Performance Consistency

Performance consistency tells us a lot about the architecture of these SSDs and how they handle internal fragmentation. The reason we do not have consistent IO latency with SSDs 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 (Logical Block Addresses) 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 dropdown selections 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.

Samsung SSD 850 EVO 250GB
Default
25% Over-Provisioning

The 850 EVO presents a healthy increase in IO consistency. The 840 EVO wasn't exactly inconsistent in the first place, but the 850 EVO takes the steady-state IOPS from ~3,000-5,000 IOPS to 5,000-8,000 IOPS, which is actually nearly on par with the 850 Pro. The 850 EVO has without a doubt one of the highest performance consistencies out of the value/mainstream drives we have tested.

Samsung SSD 850 EVO 250GB
Default
25% Over-Provisioning

 

Samsung SSD 850 EVO 250GB
Default
25% Over-Provisioning


TRIM Validation

To test TRIM, I filled a 120GB 850 EVO with sequential 128KB data and proceeded with a 30-minute random 4KB write (QD32) workload to put the drive into steady-state. After that I TRIM'ed the drive by issuing a quick format in Windows and ran HD Tach to produce the graph below.

And TRIM works as expected.

Endurance: Close to Planar MLC NAND AnandTech Storage Bench 2013
Comments Locked

97 Comments

View All Comments

  • Kevin G - Monday, December 8, 2014 - link

    What version is the firmware of the 840 EVO? There is a notorious bug out there that'll drastically reduce the speed of the drive when it attempts to read data.

    See:
    http://www.anandtech.com/show/8617/samsung-release...
  • romrunning - Monday, December 8, 2014 - link

    That's the bug about reading really old data. If he did an upgrade, then it wouldn't be really old data on the new drive.

    It also doesn't apply to the point he was making - would I subjectively "feel" the difference if I upgrade to the 850 EVO from an older <insert_brand_here> SSD?
  • doggghouse - Monday, December 8, 2014 - link

    On one of these review sites, they mentioned that the step up from HDD to SSD is enormous, but between the different generations of SSD the difference isn't nearly as noticeable. Depending on the test, the difference between HDD and SSD is anywhere from 2.5x faster (ex. sequential write) to 140x faster (ex. random write). You definitely will not see such a marked improvement going from one SSD to another.

    Also I agree with your point about the benchmarks being pointless for the average user. A lot of the benchmarks are geared towards enterprise usage. I think part of the problem is that the drives all behave pretty similarly (compared to HDD anyway), until they are pushed to the extremes. That's the only way for the reviews to differentiate between them, unless they find a particular weak point or flaw.
  • eanazag - Monday, December 8, 2014 - link

    The performance profile of the drives and your workload will dictate if you notice the better performance. I would guess that you got a lot more GBs/$ when you picked up the Evo versus the Corsair drive. Also, not you don't need to struggle to get everything on to your boot drive. At the same time your computer may only startup 2 or 3 seconds faster.

    They don't for an average user, but if you're here there's a good chance you're not an average user. And those fancy benchmark numbers cost you what?

    There are real life workloads that will benefit for an average user. I have seen the consistency play out in my usage in various drives. I have seen garbage collection routines cause some painful performance issues in software encrypted drives.
  • MrSpadge - Monday, December 8, 2014 - link

    With a half-decent SSD your I/O is probalby mostly CPU-limited in the real world anyway.
  • ummduh - Monday, December 8, 2014 - link

    In my opinion, the best reason to continue to upgrade, is capacity increases.

    My first drive was 60GB, my second is 120GB, looks like my third is going to be 250GB....

    No the speed difference isn't that drastic once you get off the HDD, but now you can start keeping more and more stuff on a drive at the same price point. All of my drives have been at the $150 point.. (Ok, so I had 3 different 60GB drives, but that was due to not realizing how bad OCZ suuuuuucked and I kept going back like a moron)
  • cm2187 - Tuesday, December 9, 2014 - link

    The other thing is that most SSDs now are limited by the SATA interface. And to be honest, unless you are running a database with thousands of queries per seconds, beating the SATA interface won't really add much to your experience. Larger capacities I think would be more useful.
  • Badelhas - Thursday, December 11, 2014 - link

    Thats exactly what I always think when I read this reviews. Altough they are great to read, for someone who already has a SSD the gains are not noticeable. I have a Vertex 3 120Gb that I bought in 2011 and been questioning myself if I would se any real gains going to a 840 Evo, for instance, since they are much newer but you just answered my question ;)
  • matej_eu - Thursday, June 25, 2015 - link

    I am in a similar situation. My first SSD was the ADATA's SP600@64GB running on SataII, I was so mesmerized by the speed of everything (boot time, installation, app run in time, ...). I really do take care of my Win installation, drivers, and apps, but in the end size was to small for boot+main drive in laptop. I bought Samsung 840 EVO@120GB, and in the recent times I thought I was going crazy, system was noticeably slowing down. I googled, and what do you now, problems, no more Samsung EVOs for me!! Buy Samsung 8x0 PRO or some other brand.
  • milli - Monday, December 8, 2014 - link

    Your Intel SSD 730 prices are wrong. They might have gone up.
    The price of the ARC 100 is amazing ATM. It scores very good in the '2013 Storage Bench service time' which in my experience tells the most about how SSDs feel in actual usage.

Log in

Don't have an account? Sign up now