![]() Note that Blackmagic is normally run with ‘stress’ set at 5 GB. While writes stabilise at just under 4 GB/s, even at 64 MiB, reads are continuing to increase in speed, at just under 10 GB/s. Results for my M1 Mac mini, with a 500 GB internal SSD, are very different. Once read and write sizes exceed about 4 MiB, transfer speeds settle at around 2.8 GB/s for reads, and 3.4 GB/s for writes. Before showing those from an M1 model, this is a typical set of results from my iMac Pro: I’m very grateful to Jay, who alerted me to different figures being reported by ATTO Disk Benchmark (available on request from ATTO). Blackmagic Disk Speed Test, available from the App Store, is now a Universal App, but returns read and write benchmarks of between 2.7 and 3.5 GB/s, which are very similar to those of a T2 model. One clear conclusion for the moment is that relative performance of M1 Macs depends on which benchmark you use. Their one crucial performance area which isn’t as clear is their internal storage: I’ve seen a range of benchmark results which either set them on a par with T2-equipped Macs like my iMac Pro, or just seem puzzling. I do use Win10 in a VM, I run a database program there that accesses many image files, all of this will be on the NVME, that program desperately needs it.There now seems no doubt that, by any standards or comparisons, Apple’s first Apple Silicon Macs, its M1 models, are wickedly quick. The goal was to prove up the MP, then shuffle drives around data SSD -> NVME and OS HDD -> SSD. Once I load data into RAM, it's very RAM and CPU (parallel) intensive, the HDD doesn't matter much. Note that I moved to this machine as a dual CPU trial, moving from a 3.7GHz i7 with 16 GB to the dual 3.33 with 64 GB and my workflow is much faster, even with the HDD. I wouldn't expect the HDD to be an issue here at all. I expected at least some gain here, since Mathematica reads and stores data in RAM, so I was assuming something else was happening. It was the same down to the hundredth of a second (16.43s). For the binary read operation, I used a program called Mathematica and measured the time it took to load the 1.5 GB file with a built in function. I knew there was more to it than read speed, but to take the same exact amount of time is surprising. And then you actually might start noticing the diff in speed between SSD and NVMe when it comes to opening big files Even a shitty cheap SSD will make a huge difference. Get the OS onto an SSD (NVMe or SATA) and you'll feel it :)ĮDIT: God, I've just realised - you say the OS is on a spinning drive? That's the damn bottleneck. ![]() ![]() It's not about the data throughput, it's about the latency. Now shifting the OS from HDD to SSD - that'll feel huge, because it's a lot of reads of lots of files, and while the machine still has lots it has to do with the data, so damn many files get read that the machine probably spends a fair bit of time waiting for the drive head to move to the right place. ![]() Loading images, even large ones, isn't really stressing the drive, but things like caching, video rendering, general file operations - they may well seem a tad faster. The new drive will be having a positive effect but it'll vary depending on what exactly it is the machine's doing. It's not that the drive is slower, it's that the machine as a whole has much more to do when reading the file than purely streaming the data off the drive, so the increase gets lost in the mix. But the machine needs another minute, pretty much, to actually process the data and present it on screen. The old drive could read the file itself in about 4 seconds, the new drive can read it in half a second. I know I need a newer firmware if I want to boot from NVME, but do I need the newer firmware to take advantage of the NVME speeds? Why does the disk test so fast but function as slow as the SSD I'm replacing? Thanks in advance!Ĭlues: old drive = 275MB/s, new drive = 2.5GB/s but opening a 1GB TIFF file takes the computer around a minute. Lastly, I loaded some big binary files, both drives loaded a 1.5 GB file in about 16 seconds. Using the 3rd party FileViewer app it took 32 second regardless of the drive. A 1 GB tiff in preview took right at 1:05 to open from either drive. The NVME RAID drive tested around 2,500 MB/s read.Īll awesome! Except when I started testing with my actual data, they're exactly the same speed. The SSD is an OWC Mercury 3G unit and it tested around 275 MB/s read. I used AmorphousDiskMark to measure the drive speeds. It formatted to RAID 0 perfectly and I was able to copy the data from the SSD to the NVME drive without issue. I recently added a Sonnet 4x4 M.2 NVME card with four, Samsung EVO 970s (500 GB each). Currently my OS and some of my data is on the original HDD, but I have a separate SSD that holds my bigger files (many are 1GB+ each). The bootrom is 00 (keep reading, I don't want to boot from NVME). I have a 2009 MP 4,1 flashed to 5,1 running High Sierra (10.13.6).
0 Comments
Leave a Reply. |