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

441–450 of 640 posts

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

#441

Post by jeffbee »

maxsilver wrote: post #374

> Why haven’t AMD/Intel been able to catch up? Is x86 just not able to keep up with the ARM architecture? When can we expect a x86 laptop chip to match the M1 in efficiency/thermals?! AMD kind of has, the "Max 395+" is (within 5% margin or so) pretty close to M4 Pro, on both performance and energy use. (it's in the 'Framework Desktop', for example, but not in their laptop lineup yet) AMD/Intel hasn't surpassed Apple…

This is not even a remotely accurate characterization of the relative performance of the Ryzen AI Max+ 395 and the Apple M4. I have both an expensive implementation of the former and the $499 version of the latter, and my M4 Mac mini beats the Ryzen by 80% or more in many single-threaded workloads, like browser benchmarks.

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

#442

Post by nodesocket »

anthonypasq wrote: post #437
maxsilver wrote: post #374

> Why haven’t AMD/Intel been able to catch up? Is x86 just not able to keep up with the ARM architecture? When can we expect a x86 laptop chip to match the M1 in efficiency/thermals?! AMD kind of has, the "Max 395+" is (within 5% margin or so) pretty close to M4 Pro, on both performance and energy use. (it's in the 'Framework Desktop', for example, but not in their laptop lineup yet) AMD/Intel hasn't surpassed Apple…

I just recently was trying to buy a laptop and was looking at that chip, but like you said, not available in anything except framework desktops and a weird tablet thats 2.5x expensive as a macbook. Its competitive on paper, but is still completely infeasible at the moment.

There there a few mini PCs using the 395+. Checkout the Beelink GTR9 Pro AMD Ryzen AI Max+ 395 and GMKtec EVO-X2.

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

#443

Post by bilbo0s »

anthonypasq wrote: post #437
maxsilver wrote: post #374

> Why haven’t AMD/Intel been able to catch up? Is x86 just not able to keep up with the ARM architecture? When can we expect a x86 laptop chip to match the M1 in efficiency/thermals?! AMD kind of has, the "Max 395+" is (within 5% margin or so) pretty close to M4 Pro, on both performance and energy use. (it's in the 'Framework Desktop', for example, but not in their laptop lineup yet) AMD/Intel hasn't surpassed Apple…

I just recently was trying to buy a laptop and was looking at that chip, but like you said, not available in anything except framework desktops and a weird tablet thats 2.5x expensive as a macbook. Its competitive on paper, but is still completely infeasible at the moment.

Also, you don't realize until you try them out that other issues make running models on the AMD chip ridiculously slow compared to running the same models on an M4. Some of that's software. But a lot is how the chip/memory/neural etc are organized.

Right now, AMD is not even in the ballpark.

In fact, the real kick in the 'nads was my fully kitted M4 laptop outperforming the AMD. I just gave up.

I'll keep checking in with AMD and Intel every generation though. It's gotta change at some point.

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

#444

Post by WorldPeas »

lelanthran wrote: post #380
lenkite wrote: post #186

Earlier quoted context omitted.

Hell, Apple CPU's are even optimized for Apple software GC calls like Retain/Release objects. It seems if you want optimal performance and power efficiency, you need to own both hardware and software. Looks like general purpose CPUs are on the losing train. Maybe Intel should invent desktop+mobile OS and design bespoke chips for those.

> Maybe Intel should invent desktop+mobile OS and design bespoke chips for those. Wouldn't it be easier for Intel to heavily modify the linus kernel instead of writing their own stack? They could even go as far as writing the sleep utilities for laptops, or even their own window manager to take advantage of the specific mods in the ISA?

they /did/ this but notice the "was" at the top of the page: https://www.clearlinux.org/

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

#445

Post by hajile »

maxsilver wrote: post #374

> Why haven’t AMD/Intel been able to catch up? Is x86 just not able to keep up with the ARM architecture? When can we expect a x86 laptop chip to match the M1 in efficiency/thermals?! AMD kind of has, the "Max 395+" is (within 5% margin or so) pretty close to M4 Pro, on both performance and energy use. (it's in the 'Framework Desktop', for example, but not in their laptop lineup yet) AMD/Intel hasn't surpassed Apple…

M4 Pro was a massive step back in perf/watt over M3 Pro. To my knowledge, there aren't any M4 die shots around which has led to speculation that yields on M4 Max were predicted to be really bad, so they made the M4 Pro into a binned M4 Max, but that comes with tradeoffs like much worse leakage current.

That said Hardware Canucks did a review of the 395 in a mobile form factor (Asus ROG Flow F13) with TDP at 70w (lower than the max 120w TDP you see in desktop reviews). This lower-than-max TDP also gets you closer to the perf/watt sweet spot.

The M4 Pro scores slightly higher in Cinebench R24 despite being 10P+4E vs a full 16P cores on the 395 all while using something like a 30% less power. M4 Pro scores nearly 35% higher than the single-core R24 benchmark too. 395 GPU performance is than M4 Pro in productivity software. More specifically, they trade blows based on which is more optimized in a particular app, but AMD GPUs have way more optimizations in general and gaming should be much better with an x86 + AMD GPU vs Rosetta 2 + GPU translation layers + Wine/crossover.

