I mean, pretty much all other benchmarks giving away some things regarding singlecore performance don't see any Hac that much in front of the M1s.Maybe, but it does match up with 5950x comparison under windows with cross platform plugs under Studio One. I'm wondering a little about these results, though. I don't personally need any numbers above 64, but you can keep going up in buffer/block sizes if desired. The main settings that are important are that the track is monitored/armed (has stereo IO assigned) green Z is on, the power buttons on the plugins are green, the playback protection is set to medium and then report how many instances of room reverb you can run at the different block (buffer) sizes before the CPU goes into the red.Īn so on, but with real numbers. I know S1 isn't AS native, and it will be worth trying this again when it is, but I'm curious how it performs under R2 as well. The same as the space designer test posted earlier, but in S1. It's just a bunch of "room reverbs" in series, with the first track input monitored and the point being to try to test low latency input monitoring performance. i will post back if i am waiting for a 16gb to arrive or just holding fast with this 8gb and waiting as well for a Big Dog !!!Ĭool to hear everything running well so far.Ĭan someone with studio one do me a favor and try this little benchmark test out for me? the Buffer issues i am having i think are from the PA Mastering Desk, the Glue Running oversampling. I'll probably get 32g assuming it's an option, if for no other reason than this next machine will be something I'll lean on for maybe 3 years. then shift it all a M? 16" MacBook Pro primarily for the larger screen and a 2tb internal. This M1 will be a good little buddy for about half my music work for next 6 months. I'm definitely pleased that I did get the 1tb so my most used libraries are always handy without hooking up the external. My gut tells me that with the new memory architecture, 8g might have actually performed well enough. I still have a lot to do in pushing this M1 with stuff from iZotope.Īs to whether I needed the 16g upgrade? Hard to say since I can't run anything head to head with an 8g M1. windows more fluid and smooth.and the case gets only mildly warm compared to a very hot (and loud) i9. The MOST notable difference is the "pleasure factor" in HOW the the M1 is handling an identical project. I have been running identical projects on a 15" 2018 MacBook Pro (32g/2tb, 6 core i9), and the M1 can handle a few more VI-laden tracks. This setup has been behaving and performing very well for me. (OWC TB3 Envoy Express enclosure with Samsung 970 EVO Plus M.2). and I have installed everything from Komplete 13 Ultimate with the libraries split between the M1's internal, and a 2tb external. The only thing from NI not working for me is Massive X, which is to be expected since it relies on AVX. My approach to using it so far has only been in Logic with a heavy dose of VI from Native Instruments. The M1 MacBook Air (16g/1tb) I've had for a few weeks now has completely lived up to my expectations. i was getting SWOD when opening a waves plugin. Waves had an update that helped a little with stability. lots of overloads due to buffer size, but no crashes. i have seen some quirky things happen in Logic last few days. OneDNN ( Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU) at 69x, oneDNN ( Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU) at 60.338x, oneDNN ( Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU) at 13.27x, oneDNN ( Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU) at 13.258x, oneDNN ( Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU) at 13.062x, oneDNN ( Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU) at 10.522x, oneDNN ( Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU) at 10.458x, oneDNN ( Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU) at 10.416x, Cryptsetup ( Serpent-XTS 512b Encryption) at 10.181x, Cryptsetup ( Serpent-XTS 256b Encryption) at 10.16x.Hey Jim, checking in to see how your testing of the macmini is going. The results with the greatest spread from best to worst included: Geometric Mean, More Is Better Geometric Mean Of All Test Results Result Composite AMD Ryzen 7 5800X 8-Core Apple M1 14 28 42 56 70 62.03 34.24ĪMD Ryzen 7 5800X 8-Core had the most wins, coming in first place for 85% of the tests.īased on the geometric mean of all complete results, the fastest (AMD Ryzen 7 5800X 8-Core) was 1.811x the speed of the slowest (Apple M1).
0 Comments
Leave a Reply. |