Hmm, so my system isn’t too too bad, I was worried something is wrong with my configuration.
Coming from first generation i7-920 I was expecting that these new CPUs will need hundreds of heavy tracks to overload them
You have got 2 more cores and a littles extra clock over your 720
About 50% increase on what you had.
I would be worried why on some of your posts it shows cpu at 21% yet Cubase maxed out
That’s a bad set up for what you need
Your target should be imo as follows
If really low latency is your goal you need fast processing which tends to mean small number of cores running fast.
For raw DAW power you need lots of core which tends to run slower both due to heat and core/thread synchronisation issues
Decide which is you goal and purchase the right basic machine for that task.
Either way your getting the balance right when your getting the Cubase performance meter showing very high on the non real time part of the meter and head room on the real time part and CPU is 90% or more at the same time on all threads/ cores with clean output and a still swiftly responsive graphics, mouse and other controllers at the minimum latency you can work with.
The pic is with 32 bit float Processing Precision and 128 samples buffer size. I tryed play with buffer at 64, 32 and 16, and work well. With buffer at 8 impossible. With 64 bit float Processing Precision is imposible to play with buffer at 128 (I need at least 256).
Wowowowoww. I try the same with Cubase Pro 9.5 and the difference is dramatic. This version work hard better! The pic with the same configuration, 32 bit and buffer at 128. With Cubase 9.5 I can play with buffer at 8 with no problems. I haven’t tryed with 64 bit float Processing Precision.
It’s difficult to tell given the screenshots are all different but how many of the Tracks playing are Rec Enabled?
Some of the screenshots seem to suggest differing number of Rec Enabled tracks which will change the number of tracks actually causing Real Time load if ASIO Guard is active.
That would be comparing Apples and Oranges, and indeed multiple Rec enabled tracks would be unusual.
Just thought I would makes this suggestion if indeed folks are looking to make comparisons.
I ran the project as is. I only loaded the performance monitor and hit play. If there are differences in each test than it is the project doing it, though I don’t see how.
Remember that the ASIO Meter only tells a roughly what performance is used.
I have had projects that show 75% utilization and I could double the tracks and plugins, now only showing 5% more.
That is of cause a rather unusual behavior, but it has happened a few times.
