Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

gadget.guy

macrumors 6502
Original poster
Oct 7, 2011
346
159
Yorkshire
Just bought the OWC i now have to decide on the SSD its probably going to be a 4TB outside chance of 2TB but i know i'll swap it out if i do.

Which SSD did you buy to go along with it how fast is it and what are the temps ?

relative newbie to enclosures but from what i've read the OWC express is a decent one i did read that some SSD are double sided which can cause slow downs in data transfer.

Any recommendations appreciated
 
thanks for the reply

Not one thats on my list seems cheaper than the rest what speeds are you getting via black magic if you don't mind me asking ?

The ones i'm currently interested in are

samsung 990 pro 2tb/4tb

crucial t500 2tb/4tb

wd black sn850x 4tb
 
4TB Samsung 990 Pro. Waiting for my backordered M4 Mini to arrive so I can't test the speed. The OWC dumbs down to Thunderbolt 2 speeds on my 2018 Mini.
 
I have the 1M2 with the Samsung 990 Pro 4TB paired with my M4 mini. I get 3.1 to 3.2 GB/s read and write using 1GB and 5GB test files.
 
  • Like
Reactions: gadget.guy
thanks for the replies i've been told the crucial T500 can cause issues as its double sided ? so sequential speeds can suffer i'm not sure about that though maybe i've misunderstood.

i'm probably still edging towards the samsung 990 pro amazon did have it 199 pounds but that included 75 pounds cashback from samsung the criteria is it must be delivered and sold by amazon

its not just amazon though so will keep a look out at the usual places in case i can get a bargain.
 
Yes KC3000
45C under load is pretty good. Initial reviews of the 2TB KC3000 had it running pretty hot under max load. In fact, power draw was one reason I chose the Samsung 990 Pro over the KC3000, but if you're only getting up to 45C, then power draw is less of a concern.

YXee827dnAEJG3vLcKT6FC-1200-80.jpeg

However, given that these drives can never max out their performance in these USB 4 enclosures, perhaps they run cooler than expected.

I'm curious. Do you have a way to measure power draw under load from your enclosure? For my Qwiizlab + 4 TB Samsung 990 Pro, the max draw is 7.5 to 8.0 Watts combined. (My Watt meter only displays in 0.5 W increments, so it's probably somewhere in between 7.5 and 8.0 Watts.)
 
The harsh reality about these "fast" drives is that you are measuring only the initial disk cache which is high speed. After a while of long read / write the cache is filled and the drive reduces to the actual disk speed, which may be about 1200 MB / sec. Some of the caches can be 300 GB large, so most people don't realize this is happening.

The only solution that is effective here is to RAID 0 two smaller drives in separate 40gbs enclosures. So when the performance hit happens it's still getting double the speed than just one. People warn that it doubles the chance of it failing, but the chance is so low it goes from a 1% to a 2% chance. Anyhow everything is backed up these days.
 
WD_BLACK SN850X - 2TB
Samsung 990 Pro - 4TB
Amorphous_OWC-1M2_WD-850X_2TB.png
Amorphous_OWC-1M2_Samsung-990Pro_4TB.png

Blackmagic_OWC-1M2_WD-SN850X_2TB.png

Blackmagic_OWC-1M2_Samsung-990Pro_4TB.png


As you can see, they’re practically the same. Speaking of practically the same, running the 64GB test in Amorphous didn’t hinder either drive. Honestly, both of these SSDs are overkill for even this enclosure but, eh, I got both drives on discount — 990 Pro during the recent BF/CM sales.
 
Last edited:
The harsh reality about these "fast" drives is that you are measuring only the initial disk cache which is high speed. After a while of long read / write the cache is filled and the drive reduces to the actual disk speed, which may be about 1200 MB / sec. Some of the caches can be 300 GB large, so most people don't realize this is happening.
For a 4 TB drive, the cache might be well over 600 GB, at least initially. So for most users here, that's sufficient.

Remember, internal drives have the same behaviour. They don't necessarily run at 3000 MB/s or 7000 MB/s max all the time either, yet people work just fine with them.

