Hacker News Reader

Comment by mhitza | Hacker News Reader

> Is it unreasonable to think Framework should be able to make a laptop competitive with the 5 years old MacBook Air M1?

Kind of unreasonable. I mean which Intel or AMD cpu can be run fanless and perform well?

On the topic of displays, my understanding is that they "kind of use what they can get". That's how there can be a 13 display with rounded corners in a straight edge case.

What you're asking are the things I'm looking for, though still every time I go into their forum I see enough thermal, fan noise issues and AMD firmware bugs, that I'm still on the fence on buying one.

I wish them luck with the 12, for me sounds like a model for "true believers" because it doesn't seem to compete well enough with run of the mill chromebooks (or an Air) that are more established in the students segment.

View on HN

It isn't the chip which determines whether it's fanless. Basically every modern chip supports power capping and then the power cap is determined by how much heat the machine can dissipate.

What that really determines is multi-thread performance. Fanless laptop that can dissipate the power of one core? No problem. Fanless laptop that can dissipate the power of all the cores? For that you have to lower the clock speed quite a bit. Which is why you see AMD chips on older TSMC process nodes getting better multithread performance than Apple's fanless ones.

The cost/benefit ratio of adding a fan is extremely attractive. The alternative way of doing it is to add more cores. If you have 8 fanless cores at 2 GHz, how do you improve multi-thread performance by 50%? Option one, clock them at 3 GHz, but now you need a fan; cost of fan ~$5. Option two, get 16 cores and cap them at 1.5 GHz to fit in the same power envelope, but now you need twice as much silicon, cost of twice as many cores $500+.

The number of people who pick the second option given that trade off is so small that hardly anybody even bothers to offer it.

Apple continues to do it because a) then they get to claim "see, they can't do this?" even when hardly anybody chooses that given the option, and b) then if you actually want the higher performance one from them, you're paying hundreds of dollars extra for more cores instead of $5 extra for the same one but with a fan in it.

Doesn't this miss differences between CPUs in their per-core efficiency?

The per-core efficiency of Apple and AMD CPUs on the same process node is pretty much identical. This has become harder to directly compare because they're now using alternate process nodes from one another, but have a look at this chart for example:

https://www.cpu-monkey.com/en/cpu_benchmark-cpu_performance_...

What do we see at the top of this chart? TSMC 3nm (M3/M4), followed by TSMC 4nm (Ryzen 7000U/8000U), TSMC 5nm (M1/M2), TSMC 5nm/6nm mixed (Ryzen 7000H), and then finally we find something made on an Intel process node instead of TSMC.

The efficiency has more to do with the process node than which architecture it is.

It's too bad they don't have Epyc on that chart. Epyc 9845 is on TSMC N3E and that thing is running cores at a >2GHz base clock at less than 2.5W per core.

You're linking to a multi-core benchmark. The story is a lot more in favor of Apple if you look at single-core efficiency, Apple is roughly 2-3x more efficient: https://www.notebookcheck.net/Intel-Lunar-Lake-CPU-analysis-...

And this benchmark doesn't even include M4, which is even more efficient.

> The story is a lot more in favor of Apple if you look at single-core efficiency

Your link is comparing the M3 against AMD chips with higher TDPs. Higher TDPs tank "single-core efficiency" because power consumption is non-linear with clock speed. Give a core near its limit three times the power budget and you're basically dividing the single-core efficiency by three because you burn three times more power and barely improve single-thread performance at all, and then that's exactly what you see there.

To have a useful comparison you have to compare the efficiency of CPUs when they're set to use the same amount of power.

Ofcourse, but that's the whole point right? At similar power levels, Apple M-series is a lot faster. The only way for AMD and Intel to compete on performance is by providing more power, but this obviously has a negative effect on efficiency. So with AMD/Intel you have to choose: efficient and slow or inefficient and fast. With Apple you can actually have it both ways.

> Ofcourse, but that's the whole point right? At similar power levels, Apple M-series is a lot faster. The only way for AMD and Intel to compete on performance is by providing more power

They're saying the exact opposite of that. Their claim is that all the extra power is only juicing performance a little bit, so at similar power levels the performance is not all that far apart.

OK, so I'm still curious about the underlying efficiency. Is there a better test of single core efficiency at the same TDP? Does Apple Silicon provide any noticeable improvement over other x86 or Arm chips in that scenario, or is it really just a straightforward function of process and node size?

Someone needs to run the same single core benchmarks on Intel and AMD while forcing those CPUs to only use 5 watts. In a few minutes of searching I was not able to find anyone doing power-limited single core benchmarks.

"The cost/benefit ratio of adding a fan is extremely attractive."

Depends on your metric. A fan makes noise, attracts dirt that needs cleaning, needs more space ...

I really love my fanless devices, even though they never will reach the speed of activly cooled ones.

