I saw that, but it is an NVMe driver and it installs just fine on the SB. The drivers before it installs are listed as "Standard NVMe controller" and after they are listed as "Samsung NVMe controller". I think they say that to distinguish it from other retail models that a customer might have when they go to that site looking for drivers. It's the brand name of the only retail M.2 drive Samsung sells over the counter.
Samsung vs Toshiba SSD (now with poll)
- Thread starter flar
- Start date
I saw that, but it is an NVMe driver and it installs just fine on the SB. The drivers before it installs are listed as "Standard NVMe controller" and after they are listed as "Samsung NVMe controller". I think they say that to distinguish it from other retail models that a customer might have when they go to that site looking for drivers. It's the brand name of the only retail M.2 drive Samsung sells over the counter.
Ah, I see that the numbers they saw improvements on were on the AS SSD test that I haven't been running. Perhaps that test was broken by something in the Standard NVMe drivers that didn't really affect the CrystalDiskMark benchmark. I would characterize my results in CDM as a "small boost", though it is muddied by the fact that the numbers seem to be trending on a very shallow curve since I purchased the device.
I could speculate a lot but that's not really the point... perhaps they contacted Samsung or Samsung contacted them and told them to use that driver to test with although not the approved official driver... IDK they didn't tell us. SOMETHĪNG happened though because they don't just print retractions/corrections without a catalyst. However, that doesn't mean they would be forthcoming with all the particulars either.
Back to the Surface Book test results vs the SP4 core m results ... *maybe* the results are that much better on the SB because it was an i7 vs a Core m, in which case having a faster drive on the Core m wont matter and the results were already much better than the SP3 with the SB and might improve still more with the eventual tuning that always happens post release with MS. IIRC the Passmark disk test I ran had the SP4 core-m version slightly ahead of an i5 SP3 disk.
Back to the Surface Book test results vs the SP4 core m results ... *maybe* the results are that much better on the SB because it was an i7 vs a Core m, in which case having a faster drive on the Core m wont matter and the results were already much better than the SP3 with the SB and might improve still more with the eventual tuning that always happens post release with MS. IIRC the Passmark disk test I ran had the SP4 core-m version slightly ahead of an i5 SP3 disk.
I just ran the AS SSD benchmark as well. Unfortunately, while I had downloaded it before, I never ran it until today with the latest 10586 build and Samsung NVMe drivers (I suppose I can try to remove the Samsung NVMe drivers and try again, but I'll need to research how to do that without borking my computer - can I just go to DM and "uninstall drivers" and let the OS reinstall the originals?).
In any case, I get a higher overall AS SSD score than NBR got for the SB, but only by a margin that is explained by the fact that I am testing the 512GB drive and they were testing the 256GB drive. The overall pattern is similar to the speeds I see with CDM, though AS tests with 64 threads and only do the 4K test in threaded mode whereas CDM uses 32 threads and tests both 4K and sequential.
My numbers on CDM still match the same pattern seen for the numbers that Anandtech got, but mine are a little lower - possibly because I don't know how to completely quiet my machine before I run the numbers (I don't run any other apps and check Task Manager to make sure that CPU and Disk are <1-2% before I run it, but I don't go out and stop any services or anything).
In any case, I get a higher overall AS SSD score than NBR got for the SB, but only by a margin that is explained by the fact that I am testing the 512GB drive and they were testing the 256GB drive. The overall pattern is similar to the speeds I see with CDM, though AS tests with 64 threads and only do the 4K test in threaded mode whereas CDM uses 32 threads and tests both 4K and sequential.
My numbers on CDM still match the same pattern seen for the numbers that Anandtech got, but mine are a little lower - possibly because I don't know how to completely quiet my machine before I run the numbers (I don't run any other apps and check Task Manager to make sure that CPU and Disk are <1-2% before I run it, but I don't go out and stop any services or anything).
OK, so I finally did the rollback. I used the Device Manager -> storage controller -> properties -> driver -> Rollback method and ended up with a Microsoft branded "Standard NVMe controller" driver dated 2006 version 10.0.10586.0 which I assume are the drivers that came with the recent 10586 Windows 10 update. Since they are dated 2006 I'm wondering if they aren't some generic drivers that simply get copied forward into every Windows release, but are usually replaced by vendor specific drivers?