M4 Pro gets around 50% better battery life for tasks like web browsing when accounting for battery size differences and more than double the battery life per watt/hr when doing something simple like playing a video. Battery life under full load is a bit better for the 395, but doing the math, this definitely involves the 395 throttling significantly down from it's 70w TDP.

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

#446

Post by gowld »

diggan wrote: post #410
RajT88 wrote: post #366

Earlier quoted context omitted.

> All that to say: M1 is pretty fast, but the reason the battery life is better has to do with everything other than the CPU cores. That's what AMD and Intel are missing. Apple is vertically integrated and can optimize at the OS and for many applications they ship with the device. Compare that to how many cooks are in the kitchen in Wintel land. Perfect example is trying to get to the bottom of why your windows lapto…

> Apple is vertically integrated and can optimize > Compare that to how many cooks are in the kitchen in Wintel land. Perfect example is trying to get to the bottom of why your windows laptop won't go to sleep and cooks itself in your backpack So, I was thinking like this as well, and after I lost my Carbon X1 I felt adventurous, but not too adventurous, and wanted a laptop that "could just work". The thinking was "I…

Apple has been vertically integrate for 50 years. Microsoft has been horizontally integrated for 50 years.

That's why Apple is good at making a whole single system that works by itself, and Microsoft is good at making a system that works with almost everything almost everyone has made almost ever.

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

#447

Post by galad87 »

RajT88 wrote: post #366
ben-schaaf wrote: post #101

Battery efficiency comes from a million little optimizations in the technology stack, most of which comes down to using the CPU as little as possible. As such the instruction set architecture and process node aren't usually that important when it comes to your battery life. If you fully load the CPU and calculate how much energy a AI340 needs to perform a fixed workload and compare that to a M1 you'll probably find s…

> All that to say: M1 is pretty fast, but the reason the battery life is better has to do with everything other than the CPU cores. That's what AMD and Intel are missing. Apple is vertically integrated and can optimize at the OS and for many applications they ship with the device. Compare that to how many cooks are in the kitchen in Wintel land. Perfect example is trying to get to the bottom of why your windows lapto…

macOS is a resource hungry pig, I wouldn't bet too much on it making a difference.

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

#448

Post by curt15 »

qcnguy wrote: post #137
throwup238 wrote: post #110

Earlier quoted context omitted.

> All that to say: M1 is pretty fast, but the reason the battery life is better has to do with everything other than the CPU cores. That's what AMD and Intel are missing. A good demonstration is the Android kernel. By far the biggest difference between it and the stock Linux kernel is power management. Many subsystems down to the process scheduler are modified and tuned to improve battery life.

And the more relevant case for laptops is macOS, which is heavily optimized for battery life and power draw in ways that Linux just isn't, neither is Windows. A lot of the problems here can't actually be fixed by intel, amd, or anyone designing x86 laptops because getting that level of efficiency requires the ability to strongly lead the app developer community. It also requires highly competent operating system deve…

>And the more relevant case for laptops is macOS, which is heavily optimized for battery life and power draw in ways that Linux just isn't, neither is Windows.

What are some examples of power draw savings that Linux is leaving on the table?

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

#449

Post by ChuckMcM »

There are really a lot of responses to this which explain it well. The summary though might be phrased as 'alignment'. Specifically, when everyone from the mainboard engineer to the product marketeer for the product have the same goals and priorities (are aligned) the overall system reflects that. In x86 land the processor guys are always trying to 'capture more addressable market' which means features for specific things which perhaps have no value to your 'laptop' but are great for cars embedding the chip. Similarly for display manufacturers who want standards that work for everyone even if they aren't precisely what everyone wants. Need a special 'sleep the pixels that are turned off' mode for your screen ASIC which isn't part of the HDMI spec? Nah we're not gonna do that because who would use it? But Apple can, specific things in the screen that minimize power that the OS can talk to through 'side channels' that aren't part of any standard? Sure they can do that too. And if everyone is aligned on long battery life (for example) that happens. I worked at both Google and Netapp and both of them bought enough hard drives that they could demand and get specific drive firmware that did things to make their systems run better. Their software knew about the specific firmware and exploited it. They 'aligned' their vendors with their system objectives which they could do because of their volume purchases.

In the x86 laptop space the 'big' vendors like Dell, HP, Asus, Lenovo, Etc. Can do that sort of thing. Framework doesn't have the leverage yet. Linux is an issue too because that community isn't aligned either.

Alignment is facilitated by mutual self interest, vendors align because they want your business, etc. The x86 laptop industry has a very wide set of customer requirements, which is also challenging (need lots of different kinds of laptops for different needs).

The experience is especially acute when one's requirements for a piece of equipment have strayed from the 'mass market' needs so the products offered are less and less aligned with your needs. I feel this acutely as laptops move from being a programming tool to being an applications product delivery tool.

Post reply on HN