Sure, and you can still find fanless devices, but then they'll typically be the ones not focused on multi-thread performance. And if you don't care about that, e.g. because you're offloading heavy workloads to a server or you just don't do anything compute heavy, then you can find a lot of fanless offerings with low core counts that are actually quite inexpensive. You can get some fanless Chromebooks for under $200.

It's a bit old at this point but the pixel slate chromebooks can be spec'ed with an i7 and 16GB of ram. I think they go for under 200 on ebay and are decent if you don't need compute heavy apps.

If someone besides Apple made a fanless laptop that had competitive performance with Apple's offerings (i.e. not a $200 Chromebook with a Celeron or a cast-off 5 year old smartphone CPU), I'd absolutely buy one. I got excited when the Qualcomm Snapdragon X was being discussed pre-launch, but then it came out with performance worse than the original M1 and it turned out that Qualcomm lied about giving it first-class Linux support. I really dislike Mac OS, but when I can't use a PC laptop in bed or on a couch or on my lap without it overheating, I'm not able to switch away. It's a shame that the entire PC industry is fine with selling laptops that will overheat when not used on a rigid flat surface.

I believe the Microsoft Surface Pro 7 is fanless. Sadly, the 8 and 9 have a fan.

I mean, you could just buy something that allows you to configure the TDP in software and then set it low enough that the fan doesn't run. You'd be sacrificing a non-trivial amount of multi-thread performance, but that's what the fanless Macbooks are doing anyway.

> The number of people who pick the second option given that trade off is so small that hardly anybody even bothers to offer it.

The number of manufacturers or the number of people? Apple was on the path to laptop irrelevancy before the M series, it doesn't seem clear to me at all that people don't care about noise and heat along with performance.

People generally have a priority between noise and heat vs. performance. If you don't do compute-heavy stuff then you might as well have something quiet. If you do, i.e. you're always waiting on the machine, how many of those people want to sacrifice a third of their performance to avoid having a fan?

But it’s not a binary choice:

Performance along which characteristics? How much performance does one need locally? At which point does heat/noise/energy cost become too much for a mobile workstation?

All of these are additional criteria that the M series laptops competes in (and in many cases wins), even amongst programmers who are some of the most compute performance sensitive consumers out there.

> thermal, fan noise issues

Anecdotal, obviously, but disabling Turbo-Core [0] on my AMD Framework 13 stopped all of my fan noise and heat complaints, with no noticeable performance impacts. It went from being so loud that my wife on the other side of the room would ask if my computer was okay to quieter than my ThinkPad, and from noticeably hot to just slightly warm.

Kind of ridiculous that it takes messing with an obscure system file to resolve it, but not any more ridiculous than issues I've had with other brands.

[0] It's `echo 0 > /sys/devices/system/cpu/cpufreq/boost` or something like that, and `echo 1` to turn it back on when you want that extra performance.

You might not have noticed, but your single-core performance will take a serious hit if you disable turbo boost. For the AMD 7480u, turbo boost frequency is 5.1Ghz vs 3.3GHz base clock frequency. If you disable turbo boost you lose 36% single-core performance.

Wouldnt something like this tool [0] be a cleaner solution?

[0]: https://github.com/FlyGoat/RyzenAdj

Quite possibly, I will check it out. Thank you.

Intel's T variants in the Core series can be passively cooled. They have pretty good burst performance in case you need it. I don't know if there are laptops using them (I only have fanless desktop systems with these CPUs).

Besides, at least in Linux, lots of kernel options can tweak Intel/AMD CPUs to make them mostly silent.

The problem is that manufacturers don't put much thought into building good cooling systems.

Lenovo, for instance, has so many SKUs that it's really random. A few are great, but some sound like a hairdyer or rev up too aggressively.

Apple gets this. By having a small product line, they usually polish all those details.

> which Intel or AMD cpu can be run fanless and perform well?

For example, AMD Ryzen 7 8840U or 7840U can be configured for the same 15W TDP as Apple M1. At 15W, the overall performance going to be about the same as M1.

> I mean which Intel or AMD cpu can be run fanless and perform well?

Lunar Lake.

> I mean which Intel or AMD cpu can be run fanless and perform well?

I don't follow CPU news and have no idea what lake they're at now, but I'd be surprised if Intel and AMD didn't have a chip competitive with an M1 by now.

When I google "fanless amd intel laptop cpu" I find this old thread https://news.ycombinator.com/item?id=31142209 which does suggest some fanless machines exist. That's from 3 years ago so surely there are even more options today, no?

To put it simply, I don't think we'll get anything closer to the M1 on the x86 architecture.

You'll have to wait for Framework to offer a Snapdragon instead of Intel/AMD but they haven't announced anything yet.

The first-generation Intel Ultra lineup is comparable to the M1 and M1 Max. See: https://nanoreview.net/en/cpu-compare/intel-core-ultra-7-165...

Intel's integrated graphics aren't as good, but they are similar in terms of power consumption & CPU performance.

Compared to M4, well, that's a different beast entirely. I'm not sure what's the latest there.

Reply on news.ycombinator.com