Completely baffling and amazing CPU behavior with Bitwig 3!
-
machinesworking machinesworking https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=8505
- KVRAF
- Topic Starter
- 6214 posts since 15 Aug, 2003 from seattle
Bear with me, I know I'm the dork posting about this on Saturday night but I've never seen anything like this!? It would be cool to hear from Bitwig on this!
In the past Bitwig scored as Badly as Live in any CPU stress test I did with it, scoring roughly 70% of what Logic, DP, and Reaper can get, track count wise, with plug ins like U-He Diva etc. Over the last 15 years the most reliable and consistent CPU test to see about track count and general performance for your DAW and your PC or Mac has been a "to failure" test.
Simply put, load a CPU pig like Diva on a track, do a one note or chord run of eighth notes (both have relevance), copy the MIDI and Diva track as many times as you can before you hear audible glitches, back off and that's the amount that DAW can handle on your system. Typically this is consistent, but not anymore. I mean, not even close!
Bitwig tested with Diva on my 2012 2.7Ghz 16GB RAM Macbook Pro is like before, terrible, 4 instances of Diva compared to 8 in DP10.
Bitwig on my heavily modified 09 Mac Pro with 12 3.33ghz cores, 24GB RAM, is scoring and I'm not making this up, 36 instances of Diva!
(in comparison DP10 is scoring only 11 tracks on the Mac Pro, it seems to be able to use the virtual cores on the macbook but not the Mac Pro?)
For the most part other DAWs like Logic, DP and Reaper are consistent with their performance but Bitwig Loves the old Mac Pro, just adores it! The two areas where it beats the macbook clearly are in graphics processing and CPU count, 12 cores as opposed to four, and the Mac Pro has a Radeon RX 580 8 GB graphics card.
No other DAW is showing these wild numbers and you can clearly see Bitwig runs each instance of Diva as a separate process in Activity Monitor on OS X, so it's not cheating somehow, all instances are using CPU etc.
Why is it so terrible on the Macbook Pro and so good on the Mac Pro? All other DAWs on my system are showing more like you would expect with the power and years difference between the two, like a little over half the power of the Mac Pro with the Macbook Pro, but not Bitwig?
In the past Bitwig scored as Badly as Live in any CPU stress test I did with it, scoring roughly 70% of what Logic, DP, and Reaper can get, track count wise, with plug ins like U-He Diva etc. Over the last 15 years the most reliable and consistent CPU test to see about track count and general performance for your DAW and your PC or Mac has been a "to failure" test.
Simply put, load a CPU pig like Diva on a track, do a one note or chord run of eighth notes (both have relevance), copy the MIDI and Diva track as many times as you can before you hear audible glitches, back off and that's the amount that DAW can handle on your system. Typically this is consistent, but not anymore. I mean, not even close!
Bitwig tested with Diva on my 2012 2.7Ghz 16GB RAM Macbook Pro is like before, terrible, 4 instances of Diva compared to 8 in DP10.
Bitwig on my heavily modified 09 Mac Pro with 12 3.33ghz cores, 24GB RAM, is scoring and I'm not making this up, 36 instances of Diva!
(in comparison DP10 is scoring only 11 tracks on the Mac Pro, it seems to be able to use the virtual cores on the macbook but not the Mac Pro?)
For the most part other DAWs like Logic, DP and Reaper are consistent with their performance but Bitwig Loves the old Mac Pro, just adores it! The two areas where it beats the macbook clearly are in graphics processing and CPU count, 12 cores as opposed to four, and the Mac Pro has a Radeon RX 580 8 GB graphics card.
No other DAW is showing these wild numbers and you can clearly see Bitwig runs each instance of Diva as a separate process in Activity Monitor on OS X, so it's not cheating somehow, all instances are using CPU etc.
Why is it so terrible on the Macbook Pro and so good on the Mac Pro? All other DAWs on my system are showing more like you would expect with the power and years difference between the two, like a little over half the power of the Mac Pro with the Macbook Pro, but not Bitwig?
