SanDisk Extreme Pro SSD (240GB, 480GB & 960GB) Review: The Fastest Just Got Faster
by Kristian Vättö on June 16, 2014 4:00 PM EST- Posted in
- Storage
- SSDs
- SanDisk
- Extreme Pro
Performance Consistency
Performance consistency tells us a lot about the architecture of these SSDs and how they handle internal defragmentation. The reason we don’t have consistent IO latency with SSD is because inevitably all controllers have to do some amount of defragmentation or garbage collection in order to continue operating at high speeds. When and how an SSD decides to run its defrag or cleanup routines directly impacts the user experience as inconsistent performance results in application slowdowns.
To test IO consistency, we fill a secure erased SSD with sequential data to ensure that all user accessible LBAs have data associated with them. Next we kick off a 4KB random write workload across all LBAs at a queue depth of 32 using incompressible data. The test is run for just over half an hour and we record instantaneous IOPS every second.
We are also testing drives with added over-provisioning by limiting the LBA range. This gives us a look into the drive’s behavior with varying levels of empty space, which is frankly a more realistic approach for client workloads.
Each of the three graphs has its own purpose. The first one is of the whole duration of the test in log scale. The second and third one zoom into the beginning of steady-state operation (t=1400s) but on different scales: the second one uses log scale for easy comparison whereas the third one uses linear scale for better visualization of differences between drives. Click the buttons below each graph to switch the source data.
For more detailed description of the test and why performance consistency matters, read our original Intel SSD DC S3700 article.
 |
|||||||||
| SanDisk Extreme Pro | SanDisk Extreme II | Intel SSD 730 | Intel SSD 530 | OCZ Vector 150 | |||||
| Default | |||||||||
| 25% Spare Area | |||||||||
Similar to the Extreme II, the IO consistency is just awesome. SanDisk's firmware design is unique in the sense that instead of pushing high IOPS at the beginning, the performance drops close to 10K IOPS at first and then rises to over 50K and stays there for a period of time. The higher the capacity, the longer the high IOPS period: the 960GB Extreme Pro takes ~800 seconds before the IOPS drops to 10K (i.e. the drive reaches steady-state). I do not know why SanDisk's behavior is so different (maybe it has something to do with nCache?) but it definitely works well. Furthermore, SanDisk seems to be the only manufacturer that has really nailed IO consistency with a Marvell controller because Crucial/Micron and Plextor have had some difficulties and their performance is not even close to SanDisk.
However, I would not say that the Extreme Pro is unique. Both Intel SSD 730 and OCZ Vector 150 provide the same or even better performance at steady-state, and with added over-provisioning the difference is even more significant. That is not to say that the Extreme Pro is inconsistent, not at all, but for a pure 4KB random write workload there are drives that offer (slightly) better performance.
 |
|||||||||
| SanDisk Extreme Pro | SanDisk Extreme II | Intel SSD 730 | Intel SSD 530 | OCZ Vector 150 | |||||
| Default | |||||||||
| 25% Spare Area | |||||||||
 |
|||||||||
| SanDisk Extreme Pro | SanDisk Extreme II | Intel SSD 730 | Intel SSD 530 | OCZ Vector 150 | |||||
| Default | |||||||||
| 25% Spare Area | |||||||||
TRIM Validation
To test TRIM, I filled the drive with sequential data and proceeded with 60 minutes of 4KB random writes at queue depth of 32. I measured performance with HD Tach after issuing a single TRIM pass to the drive.
TRIM works for sure as the write speed is at steady 400MB/s.
.