In any case, the results are that the AS SSD benchmark is well and truly ruined by these drivers. It's not a case of "gee, that doesn't seem like the fastest SSD" kind of performance, it is over 100 times slower on the 4K random benchmark. The CDM benchmark had similar results to before I did the Samsung driver update so this isn't a case of botching the driver rollback, the AS SSD benchmark seems to trigger a really bad bug in the standard drivers. Here is my history of results.
Right before I updated to the Samsung NVMe driver I ran a CDM benchmark to test the results of the 10586 Windows 10 update:
Then I did the Samsung NVMe update and ran both CDM and AS SSD and got these results:
Not the greatest PCIe SSD results, but they look valid. Since I hadn't run the AS SSD benchmark before the Samsung driver update I decided to roll it back and test again and got these results:
That AS SSD result is just ridiculous, it's not representative of the SSD performance and I don't think I've seen any real world results to indicate such performance. Still, it may be worth updating to the Samsung drivers to avoid whatever problem is shown there. It won't fix the "Microsoft chose a cheap TLC SSD" issue, but it will potentially avoid a nasty performance bottleneck.
I ran with the Samsung drivers for around 3 days and saw no problems at all. I'm going to update back...
In any case, the results are that the AS SSD benchmark is well and truly ruined by these drivers. It's not a case of "gee, that doesn't seem like the fastest SSD" kind of performance, it is over 100 times slower on the 4K random benchmark. The CDM benchmark had similar results to before I did the Samsung driver update so this isn't a case of botching the driver rollback, the AS SSD benchmark seems to trigger a really bad bug in the standard drivers. Here is my history of results.
Right before I updated to the Samsung NVMe driver I ran a CDM benchmark to test the results of the 10586 Windows 10 update:
Then I did the Samsung NVMe update and ran both CDM and AS SSD and got these results:
Not the greatest PCIe SSD results, but they look valid. Since I hadn't run the AS SSD benchmark before the Samsung driver update I decided to roll it back and test again and got these results:
That AS SSD result is just ridiculous, it's not representative of the SSD performance and I don't think I've seen any real world results to indicate such performance. Still, it may be worth updating to the Samsung drivers to avoid whatever problem is shown there. It won't fix the "Microsoft chose a cheap TLC SSD" issue, but it will potentially avoid a nasty performance bottleneck.
I ran with the Samsung drivers for around 3 days and saw no problems at all. I'm going to update back...
Just updated back to the Samsung driver - it was easy, I just went to the controller and clicked update driver and it found the Samsung. That may have been due to the fact that I had downloaded them from Samsung's web site, but perhaps they are the latest drivers available via the standard update mechanism?
In any case, my AS SSD benchmark results returned to "non-botched" values:
If someone else wants to try this, can you please answer/confirm the following?
- What is your driver version before you update? (10.0.10586.0?)
- Run AS SSD before you update and see if it is similar to my rollback results above (previous message)
- Try using Device Manager -> update driver and see if it finds the Samsung drivers without manually downloading them. Does it? If not, then use the link below to manually download and install them.
- What is your driver version after the update? (1.4.7.16?)
- Run AS SSD after the update and see if it is similar to my Samsung driver results above.
Link for Samsung drivers: Samsung SSD Downloads | Samsung SSD
In any case, my AS SSD benchmark results returned to "non-botched" values:
If someone else wants to try this, can you please answer/confirm the following?
- What is your driver version before you update? (10.0.10586.0?)
- Run AS SSD before you update and see if it is similar to my rollback results above (previous message)
- Try using Device Manager -> update driver and see if it finds the Samsung drivers without manually downloading them. Does it? If not, then use the link below to manually download and install them.
- What is your driver version after the update? (1.4.7.16?)
- Run AS SSD after the update and see if it is similar to my Samsung driver results above.
Link for Samsung drivers: Samsung SSD Downloads | Samsung SSD
I got the i7 512GB and it comes with samsung as well. I used CrystalDisk5 and got these:
Seq (Q32T1) Read 1436MB/s
Seq (Q32T1) Write 573.4MB/s
4K (Q32T1) Read 319.4MB/s
4K (Q32T1) Write 201.6MB/s
Seq Read 749.4MB/s
Seq Write 418.6MB/s
4K Read 23.56MB/s
4K Write 72.82MB/s
Seq (Q32T1) Read 1436MB/s
Seq (Q32T1) Write 573.4MB/s
4K (Q32T1) Read 319.4MB/s
4K (Q32T1) Write 201.6MB/s
Seq Read 749.4MB/s
Seq Write 418.6MB/s
4K Read 23.56MB/s
4K Write 72.82MB/s
Attachments
Last edited:
Thanks. Corrected it