-
machinesworking machinesworking https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=8505
- KVRAF
- Topic Starter
- 6214 posts since 15 Aug, 2003 from seattle
OK in subsequent tests on the Macbook Pro, whatever the intermittent spiking was that messed up Bitwig is gone, and it's routinely scoring 14-15 instances on the Macbook Pro here. Something before was messing up it's pre-rendering or buffering etc. that all DAWs do, that Bitwig seems to be jumping ahead of the pack with.
-
machinesworking machinesworking https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=8505
- KVRAF
- Topic Starter
- 6214 posts since 15 Aug, 2003 from seattle
To be clear, I'm a skeptic as well, but the numbers don't lie.
Bitwig are doing proper multithreading on both my old Mac Pro and old Macbook Pro. I'm certain the numbers wouldn't be surprisingly different for more modern macs or PCs. These numbers represent single instances of Diva with the INIT patch, playing an eighth note four bar loop at 120 BPM with a 256 buffer in each DAW.
09 Mac Pro 12 core 3.33 Xeon CPU use
Bitwig - 36
Live 10 - 9
Logic - 10
DP10 - 11
Reaper - 11
2012 4 core 2.7 i7 Macbook Pro
Bitwig - 14
Live10 - 7
DP10 - 9
Reaper - 15
Bitwig are doing proper multithreading on both my old Mac Pro and old Macbook Pro. I'm certain the numbers wouldn't be surprisingly different for more modern macs or PCs. These numbers represent single instances of Diva with the INIT patch, playing an eighth note four bar loop at 120 BPM with a 256 buffer in each DAW.
09 Mac Pro 12 core 3.33 Xeon CPU use
Bitwig - 36
Live 10 - 9
Logic - 10
DP10 - 11
Reaper - 11
2012 4 core 2.7 i7 Macbook Pro
Bitwig - 14
Live10 - 7
DP10 - 9
Reaper - 15
-
- KVRAF
- 11194 posts since 2 Dec, 2004 from North Wales
That is very interesting. You would think the Dev's would be making a bigger deal about this!
X32 Desk, i9 PC, S49MK2, Studio One, BWS, Live 12. PUSH 3 SA, Osmose, Summit, Pro 3, Prophet8, Syntakt, Digitone, Drumlogue, OP1-F, Eurorack, TD27 Drums, Nord Drum3P, Guitars, Basses, Amps and of course lots of pedals!
- Banned
- 11467 posts since 4 Jan, 2017 from Warsaw, Poland
Positive Bitwig topic from @machinesworking??
I must be still sleeping...
I must be still sleeping...
-
- KVRian
- 503 posts since 1 Jul, 2009
I did a test with the same settings in Bitwig, under Windows 10, on a i7-4771@3.5GHz (4 cores old CPU but still powerful) and barely can get 7 Diva's running without buffer under-runs. I had to enable multicore on the last 2 Diva's. CPU is at ~95%machinesworking wrote: ↑Sun Jan 05, 2020 10:25 am These numbers represent single instances of Diva with the INIT patch, playing an eighth note four bar loop at 120 BPM with a 256 buffer in each DAW.
Your numbers are impressive.
- 4 core 2.7 i7 Macbook Pro - Bitwig - 14 ... HOW?? What CPU exactly?
- Mac Pro 12 core 3.33 Xeon CPU use - Bitwig - 36
-
machinesworking machinesworking https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=8505
- KVRAF
- Topic Starter
- 6214 posts since 15 Aug, 2003 from seattle
So with the 2012 Macbook Pro (2.7 GHz Core i7 (I7-3820QM)), the first time I tested it, the CPU was audibly spiking at regular intervals. I wish I had screenshot it?! I did screenshot the results with 14. This is why I mentioned that maybe this is the reason they aren't bragging out right about this? I'll test it again a third time and see which way it goes, but it does seem to be some random spike that made it drop the ball the first time. What does the CPU meter show in Bitwig when you test it?anoise wrote: ↑Sun Jan 05, 2020 1:03 pmI did a test with the same settings in Bitwig, under Windows 10, on a i7-4771@3.5GHz (4 cores old CPU but still powerful) and barely can get 7 Diva's running without buffer under-runs. I had to enable multicore on the last 2 Diva's. CPU is at ~95%machinesworking wrote: ↑Sun Jan 05, 2020 10:25 am These numbers represent single instances of Diva with the INIT patch, playing an eighth note four bar loop at 120 BPM with a 256 buffer in each DAW.
