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

AMD's Strix Point: Zen 5 Hits Mobile

chipsandcheese.com

Post reply on HN
Print view View original

221–230 of 241 posts

Re: AMD's Strix Point: Zen 5 Hits Mobile

#221

Post by akvadrako »

jml7c5 wrote: post #56

I sit firm in my belief that the best thing Microsoft could do for their laptop ecosystem is to add support for a "max fan speed" slider somewhere prominent in the Windows UI. People want the option to make their laptop silent or nearly silent. And when users do need the power, they generally prefer a slightly slower laptop at a reasonable volume rather than the roar of a jet engine. Laptop manufacturers want their d…

This exists.

My friend's new ASUS ROG G16 let's you control maximum power draw for the CPU / GPU and draw the fan curve (temp → RPM).

The only thing that seems missing is a (temperature → max power curve) to handle spikes.

Re: AMD's Strix Point: Zen 5 Hits Mobile

#222

Post by AnthonyMouse »

aurareturn wrote: post #219
AnthonyMouse wrote: post #217

Earlier quoted context omitted.

7730U (7nm) vs. M1 (5nm) for MT: https://nanoreview.net/en/cpu-compare/apple-m1-vs-amd-ryzen-... Faster in Passmark MT, basically tied in Geekbench MT, faster in average MT score. HX 370 vs. M1: https://nanoreview.net/en/cpu-compare/apple-m1-vs-amd-ryzen-... Faster in everything, ST and MT. ST difference is significant, MT difference is huge. Obviously this is expected because in this comparison AMD has the process a…

Let's not use Passmark MT. Stick to the better benchmarks that are optimized for both ARM and x86. GB5 and GB6, M1 is faster in MT despite having 4 fewer cores. If you can find SPEC scores, that'd be great too. HX 370 vs M1, what's the perf/watt for SPEC and GB5/6 and Cinebench 2024? HX370 consumes a lot more power. Hence, there aren't any fanless laptops available for it. 4 years later, AMD's chips still can't work…

> Let's not use Passmark MT. Stick to the better benchmarks that are optimized for both ARM and x86.

At some point you just run out of benchmarks. The majority of benchmarks people ordinarily use already don't run on Macs.

> GB5 and GB6, M1 is faster in MT despite having 4 fewer cores.

It has the same number of cores as the 7730U. Half the M1's cores are E-cores, but that should be an advantage on a comparison at a given power level because E-cores have better performance per watt. The M1 gets within the margin of error of the same MT score on the benchmark you actually like even though the M1 is built on a newer process.

> HX 370 vs M1, what's the perf/watt for SPEC and GB5/6 and Cinebench 2024?

You keep asking for benchmarks that probably nobody has published.

> HX370 consumes a lot more power. Hence, there aren't any fanless laptops available for it.

It has a configurable TDP down to 15W. You can make a laptop that passively dissipates 15W. But you can also make a laptop with a fan in it which is capable of higher performance from the same silicon and then have a setting for "silent mode" that lets you switch between them at will. People generally like that better so that's what they make.

Re: AMD's Strix Point: Zen 5 Hits Mobile

#224

Post by acchow »

sm_1024 wrote: post #2

IMO, the most interesting thing about this line is the battery life---within an hour of MBP3 and within 2 hours of Asus's Qualcomm. Making it comparable to ARM architectures. Which is a little surprising because ARM is commonly believed to be much more power efficient than x86. [1] https://youtu.be/Z8WKR0VHfJw?si=A7zbFY2lsDa8iVQN&t=277

> within an hour of MBP3

Not a good way to measure. The Zenbook S16 has a larger 78Wh battery vs the MacBook Pro’s 69.6Wh.

So that’s 11% less battery life despite 12% more battery capacity.

Re: AMD's Strix Point: Zen 5 Hits Mobile

#225

Post by Alifatisk »

EVa5I7bHFq9mnYK wrote: post #223

Do all new processors include NPUs now? What if I don't need AI, should I still pay for the unneeded transistors? If only things could be made modular/reconfigurable.

I wonder if the NPUs can help other parts of the computer offload some tasks anyway?

Re: AMD's Strix Point: Zen 5 Hits Mobile

#226

Post by cubefox »

hajile wrote: post #137
cubefox wrote: post #34

Earlier quoted context omitted.

Unlike AMD and Qualcomm, Apple uses an expensive TSMC 3nm process, so you would expect better battery life from the "MBP3". I assume they used the process improvements to increase performance instead.

Perf per watt is higher for M1 on N5 vs Zen5 on N4P, so the problems go deeper than just process. X Elite also beats AMD/Intel in perf/watt while being on the same N4P node as HX370. https://www.notebookcheck.net/AMD-Zen-5-Strix-Point-CPU-anal...

