Samsung 970 EVO M.2 NVMe SSD

[NazariyUA]

I just installed Mojave with this drive using APFS.
My specs are:
Gigabyte Z370 HD3P(using the m.2 port)
Corsair Dominator 16GB(2x8)
Intel i5 8600K @ stock
Evo 212

My read speeds fluctuate around 3K, and my write speeds fluctuate around 2,500.
Running `system_profiler SPSerialATADataType | grep 'TRIM'` returns `no`. However, I read on the forums trim support is enabled by default. Also boot speeds are relatively quick--almost the same as windows.

Also in this picture, trim support is enabled. So it looks like it comes enabled by default.

 

[wstrohm]

Does the NVMe Samsung 970 EVO/PRO boot OK using Clover and having chosen it as the #1 boot preference in the UEFI (BIOS) firmware? And what apps (other than "disk speed test" benchmark apps) show improved performance? CCC? Any other advantages or disadvantages? Does TRIM work?

 

[pastrychef]

Does the NVMe Samsung 970 EVO/PRO boot OK using Clover and having chosen it as the #1 boot preference in the UEFI (BIOS) firmware? And what apps (other than "disk speed test" benchmark apps) show improved performance? CCC? Any other advantages or disadvantages? Does TRIM work?

Yes, the 970 EVO boots Clover fine.

AJA System Test is another benchmark you can use.

Carbon Copy Cloner works amazingly fast when cloning to and/or from NVMe.

Try unRARing a big file on an NVMe vs SSD or mechanical hard drive and you will clearly see how big of a difference there is.

I believe Trim is enabled by dafault on APFS partitions. If you are on HFS+, you may have to enable Trip manually or with a KextsToPatch.

 

[wstrohm]

Thanks much for that information. My guess is that required time of a CCC backup from the 970 to a SATA III SSD would be governed by the sequential write speed of the SATA SSD. Whereas a "disk speed test" app, or "unRARing" a file, reads and writes to the same storage device, most copy operations involve more than just the source, so it seems to me that the slower of the two devices would determine the transfer rate. Is that right or wrong?

 

[pastrychef]

Thanks much for that information. My guess is that required time of a CCC backup from the 970 to a SATA III SSD would be governed by the sequential write speed of the SATA SSD. Whereas a "disk speed test" app, or "unRARing" a file, reads and writes to the same storage device, most copy operations involve more than just the source, so it seems to me that the slower of the two devices would determine the transfer rate. Is that right or wrong?

For copying, it depends on the size of the files. But, generally, I think the bottleneck is in the write speed of the destination.

 

[NazariyUA]

For copying, it depends on the size of the files. But, generally, I think the bottleneck is in the write speed of the destination.

Yes, if its a small, the speeds will appear low, however if its a large file, it will be close to the rating. I transferred a 5gb file from one place to another on the drive in around 5 seconds. I also installed it directly from a USB installer, I didn't clone my drive.

 

[wstrohm]

FWIW, below is a comparison between Black Magic Disk Speed Test results for my Crucial M500 500 GB SATA III SSD and my newly-installed (today) Samsung 970 PRO.

 

[NazariyUA]

FWIW, below is a comparison between Black Magic Disk Speed Test results for my Crucial M500 500 GB SATA III SSD and my newly-installed (today) Samsung 970 PRO.

View attachment 362256 View attachment 362257

How does the Pro differ from the EVO, more writes thus longer lifetime?


Those are my speeds on EVO 970.

 

[wstrohm]

How does the Pro differ from the EVO, more writes thus longer lifetime?

I have read that the EVO = PRO up to a transfer maximum of 22 GB, and from there on up the PRO is faster. Don't know about lifetime.

 

[MAC_OS]

unfortunately I got only very slow speed on my new Samsung 970 Pro NVMe m.2
1460 read - 1480 write

I install NVMe SSD direct on M.2 slot, should I really buy PCI-e M.2 Adapter or someone got also success on M.2 slot?

My system: GA-Z170X-UD5 TH / i7-7700K