When recording 32 stereo tracks with 2 Revelation Reverbs my real-time DSP usage is 15-20% lower than previous versions. This is using a 64 sample buffer on my MacPro 5.1 twelve core MacPro. I’m using a Focusrite RedNet PCIe Card with Rednet converters and Dante running Sierra. Works great! From what another user stated and from what I’ve seen when mixing at high latency the DSP usage also seems lower. Well done Steinberg!
I feel better CPU usage on the this update too…
That’s great! I wish I was experiencing that on my Mac. 
Feels about the same here with CPU usage.
No difference for me unfortunately.
It feels a tiny bit better once I trashed prefs. I wonder what the difference is between those who see no to very little improvement vs. those who see a lot? In the VI_Control forum there are composers like myself with huge sessions and VE Pro that are experiencing much better CPU utilization in 10.0.30.
I think it’s the different versions of OSX. I’m on Sierra. Also, do this. TRY SIGNING OUT OF ICLOUD. THIS LOWERED REAL-TIME DSP USAGE ABOUT 20% AT LOW BUFFERS. I did this months ago and it really helped. Seems that iCloud is a real DSP hog for real-time Audio.
I tried this, thanks for the suggestion. No dice, nearly the same CPU usage (I’m on Mojave). I guess I’m stuck with the lesser former CPU efficiency unless Steinberg support writes back and can help me specifically get there with my very-up-to-par Mac Pro (they rarely write back to begin with, so forget that option I’m afraid to say). So I’m stuck with not-really-improved CPU efficiency until I can figure out why on my own, or until I’m so tired of it that I go to another DAW> 
I’m OK with my system needing to be adjusted to make this happen – just need to figure out what it is I need to change/adjust.
Mmm, this is a really interesting statement, never heard of that.
Thank you, i am gonna check this out !
BTW: I am on OSX Mojave 10.14.6 beta.
Kind Regards.
HERES ALL THE THINGS I TRIED THAT SEEMED TO MAKE NO NOTICABLE DIFFERENCE (perhaps it was so slight I couldn’t tell) maybe they will on a different OS or system
Trashed preferences
Restart
Reset P Ram
Reset SMC
Removed all USB except for keyboard, mouse, dongles and USB to SSL Matrix Mixer
Disabled all networks
Tried all different VST Audio settings in Studio menu (ASIO guard and all the others on that page)
Selected the Mac internal Audio I/O in studio menus instead of Focusrite PCIeR Audio card
Change check mail to manual
Deleted network printers
Turned off WiFi
Turned off printer sharing
Turned off Bluetooth
Disabled time sync and auto time zone
Disabled notifications
Turned off location services
Turned off Mission Controller hot keys
Removed FireWire and Bluetooth in Network Pref setting
Disabled Auto updates in App Store
Removed all unneeded Login items
Removed Startup items
Disconnected additional monitors
How were you measuring things to determine if you’re getting an improvement. I did a fairly methodical test, including a comparison with other daw, under 10.0.20 and Cubase 10.0.20 performed so terribly worse then all the other DAW’s that I stopped using it! After updating to 10.0.30, it now competes with the other DAW’s very well.
Specifically with cubase under 10.0.20 i noticed that for the first 10-20 tracks, everything seemed more or less normal, compared to the other DAW’s, but as the tracks added up, the CPU started bogging down. You could notice it at 50 tracks and at around 75 tracks it would not even play back the project anymore. Under 10.0.30 I can easily play the whole project with lots of room to spare in the CPU and on average its using about the same amount as the other DAW’s.
Here was my test: DAW Performance Test Results | VI-CONTROL
I think if you are going by how it “feels” you may not notice any difference. I was not measuring perceived sluggishness or GUI related performance. I was simply measuring how much CPU it needed to play back a 90 track orchestral score with VSL instruments. 10.0.20 could not do it. 10.0.30 can do it with a lot of room to spare. Hugely improved.
I was focused on low latency recording (32 or 64 samples). I simply watched to see if the DSP meter overloaded when recording about 40 stereo tracks with a couple of reverbs. I’ve never had an issue with playback using high latency settings and I don’t use virtual instruments.