The only solution that is effective here is to RAID 0 two smaller drives in separate 40gbs enclosures. So when the performance hit happens it's still getting double the speed than just one. People warn that it doubles the chance of it failing, but the chance is so low it goes from a 1% to a 2% chance. Anyhow everything is backed up these days.
You're suggesting this is a major problem, but for most people it is not. And yes, software RAID in separate USB enclosures is a great way lose data. Definitely not recommended, for the most users. And in this case because we're not talking about two drives in the same fixed enclosure. We are talking about two separate USB drives which can get inadvertently unplugged at any time. That said, you can buy dual NVMe enclosures for RAID if you really want RAID, but for the vast majority of the population, it is complete overkill and added risk for little benefit.
 
well i decided on the samsung pro 990 mostly due to the cashback samsung now offer here in the uk and the discount i can get off it

I've set it up all went well i'm not getting anywhere near the figures your getting but i am using it ex fat

ive seen people mention that using ex fat can reduce speeds by 8-10%

My speeds are 3100 write and 3050 read which i'm happy with
 
My speeds are 3100 write and 3050 read which i'm happy with
That’s still plenty good.

FYI, the speeds were lower when connected to my M1 Mac mini (with 512GB internal SSD).
full

I’m not sure if the TB5 controller and/or internal SSD speed of the M4 Pro is what gave a boost in performance.

Additionally, you’re within reach of the 1M2’s marketed performance:
OWC said:
  • Ultra-fast: up to 3151MB/s real-world performance1
 
well i decided on the samsung pro 990 mostly due to the cashback samsung now offer here in the uk and the discount i can get off it

I've set it up all went well i'm not getting anywhere near the figures your getting but i am using it ex fat

ive seen people mention that using ex fat can reduce speeds by 8-10%

My speeds are 3100 write and 3050 read which i'm happy with
Yes, exFAT will reduce speeds on macOS. However, more important for me is that I find exFAT a bit problematic on macOS. I know some people have good results with it, but it doesn't play nice with certain macOS features (eg. exFAT is not supported for the Apple Photos Library), and I've encountered errors from time to time with exFAT on USB flash drives, so I've stopped using it. YMMV.
 
Last edited:
You need to get a drive with DRAM if you want run the OS or projects off it and let the M chip worry about other things and not caching the drive.
 
Hi, just received my OWC Express 1M2 today and I installed my Samsung 990 Pro 4TB and connected it to my Mac Mini M4 Pro. Power consumption (based on iStats) went up quite a bit (total power: ~6W --> 15W).
 
I just bought another 1M2 and I installed a WD_Black 8TB SN850X. Working fine so far
Any cooling concerns? Being double-sided, how does it dissipate heat from the side facing the PCB?
 
Last edited:
Put a WD850X in a 1M2 and I ran the black magic test. I'm not getting 3000 reads and writes like others are. More like 900-1000.

I transfered 100GB to it on the M4 Pro and it was the same speed as my Samsung T7. It actually slowed down to 140MB/sec speeds wtf? Sounds like my mac doesn't have TB4. Maybe I got scammed by Apple.

Either TB4 is way overhyped Or I'm wondering if my 1M2 could be defective.
 
Put a WD850X in a 1M2 and I ran the black magic test. I'm not getting 3000 reads and writes like others are. More like 900-1000.

I transfered 100GB to it on the M4 Pro and it was the same speed as my Samsung T7. It actually slowed down to 140MB/sec speeds wtf? Sounds like my mac doesn't have TB4. Maybe I got scammed by Apple.

Either TB4 is way overhyped Or I'm wondering if my 1M2 could be defective.
That's USB 3 speed. Are you using the proper cable? Either the original cable or else a certified USB 4 or certified Thunderbolt 4 cable.

Also, if you're transferring data in a file folder, often the speeds will be much slower depending upon the file mix in that folder. The 3000+ speeds are in synthetic benchmarks. Furthermore, with extended writes, the data speed will drop after a while when the write cache is exhausted. Interestingly, that speed after the cache is exhausted is right around 1000 MB/s.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.