u-he

medienhexer · Post by **medienhexer** » Tue Dec 08, 2020 4:36 pm

So that ist 6/32 on the 2015 MBP vs. 32/192 on the M1. Okay, that just happened.

I really want to wait for an iMac with a little more of everything. Especially RAM, since I play with larger sample libraries a lot. But those are really impressive first numbers.

And in my case, that odd Diva lead melody or pad always comes last when the system is already at its limits. Having that much headroom and be able to keep latency low sounds like a dream come true.

Funkybot's Evil Twin · Wed Dec 09, 2020 12:32 am

Are there any other laptops that can pull of comparable performance? I mean, I've always been a hardcore Windows guy but holy crap, I can't imagine not buying an M1 MPB the second they announce they 16 inch models. I apologize in advance for all my future complaining about how a new Mac OS update broke something.

zvenx · Post by **zvenx** » Wed Dec 09, 2020 12:54 am

Funkybot's Evil Twin wrote: Wed Dec 09, 2020 12:32 am Are there any other laptops that can pull of comparable performance? I mean, I've always been a hardcore Windows guy but holy crap, I can't imagine not buying an M1 MPB the second they announce they 16 inch models. I'm apologize in advance for all my future complaining about how a new Mac OS update broke something.

Lol.
rsp

Urs · Post by **Urs** » Wed Dec 09, 2020 6:09 am

I think we have to take all of this with a grain of salt. As far as we know now, one can't add more RAM or upgrade pretty much anything, as speed, cost and power consumption are paid for by System-on-a-Chip design. In that sense, we're dealing with an entirely different architecture since terms like clock rate, data throughput and so on are opaque and may behave different to common systems. I mean, who knows how context sensitive all of this is?

For instance, a single Diva uses a lot more CPU than adding two more. We reckon that's because a single instance will still run on one of the low power cores. Likewise, we now might have the answer why Apple wants developers to add extra effort into multithreaded processes, which for Diva's multicore option requires us to publish context information and some form of group tags for voices. We think the poor multicore performance we observe is result of not having that, and thus having many voices render on slow cores. (Our native builds scale even worse for multicore, we guess that Rosetta generally palms stuff off to faster cores than native code... until we do that extra work... we hope...)

carrieres · Post by **carrieres** » Tue Dec 15, 2020 2:14 pm

tasmaniandevil wrote: Fri Dec 04, 2020 4:04 pm As written in the initial post, this was just a quick first test.
We compared it to the machine that was at hand, which was a 2015 MacBook Pro.
We will conduct lots more tests over time, but we simply don't have that many different systems with all kinds of CPUs we could compare it with (especially not during Covid times, when the majority of people is working from home office).

What is the cpu inside the 2015 MacBook Pro ?

zvenx · Post by **zvenx** » Tue Dec 15, 2020 2:32 pm

If only Tas had listed that on his first post (OP) here

And to have something to compare it to, we did the same test on an old, trustworthy MacBook Pro from 2015 with a 2.2GHz Intel i7 (this is starting to feel a bit like on an episode of Mythbusters).

rsp

carrieres · Post by **carrieres** » Tue Dec 15, 2020 2:36 pm

zvenx wrote: Tue Dec 15, 2020 2:32 pm If only Tas had listed that on his first post (OP) here
And to have something to compare it to, we did the same test on an old, trustworthy MacBook Pro from 2015 with a 2.2GHz Intel i7 (this is starting to feel a bit like on an episode of Mythbusters).
rsp

Thanks, I saw it after i wrote my post, i suppose it's an i7 4770HQ

zvenx · Post by **zvenx** » Tue Dec 15, 2020 2:39 pm

Welcome..
most likely indeed:
Details: This model is powered by a 22 nm, 64-bit "Fourth Generation" Intel Mobile Core i7 "Haswell/Crystalwell" (I7-4770HQ) processor which includes four independent processor "cores" on a single silicon chip. Each core has a dedicated 256k level 2 cache, shares 6 MB of level 3 cache, and has an integrated memory controller (dual channel).

https://everymac.com/systems/apple/macb ... specs.html

tasmaniandevil · Post by **tasmaniandevil** » Tue Dec 15, 2020 4:04 pm

Yep, exactly that.

