Exploring The Performance Of A Full SandForce-Based SSD
One of the reasons we brought back fresh out-of-box testing in Iometer was to free up time to explore other aspects of performance. For this round-up, we specifically want to investigate the read performance of a full drive. After all, a 60 GB drive isn't very large, so it's very likely that you'll be dealing with this exact scenario on and off.
In theory, read speeds shouldn't be significantly impacted by the drive's free space. But there is a noticeable drop-off on the lower-capacity SF-22xx-based SSDs. Compression is one reason for this. Compressed data has to be decompressed, which requires sufficient free space. SandForce's garbage collection mechanism is a second explanation. Because most of its clean-up happens in the foreground, decompression requires a rotation of scratch space so that the NAND wears out evenly. This is known as wear leveling.
Both factors penalize the 60 GB drives we're rounding up more than the larger models as a result of limited NAND bandwidth, a side-effect of leaning on fewer NAND dies. We're using HD Tune to illustrate.
The image above shows how read performance varies. After using Acronis True Image Home to clone our system drive, we set HD Tune to read the entire LBA space sequentially. As you can see, reading back information written to our 60 GB Agility 3 falls between 150-200 MB/s. Once we move on to free space, performance accelerates to 450 MB/s. That's a big difference.
But what happens when we get more specific and narrow our focus to incompressible data? After adding about 6 GB of MP3s and H.264-encoded video, we can clearly discern performance at its worst on a SandForce-based drive. Read performance drops to about 150 MB/s. That's why you see performance bounce up and down when we only have Windows and a few apps installed: some of that data is compressible, and some isn't.
That's not the end of the story, though. We repeated this same tests on all of the drives in our round-up and found that the results only apply to 60 GB SF-22xx-based SSDs accessing asynchronous NAND. Let's take Intel's SSD 520 for a spin to compare the performance of a drive with higher-performing synchronous flash.
We already established (using early Iometer testing) that reading back free space is similar regardless of a drive's NAND interface. Yet, comparing the SSD 520 to OCZ's Agility 3, we see read performance in the space occupied by Windows and our application data falls between 250-300 MB/s. In a worst-case scenario, incompressible data is read back at 250 MB/s. Sure, that's low compared to the 450 MB/s we see otherwise. But it's also 100 MB/s faster than the 60 GB Agility 3 and its asynchronous NAND-based competition.
Now, to put it all in context. Check out the results from our 64 GB m4 in the chart below. Crucial mostly employs background garbage collection, and it doesn't lean on compression for its commendable performance. As a result, read performance is largely unaffected by fill state.
Current page: Exploring The Performance Of A Full SandForce-Based SSDPrev Page Endurance Testing Next Page Performance Is Defined By Flash
Stay on the Cutting Edge
Join the experts who read Tom's Hardware for the inside track on enthusiast PC tech news — and have for over 25 years. We'll send breaking news and in-depth reviews of CPUs, GPUs, AI, maker hardware and more straight to your inbox.
As these drives are basically boot drives, i would have liked a test where you measure the total time taken to install a fresh wi7-sp1 on it and install updates and install a few softwares likeReply
Adobe pdf reader
a web browser, a photo manipulating program
a music/video player.
Install a game from a ISO.
And all these apps should be installed from the SSD itself (meaning their setups should be on the SSD).Then you should test the startup and shutdown times.
All these synthetic benchies dont make much sense, IMHO.
I have found that when working with SSD's, single core CPU performance becomes a big bottleneck in some tasks.Reply
A lot of operations use only a single core and the SSD cant use its true potential. That is, the CPU cant process data as fast as the SSD can provide.
This is just reverse of what happens in case of mechanical HDD's.
You're not going to see a major difference.
mayankleoboy1I have found that when working with SSD's, single core CPU performance becomes a big bottleneck in some tasks.A lot of operations use only a single core and the SSD cant use its true potential. That is, the CPU cant process data as fast as the SSD can provide.This is just reverse of what happens in case of mechanical HDD's.Reply
Well, it is pointless though since everything you are doing is so fast that it doesn't matter anymore. I however see your point since I can be loading a program and my SSD is not even at max speed my CPU frequency is maxed out. The only way to get more speed is to just overclock as much as you can.
ackuhttp://www.tomshardware.com/review 24-14.htmlYou're not going to see a major difference.Reply
that is the point of buying a cheaper SSD based on a chepaer NAND.
Considering the conclusion that performance is defined by flash, I find it interesting that the one SF2281 with Toggle NAND at 60GB is not in the roundup (in North America anyway). The Mushkin Chronos Deluxe 60 is substantially cheaper now at $99. It's performance characteristics are much more profound than the 25nm ONFI sync/async models. They're often out of stock at Newegg, and for good reason.Reply
Is there a benchmark to compare virtual memory performance? My current workstation has 24gb of memory, which means Windows eats up 36gb of my boot drive for virtual memory. (yes, I know I can change/disable it, but some programs act wonky when it's screwed with). A dedicated virtual memory drive would free up space on my primary ssd, as well as keep the writes down.Reply
I'd also like to see small drives benchmarked as swap drives in video editing machines. Currently I'm using a raid 0 array of 1tb samsung drives that keeps up well enough, but I'd be interested to see if there are tangible productivity differences.
fwiw...intel uses its own premium binned 25nm sych...that why 4k read were so good.Reply
With a final page heading "Performance Is Defined By Flash" I would have like to see that difference looked at more closely. For example, the Mushkin Chronos Deluxe uses premium 3Xnm Toshiba Toggle Mode Flash (as does Patriot Wildfire, Vertex 3 Max IOPS and OWC Mercury Extreme Pro) and I would love to see for example how just changing the Flashin in an SSD from the same manufacturer and line (i.e Chronos standard versus Deluxe, Vertex 3 versus Vertex 3 Max IOPS). With that info, a user can decide whether it's makes sense to invest in say the premium Toshiba stuff as compared to the "same SSD w/o the premium Flash. That was what I expected to see when I read the referenced page heading.Reply
I'm wondering why Toms' own trace-based benchmark didn't make it into this round-up? Does it take much longer to run than the other tests? While comparing synthetics is important to determine why a certain drive behaves a certain way, trace-based benchmarks (PCMark 7 could be considered trace-based) is what makes the final purchasing decision. In this case, PCMark was the one with the most clear-cut differences, ones that would likely be mirrored in a trace-based benchmark.Reply
For a future SSD review/roundup could you take, for example, 10 real-life traces from 10 different editor's machines (the more variation in workload, the better), and then compare the %change in execution time vs. a reference drive?