Your numbers are impressive.
- 4 core 2.7 i7 Macbook Pro - Bitwig - 14 ... HOW?? What CPU exactly?
- Mac Pro 12 core 3.33 Xeon CPU use - Bitwig - 36
With the Mac Pro 36 instances, you could watch the CPU readjust as you add instances, and Bitwig was flatly the only DAW that actually used all possible 24 threads on this old 12 core Mac Pro. Even Reaper simply didn't use the possible two threads per core. In the Activity Monitors CPU history for both OS X and Bitwig built in CPU history you could clearly see it working cleanly to get all the power it could out of the Xeon's.
-
- KVRist
- 421 posts since 11 Dec, 2002 from Los Angeles
Anyone able to benchmark Windows in the same way? I only have Bitwig these days.
-
- KVRian
- 1262 posts since 15 May, 2002 from Finland
I can try later. I asked the devs about this, BW does use as many cores as you can throw at it. I'd love one of those new Epycs if I was a millionaire Altough 64-core threadrippers are being released this year, I'll be interested to see them, since apparently the 2nd gen Threadrippers are much better for DAW use than the lackluster 1st gens.
-
- KVRian
- 503 posts since 1 Jul, 2009
The meter in Bitwig is almost full.machinesworking wrote: ↑Sun Jan 05, 2020 2:13 pm What does the CPU meter show in Bitwig when you test it?
I don't know how your laptop CPU can handle 14 instances of Diva, double than mine and according to cpubenchmark.net, my CPU is somewhat faster than yours. Double the performance on a weaker CPU is hard to believe.
I doubt that it is OS related, or that Bitwig and Diva are more efficient on a Mac, but who knows. We need more testers.
- KVRian
- 1295 posts since 7 Dec, 2017
I have to be doing this wrong. I'm getting 82 instances before it croaks.
-JH
- KVRian
- 985 posts since 10 Sep, 2014
I get 31 instances (2bar 8 note chord) without crackling on Ryzen 7 2700x. 32 if I turn off performance graph.
-
machinesworking machinesworking https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=8505
- KVRAF
- Topic Starter
- 6214 posts since 15 Aug, 2003 from seattle
The laptop 2.7 is supposed to have much better CPUs than than desktop, even if they're 2.7ghz VS 3.33ghz. It seems that each core is handling 3.5 instances on the laptop and 3 on the desktop, which backs that up, it actually adds up more or less correctly.anoise wrote: ↑Sun Jan 05, 2020 3:47 pmThe meter in Bitwig is almost full.machinesworking wrote: ↑Sun Jan 05, 2020 2:13 pm What does the CPU meter show in Bitwig when you test it?
I don't know how your laptop CPU can handle 14 instances of Diva, double than mine and according to cpubenchmark.net, my CPU is somewhat faster than yours. Double the performance on a weaker CPU is hard to believe.
I doubt that it is OS related, or that Bitwig and Diva are more efficient on a Mac, but who knows. We need more testers.
I know that CPU benchmarks are important but if I recall correctly, ( and there's plenty of people on here more informed than I am on this to correct me if I'm wrong ), it's all broken down into different types of operation and various CPUs can score better or worse on various operations etc. At that site you mentioned cpubenchmarks.com it scores the 3.33 at 8354 and the 2.7 at 8402, when clearly the 2.7 is a better CPU per core, just not at four cores VS 12.
-
machinesworking machinesworking https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=8505
- KVRAF
- Topic Starter
- 6214 posts since 15 Aug, 2003 from seattle
Cool!JHernandez wrote: ↑Sun Jan 05, 2020 6:03 pm I have to be doing this wrong. I'm getting 82 instances before it croaks.
Hey would both of you toss out more system specs? name and number of CPU, how many cores? RAM?