Results: Tom's Storage Bench v1.0
Storage Bench v1.0 (Background Info)
Our Storage Bench incorporates all of the I/O from a trace recorded over two weeks. The process of replaying this sequence to capture performance gives us a bunch of numbers that aren't really intuitive at first glance. Most idle time gets expunged, leaving only the time that each benchmarked drive was actually busy working on host commands. So, by taking the ratio of that busy time and the the amount of data exchanged during the trace, we arrive at an average data rate (in MB/s) metric we can use to compare drives.
It's not quite a perfect system. The original trace captures the TRIM command in transit, but since the trace is played on a drive without a file system, TRIM wouldn't work even if it were sent during the trace replay (which, sadly, it isn't). Still, trace testing is a great way to capture periods of actual storage activity, a great companion to synthetic testing like Iometer.
Incompressible Data and Storage Bench v1.0
Also worth noting is the fact that our trace testing pushes incompressible data through the system's buffers to the drive getting benchmarked. So, when the trace replay plays back write activity, it's writing largely incompressible data. If we run our storage bench on a SandForce-based SSD, we can monitor the SMART attributes for a bit more insight.
| Mushkin Chronos Deluxe 120 GBSMART Attributes | RAW Value Increase |
|---|---|
| #242 Host Reads (in GB) | 84 GB |
| #241 Host Writes (in GB) | 142 GB |
| #233 Compressed NAND Writes (in GB) | 149 GB |
Host reads are greatly outstripped by host writes to be sure. That's all baked into the trace. But with SandForce's inline deduplication/compression, you'd expect that the amount of information written to flash would be less than the host writes (unless the data is mostly incompressible, of course). For every 1 GB the host asked to be written, Mushkin's drive is forced to write 1.05 GB.
If our trace replay was just writing easy-to-compress zeros out of the buffer, we'd see writes to NAND as a fraction of host writes. This puts the tested drives on a more equal footing, regardless of the controller's ability to compress data on the fly.
Average Data Rate
The Storage Bench trace generates more than 140 GB worth of writes during testing. Obviously, this tends to penalize drives smaller than 180 GB and reward those with more than 256 GB of capacity. It's a little unfair in that way, which is why it's a good metaphor for life.
First, the Vector is fast. Seagate's 600 might be an iron fist in a silk glove, but the Vector is a iron fist that rockets around at 286 MB/s. Pulling down an average data rate so far ahead of everything else is impressive, though it doesn't necessarily guarantee better performance quality on its own. Second, among all the 240/256 GB SSDs in this collection, all models (except for the Vector) are within spitting distance of each other.
Corsair's Neutron GTX falls further in the standings than Seagate. In absolute terms, the Neutron is 30 MB/s behind the 600. The GTX's 197 MB/s is certainly quick. But the Seagate implementation earns its first real performance break from the other LAMD-controlled drive with a whopping 227 MB/s.
Also notable is the value-oriented Intel SSD 335 and the pricier SSD 510. The 250 GB SSD 510 posts an average data rate just a few KB/s slower than Corsair's solution, and the SSD 335 pushes slightly ahead of Seagate's 600. Considering the vast majority of transfers are 4 KB in size, the week of system activity doesn't really punish the dilatory SSD 510 as much as the previous page's charts might have led you to believe.
It'd be reasonable to expect the Seagate and Corsair drives to hand in nearly identical numbers, based on the previous few pages of results. Given that both are likely very similar at the firmware level, the differences in performance could amount to a few gigabytes of spare NAND on the Seagate.
Service Times and Standard Deviation
There is a wealth of information we can collect with Tom's Storage Bench above and beyond the average data rate. Mean (average) service times show what responsiveness is like on an average I/O during the trace. It would be difficult to plot the 10 million I/Os that make up our test, so looking at the average time to service an I/O makes more sense. We can also plot the standard deviation against mean service time. That way, drives with quicker and more consistent service plot toward the origin (lower numbers are better here).
Given its average data rate performance, we're a little surprised to see the "slower" Neutron GTX serving up I/O as rapidly as the Seagate 600. Corsair might be doing its work in a more consistent fashion than Seagate, though. Overall, the Neutron demonstrates better quality of service, though the differences aren't spectacularly huge.
In a news flash that should surprise nobody, OCZ's Vector is still fast, breaking through with the lowest latency of all.