Tronam · Post by **Tronam** » Mon Dec 21, 2020 9:45 pm

zvenx wrote: Tue Dec 15, 2020 2:39 pm Welcome..
most likely indeed:
Details: This model is powered by a 22 nm, 64-bit "Fourth Generation" Intel Mobile Core i7 "Haswell/Crystalwell" (I7-4770HQ) processor which includes four independent processor "cores" on a single silicon chip. Each core has a dedicated 256k level 2 cache, shares 6 MB of level 3 cache, and has an integrated memory controller (dual channel).

https://everymac.com/systems/apple/macb ... specs.html

A rough compute power comparison between the quad-core i7 2.2GHz and M1 MacBook Pro, at least according to GeekBench. Double the performance at 1/4 the power consumption. The comparisons seem to vary a bit though depending on what's being processed since the M1 appears to employ hardware acceleration for certain types of tasks.
.

2015 i7 vs 2020 M1.png

midi_transmission · Post by **midi_transmission** » Wed Dec 30, 2020 1:29 am

medienhexer wrote: Tue Dec 08, 2020 4:36 pm So that ist 6/32 on the 2015 MBP vs. 32/192 on the M1. Okay, that just happened.

I really want to wait for an iMac with a little more of everything. Especially RAM, since I play with larger sample libraries a lot. But those are really impressive first numbers.

And in my case, that odd Diva lead melody or pad always comes last when the system is already at its limits. Having that much headroom and be able to keep latency low sounds like a dream come true.

Thanks so much for sharing u-he.

What I don't understand is that the 2015 is so much slower. When I look up the cpu benchmarks of both, the M1 should be max. 4 times faster than the 2015 i7. Any idea why the difference is so unbelievable huge in this case? 6/32 vs. 32/192, that's 32x as much voices! And more than 5 times as much instances. Any possibility that there was a overlooked thing in the comparison, something like thermal throttling of the 2015 or something? Or do I maybe overlook something?

jacobh · Post by **jacobh** » Wed Jan 06, 2021 11:36 pm

I'd really love to see what the performance is on the Macbook Air M1 processor, which I know would be a little less. The Macbook Air M1 is most appealing to me because it doesn't need fans to cool it! Cooling fan noise has been a problem for me with many generations of laptops.

If a Macbook Pro M1 is as high performance with Diva as the specs you're reporting, then I think a Macbook Air M1 could likely still be high enough performance for me.

Any chance anyone could test on a Macbook Air M1?

Thanks!

carrieres · Post by **carrieres** » Thu Jan 07, 2021 9:41 am

https://www.apple.com/fr/mac/compare/?m ... ac-mini-M1
Every Computer have the same M1 CPU, so they have the same performance

Urs · Post by **Urs** » Thu Jan 07, 2021 9:44 am

The problem with the Air is exactly the lack of fans. It will therefore need to run more stuff on the slower cores, just like it already has one core switched off completely. I would expect this to have a rather great impact on plug-in performance.

We did not buy an Air to test things since we think it would not be of much use for us, we only bought the MBP and Mac Mini for our own purposes, hence could only test on these.

pdxindy · Post by **pdxindy** » Thu Jan 07, 2021 5:00 pm

carrieres wrote: Thu Jan 07, 2021 9:41 am https://www.apple.com/fr/mac/compare/?m ... ac-mini-M1
Every Computer have the same M1 CPU, so they have the same performance

The MB Air has one less graphics core enabled... and it does throttle back compared to the MBP and Mini

u-he

KVR Audio

Diva with Rosetta2 on Apple M1