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

2020 Mac Mini – Putting Apple Silicon M1 To The Test

anandtech.com

Post reply on HN
Print view View original

481–490 of 767 posts

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#481

Post by sickygnar »

kissiel wrote: post #76
geerlingguy wrote: post #55

Earlier quoted context omitted.

I think the overall point is that for the average user, who doesn't need all those cores or could make good use of them, the M1 may in fact feel / be faster. For users like you or I, of course we'd see a huge difference, but not everyone is running workloads that need more than 2 or 4 cores.

I have a theory on why ST perf is always the most important metric for me, and some other folks. When you're waiting for something synchronously, like rendering a webpage, stuff to open, etc. you're usually running a ST load. For stuff that can benefit from multithreading it's usually planned task. So does it make a difference if it takes 4 minutes compared to 3? You will still context switch.

Right now I have ~20 tabs open and a few apps, a workload which is probably similar to the average user. My machine currently has 510 processes running with 2379 threads, though most of them are background. I'd wager core count is more important than ST performance nowadays, especially considering the fact that applications seem to be multicore optimized.

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#482

Post by tachyonbeam »

bscphil wrote: post #459
jmagoon wrote: post #446

Earlier quoted context omitted.

It is, in fact, already used by everyone, because it's an evolution of the chipset in basically every smartphone in the world with widely divergent target audiences, income brackets, and technical limitations.

I don't know that that's a fair comparison. Just because it's an ARMv8 chip doesn't mean it's directly comparable to what's in smartphones. (I assume you aren't comparing it to Apple made chips for iPhone specifically, since then it wouldn't be true that it's in "basically every smartphone in the world".) In particular, this is the first 5nm chip to be widely available, and by most accounts on performance it competes…

AFAIK you are correct. Apple has completely redesigned their own ARM chip. It has the same instruction set (or a superset of the instruction set) as what runs in a cellphone, but the design is completely different from say, Qualcomm chips.

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#483

Post by bengale »

zaroth wrote: post #426
em500 wrote: post #228

Earlier quoted context omitted.

Here's one data point: a WebKit compile took 25min on the Air vs 20min on the Mini/Pro. That 25min is still a bit faster than the Intel 16-inch Pro, which took 27min and waaay faster than Intel 13-inch Pro at 46min. The crazy thing is that both M1 MacBooks still had 91% battery left after the compile, vs 61% on the 16-inch Pro and 24% on the 13-inch Intel Pro. [1] https://techcrunch.com/2020/11/17/yeah-apples-m1-macb…

How is that anything less than mind blowing ? Twice as fast, using 1/10th the battery life.... and that’s for a part that costs Apple $70 instead of, what, $400?

Can you imagine how frustrating it must have been at Apple knowing what you had and having to deal with intel’s crap over the last year or two.

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#484

Post by Razengan »

xuki wrote: post #224
interestica wrote: post #210

Earlier quoted context omitted.

These gains are just through improvements in processors. Now, combine that with parallel leaps for battery tech... It's also good that battery tech hasn't jumped too far ahead already - the constraint has helped in the development in these mobile chips. The display panel is still the biggest energy draw on my laptop. There's a lot of chance for improvement there too. A 2 day battery in 5 years isn't that crazy.

Apple is the kind of company that will choose to slim down the laptop instead of making it 2 day battery life.

Apple is the kind of company that will choose to slim down the laptop and improve the performance AND increase the battery life.

Like they just fucking did with the M1 MacBooks.

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#485

Post by michaelmrose »

Why did the single threaded cinebench single threaded measure against the Ryzen 5950x a high end desktop processor then the multithreaded version of the same benchmark only list the 4900HS a high end laptop part with a fraction of the thermal budget and half the cores?

Other sites had the Ryzen 5950x pegged at 28,641 in the multithreaded version vs 7833 for the m1 mac mini.

Its not really surprising that something with 4x as many high performance cores as the m1 with a much higher thermal budget is almost 4x faster than the m1.

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#486

Post by skavi »

jeswin wrote: post #162

This is as much a ringing endorsement of AMD Ryzen as it is of the M1. The 15W 4800U is just as impressive as the M1, and the performance per watt gap seems small enough to bridge with AMDs upcoming 5nm switch.