Facebook
Twitter
RSS
85 Comments
View All Comments
Solix - Tuesday, June 17, 2014 - link
I get the enthusiasm for the increase in bandwidth by going PCIe, but there are other constraints for some of us. What about those of us that already have all of their PCIe slots crammed full to the brim with GPUs and Sound cards and stuff? Maybe vendors can put internal PCIe slots in without the need for using a chassis slot. After all, most of them just have a back plate with nothing right? Or something that lets me use a riser cable and mount it elsewhere instead of burning a precious one of my back panel slots.457R4LDR34DKN07 - Tuesday, June 17, 2014 - link
That is essentially what M.2 does...TelstarTOS - Tuesday, June 17, 2014 - link
Thanks. I really needed to see the performance of the new model before buying the old one on sale :)mapesdhs - Tuesday, June 17, 2014 - link
Kristian, just curious, why did the Vector 512GB top the IOMeter Sequential
Write table? Just seems a bit of an oddity given where it ends up in all the
other tables.
Btw, your AS-SSD tables are amusing in this regard; I ran AS-SSD on my
Vector 512GB, it gave 516.94 seq. read, 502.78 seq. write (topping both
tables), overall score of 1091. Interestingly, although my Vector 128GB
drops its seq. write to 386, the seq. read still stays healthy at 515.18.
I do like the Vector series, they work very well. I bought a Vector-150
128GB; it's seq. write is 10% higher, but the 4K numbers are lower,
resulting in a reduced overall score (989 vs. 1040).
The highest overall score I've had so far was 1147 from an 840 Pro 512GB.
(all done via Intel SATA3 on Z68)
Ian.
mapesdhs - Tuesday, June 17, 2014 - link
Caveat: in general prefer Samsung models for C-drives, for better long termconsistency. Here's the 840 250GB in my 3930K setup:
http://www.sgidepot.co.uk/misc/samsung_840_250GB_H...
Vectors are great for AE cache drives (ditto Vertex4), though my 2700K does
have a Vector 256GB for its C-drive; here's an HDTach:
http://www.sgidepot.co.uk/misc/OCZ_Vector_256GB_HD...
Pretty decent; not as quick/smooth as my 3930K's 840 250GB, though an AS-SSD
run is good:
http://www.sgidepot.co.uk/misc/OCZ_Vector_256GB_AS...
These Extreme Pros look interesting. I might get one to see how they fare for AE, etc.
Ian.
KAlmquist - Wednesday, June 18, 2014 - link
In case anyone missed it,the Sandisk Extreme Pro prices listed on the last page of the article are not real prices because the items were out of stock at the time the reviewer pulled the prices from Newegg. Now the 240 and 480 GB models are in stock for $270 and $400, respectively. All three models are in stock at Amazon.com with prices of $240, $430, and $700.The SSD market is looking like what we often see in the CPU market, with large price premiums for the fastest models. I won't be buying a Sandisk Extreme Pro, but there probably are people who could really benefit from the performance.
Xajel - Wednesday, June 18, 2014 - link
Interesting, the price went up..The 240GB model is cheaper in Amazon now compared to newegg
( A vs. NE ) : $229.99 vs. $269.95 or $0.958/GB vs. $1.125/GB
The 480GB model is cheaper in newegg
( A vs. NE ) : $429.99 vs. $399.95 or $0.896/GB vs. $0.833/GB
The 960GB model is not available in newegg.. but it cost $699.99 or $0.729/GB in Amazon
milli - Wednesday, June 18, 2014 - link
"SanDisk is turning out to be a very dangerous player in the client SSD space. With nearly perfect vertical integration model (they just lack client controller silicon and DRAM), SanDisk has the ability to put against Samsung and Intel who have traditionally held the performance crown."They actually do have their own controller. The U110 and the Standard (and maybe more) use Sandisk's own controller. Granted these are low end.
skarthikeyan - Saturday, June 21, 2014 - link
Hi, shouldn't writes be slower than reads in drives, more so in SSDs? For eg, on http://anandtech.com/show/8170/sandisk-extreme-pro... sequential read for the Extreme Pro 960GB is 438 MBps while sequential write is 371.4. But in the case of random read(98.5) vs write(257MBps for QD=1), write is faster. Why is that the case?skarthikeyan - Saturday, June 21, 2014 - link
Sorry, my comma broke the link, the correct link to that page is http://anandtech.com/show/8170/sandisk-extreme-pro...