Performance per watt also depends on clock speed. Other things equal, higher clock speed means worse performance per watt.

Re: AMD's Strix Point: Zen 5 Hits Mobile

#227

Post by esperent »

rafaelmn wrote: post #86
cmeacham98 wrote: post #85

Earlier quoted context omitted.

The cheapest MacBook Air is $1000, and it's more like $1500+ if you want a reasonable amount of RAM and storage. There are similarly expensive Windows laptops available that are fanless.

Mind linking some (genuinely curious, would like to checkout potential Linux machine for the next upgrade)

The Huawei Mate book X.

I think there's a fanless Asus ZenBook too, or the Surface Pro X.

https://www.notebookcheck.net/Huawei-MateBook-X-in-review-Ul...

Re: AMD's Strix Point: Zen 5 Hits Mobile

#228

Post by flostk »

mshockwave wrote: post #150

Switching from individual schedulers to unified one for integer execution makes sense to me, but I still don’t quite understand why FP execution units do the opposite, could somebody explain why?

In this interview Mike Clark from AMD explains that a little, search for "third scheduler":

https://chipsandcheese.com/2024/07/15/a-video-interview-with...

The way I understand it, it's a combination of a unified scheduler for floating point being difficult to implement because many FP instructions need multiple cycles, and FP code being more regular in practice, so you don't need the scheduler to be as powerful.

Re: AMD's Strix Point: Zen 5 Hits Mobile

#229

Post by aurareturn »

AnthonyMouse wrote: post #222
aurareturn wrote: post #219

Earlier quoted context omitted.

Let's not use Passmark MT. Stick to the better benchmarks that are optimized for both ARM and x86. GB5 and GB6, M1 is faster in MT despite having 4 fewer cores. If you can find SPEC scores, that'd be great too. HX 370 vs M1, what's the perf/watt for SPEC and GB5/6 and Cinebench 2024? HX370 consumes a lot more power. Hence, there aren't any fanless laptops available for it. 4 years later, AMD's chips still can't work…

> Let's not use Passmark MT. Stick to the better benchmarks that are optimized for both ARM and x86. At some point you just run out of benchmarks. The majority of benchmarks people ordinarily use already don't run on Macs. > GB5 and GB6, M1 is faster in MT despite having 4 fewer cores. It has the same number of cores as the 7730U. Half the M1's cores are E-cores, but that should be an advantage on a comparison at a g…

>At some point you just run out of benchmarks. The majority of benchmarks people ordinarily use already don't run on Macs.

You just need GB5 or GB6. They are correlated to SPEC. Anything else is sort of worthless in 2024.

>It has the same number of cores as the 7730U. Half the M1's cores are E-cores, but that should be an advantage on a comparison at a given power level because E-cores have better performance per watt. The M1 gets within the margin of error of the same MT score on the benchmark you actually like even though the M1 is built on a newer process.

You're right, the 7730U does have 8 cores only. My mistake.

Re: AMD's Strix Point: Zen 5 Hits Mobile

#230

Post by dagmx »

AnthonyMouse wrote: post #220
dagmx wrote: post #216

Earlier quoted context omitted.

Rendering is parallel, yes, but it also makes heavy use of SIMD to accelerate the operations per thread. One does not obviate the other. In the most trivial case, sure, the JavaScript runtimes won’t compile to use SIMD but there’s lot of cases where they will as well as part of their JIT. I think you’re trivializing how they work. And back to the main point, it doesn’t really matter if you believe Cinebench reflects…

> The fact is that R23 uses SIMD for x86, but not for ARM. R24 rectifies that. Both R23 and R24 use the same rendering code path regardless of running in single or multi threaded mode. But that's not the issue. Even if Cinebench R23 isn't a valid comparison, Zen5 also faster for Cinebench R24 MT. Cinebench ST (R24 or R23) turns out to be a silly benchmark, because nobody in real life is going to artificially limit th…

Again, I think your logic here is flawed. It’s still valuable to know how a single core behaves for a rendering work load.

I’ve worked professionally in feature film production. We test single threaded performance all the time.

It tells us the performance characteristics of the nodes on our farm, and lets us more accurately get a picture of how jobs will scale and/or be packed.

It’s not uncommon to have a single thread rendering, for example to update part of an image while an artist works on other parts.

It’s not just a test of how wide the chip is. It also tests things like how it actually handles the various instruction sets from a real world codebase. Not all processors that are equally “wide” (not the right term but whatever) handle AVX the same, and you need to know how a single core behaves for that. It’s also useful to see how the cores actually behave on their own so you can eliminate the overhead of thread synchronization and system scheduling affecting you.

Post reply on HN