You are making the mistake of equating TDP to actual power draw. The Yoga Slim 7, which uses a 4800U has an average load power of ~50W[0] vs the Mini's load power of ~30W.

[0]: https://www.notebookcheck.net/The-Ryzen-7-4800U-is-an-Absolu...

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#487

Post by amelius »

blunte wrote: post #196
mattlondon wrote: post #33

Wow it must suck to be Intel right now. It wasn't so long ago that the trope was while others had better multi-core performance "...Intel still holds the lead for single core performance" Now not only do AMD have a better product, but also Apple now offer equal or better performance than the best that Intel can offer. I wonder what is next for Intel now their £1000+ CPUs are firmly in third place. Looking forward to…

This happens to every giant eventually (and to countries or civilizations). They get climb to the top, and then they hold such dominant positions that they aren't forced to try. They get lazy or sloppy (and in Intel's case, I'm not suggesting the engineers were the sloppy ones... more likely strategic decisions from management and quarterly earnings per share-focused execs). Eventually they are dethroned, and some ne…

> In their industry it takes years or even a decade to see the fruits of your effort.

So how long has Apple been working on this chip?

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#488

Post by spijdar »

Hamuko wrote: post #433
StreamBright wrote: post #384

Earlier quoted context omitted.

Absolutely not. We need more competition because the #1 reason we got to this situation is mono culture and a single platform (x86). We need Apple to succeed of creating an alternative ARM based desktop/laptop platform and for more competition we could add in Mips64 from China to the mix. I am really hoping that by 2025 we are going to have 3 major platforms available for end customers, so that there is real competit…

Isn't having a whole bunch of different processor architectures at the same time kind of bad for end-users?

This really depends. Once-upon-a-time, at least in the UNIX (tm) world, there were a plethora of ISAs, and this was the environment where ideas like Java really made sense. Write once, run anywhere.

Most OSes are still pretty well situated to handle this. Java remains, and is easily cross platform. I can run Java-based proprietary games like Minecraft on my POWER9 desktop, despite no-one involved probably ever considering ppc64le a valid target.

The CLR on Windows is also pretty easily cross-platform, although it won't help legacy x86 PE executables. Apple has solved this for ages on the tooling side, encouraging production of "fat" binaries with many arches since OS X was NeXT, and your .app packages needed to run on x86 + m68k + SPARC + PA-RISC.

Emulators like Rosetta (and qemu's usermode emulation) can fill the gap of legacy executables, while these other technologies can make the end-user experience good. Of course, that's only if a) someone writes your platform's equivalent of Rosetta, and b) developers write crossplatform apps.

So, the answer depends on how cynical or optimistic you are :-)

Re: 2020 Mac Mini – Putting Apple Silicon M1 To The Test

#489

Post by hajile »

mcintyre1994 wrote: post #429
arvinsim wrote: post #241

Earlier quoted context omitted.

As someone who owns both Mac and PC, I am excited on what Ryzen can offer on 5nm. If Apple has these gains, I am sure Ryzen will have great performance leaps too.

I’m not massively familiar with CPU architectures, would you expect to see similar performance gains going to 5nm on x86 as you do on ARM?

Not strictly because of 5nm itself.

5nm will be Zen 4 which should bring 10-20% IPC uplift if AMD's current trend continues.

TSMC's N5 5nm transistors are 85% smaller than their N7 transistors which should lower power consumption significantly though SRAM only shrinks a modest 35% (this especially affects desktop Ryzen with tons of cache compared to their laptop versions).

AMD currently makes the Zen 2/3 IO die on Global Foundries 12nm for contractual reasons. When they finally shrink that to 7 or 5nm, the power savings should be significant.

Zen 4 is expected to bring DDR5 support which will both drastically increase bandwidth and lower RAM power consumption. Likewise, it is expected to support PCIe 5 which doubles the bandwidth per lane to a little shy of 4GB/s.

All of these things together could mean a decent improvement in IPC and total performance and a very big improvement in performance per watt.

Meanwhile, I suspect we'll start seeing large "Infinity Cache" additions to their APUs that is shared between the CPU and GPU as the bus width of DDR just doesn't offer the bandwidth to keep larger GPUs from fighting the CPU for bandwidth. This should not only improve APU total performance, but fewer trips to RAM has a significant effect on power consumption (it costs more to move 2 bytes than to add them together).

Post reply on HN