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#121

Post by ozgrakkurt »

Remnant44 wrote: post #16
trashface wrote: post #6

I may be out of date or wrong, but I recall when the M1 came out there was some claims that x86 could never catch up, because there is an instruction decoding bottleneck (instructions are all variable size), which the M1 does not have, or can do in parallel. Because of that bottleneck x86 needs to use other tricks to get speed and those run hot.

ARM instructions are fixed size, while x86 are variable. This makes a wide decoder fairly trivial for ARM, while it is complex and difficult for x86. However, this doesn't really hold up as the cause for the difference. The Zen4/5 chips, for example, source the vast majority of their instructions out of their uOp trace cache, where the instructions have already been decoded. This also saves power - even on ARM, decod…

But is the uOp trace cache free? It surely doesn’t magically decode and put stuff in there without cost

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#122

Post by danb1974 »

Macbooks are more like "phone/tablet hardware evolved into desktop" mindset (low power, high performance). x86 hardware is the other way around (high power, we'll see about performance).

That being said, my M2 beats the ... out of my twice as expensive work laptop when compiling an arduino project. Literall jaw drop the first time I compiled on the M2.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#123

Post by marshray »

JustExAWS wrote: post #5
roscas wrote: post #4

RISC vs CISC. Why you think a mainframe is so fast? ARM is great. Those M are the only thing I could buy used and put Linux on it.

I thought people stopped believing this around 2005 when Apple users finally had to admit that PPC was behind x86. Even though this was the case for the most part during the entire history of PPC Macs (I owned two during these years) https://chipsandcheese.com/p/arm-or-x86-isa-doesnt-matter

RISC lost its meaning once SPARC added an integer multiply instruction.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#124

Post by Panzer04 »

Software.

If you actually benchmark said chips in a computational workload I'd imagine the newer chip should handily beat the old M1.

I find both windows and Linux have questionable power management by default.

On top of that, snappiness/responsiveness has very little to do with the processor and everything to do with the software sitting on top of it.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#125

Post by kangs »

aurareturn wrote: post #117
noelwelsh wrote: post #99

Like a few other comments have mentioned, AMD's Strix Halo / AI Max 380 and above is the chip family that is closest to what Apple has done with the M series. It has integrated memory and decent GPU. A few iterations of this should be comparable to the M series (and should make local LLMs very feasible, if that is your jam.)

On Cinebench 2025 single threaded, M4 is roughly 4x more efficient and 50% faster than Strix Halo. These numbers can be verified by googling Notebookcheck. How many iterations to match Apple?

yes and no. i have macbook pro m4 and a zbook g1a (ai max 395+ ie strix halo)

In day to day usage the strix halo is significantly faster, and especially when large context LLM and games are used - but also typical stuff like Lightroom (gpu heavy) etc.

on the flip side the m4 battery life is significantly longer (but also the mpb is approx 1/4 heavier)

for what its worth i also have a t14 with a snapdragon X elite and while its battery is closer to a mbp, its just kinda slow and clunky.

so my best machine right now is the x86 actually!

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#126

Post by Panzer04 »

dmitrygr wrote: post #11

There is one positive to all of this. Finally, we can stop listening to people who keep saying that Apple Silicon is ahead of everyone else because they have access to better process. There are now chips on better processes than M1 that still deliver much worse performance per watt.

Because of a random anecdote on hackwrnews?

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#127

Post by Remnant44 »

Tuna-Fish wrote: post #65
PaulKeeble wrote: post #60

I tend to think its putting the memory on the package. Putting the memory on the package has given the M1 over 400GB/s which is a good 4x that on a usual dual channel x64 CPU and the latency is half that of going out to a DRAM slot. That is drastic and I remember when the northbrige was first folded into the CPU by AMD with the Athlon and it had a similarly big improvements in performance. It also reduces power consu…

> and the latency is half that of going out to a DRAM slot. No, it's not. DRAM latency on Apple Silicon is significantly higher than on the desktop, mainly because they use LPDDR which has higher latencies.

I was going to mention this as well.

Source: chipsandcheese.com memory latency graphs

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#128

Post by saati »

Remnant44 wrote: post #16
trashface wrote: post #6

I may be out of date or wrong, but I recall when the M1 came out there was some claims that x86 could never catch up, because there is an instruction decoding bottleneck (instructions are all variable size), which the M1 does not have, or can do in parallel. Because of that bottleneck x86 needs to use other tricks to get speed and those run hot.

ARM instructions are fixed size, while x86 are variable. This makes a wide decoder fairly trivial for ARM, while it is complex and difficult for x86. However, this doesn't really hold up as the cause for the difference. The Zen4/5 chips, for example, source the vast majority of their instructions out of their uOp trace cache, where the instructions have already been decoded. This also saves power - even on ARM, decod…

Zens don't have a trace cache, just an uop cache.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#129

Post by mschuster91 »

> I am sorely disappointed, using the Framework feels like using an older Intel based Mac. If I open too many tabs in Chrome I can feel the bottom of the laptop getting hot, open a YouTube video and the fans will often spin up.

A big thing is storage. Apple uses extremely fast storage directly attached to the SoC and physically very very close. In contrast, most x86 systems use storage that's socketed (which adds physical signal runtime) and that goes via another chip (southbridge). That means, unlike Mac devices that can use storage as swap without much practical impact, x86 devices have a serious performance penalty.

Another part of the issue when it comes to cooling is that Apple is virtually the only laptop manufacturer that makes solid full aluminium frames, whereas most x86 laptops are made out of plastic and, for higher-end ones, magnesium alloy. That gives Apple the advantage of being able to use the entire frame to cool the laptop, allowing far more thermal input before saturation occurs and the fans have to activate.

Re: Ask HN: Why hasn't x86 caught up with Apple M series?

#130

Post by mschuster91 »

BlindEyeHalo wrote: post #112
mmcnl wrote: post #92

A lot of insightful comments already, but there are two other tricks I think Apple is using: (1) the laptops can get really hot before the fans turn on audibly and (2) the fans are engineered to be super quiet. So even if they run on low RPM, you won't hear them. This makes the M-series seem even more efficient than they are. Also, especially the MacBook Pros have really large batteries, on average larger than the co…

The macbook air doesn't even have a fan. I don't think you could built a fan-less x86 laptop.

You can, the thing is you have to build it out of a solid piece of metal. Either that's patented by Apple or it is too expensive for x86 system builders.
Post reply on HN