Silicon Power E10 and M10 SSDs
Published: 26th February 2010 | Source: Silicon Power | Price: £TBC |
Todays test setup is our standard P55 based test hardware, and as we still had the ASUS Maximus III Extreme in the OC3D bunker from last weeks testing it was decided to use that to make sure we provided the best possible scaffolding for this house to be built on.
Motherboard : ASUS Maximus III Extreme
CPU : Intel Core i7 870
Graphics : ASUS GTX275
Cooler : Cooler Master Hyper 212 Plus with Arctic Cooling MX3 Thermal Compound
PSU : Cougar 1000CM
OS : Windows 7 64
RAM : G.Skill ECO PC3-12800 CL7
Testing SSD drives is always a strange experience. Most of the standalone programs designed to test drive speeds are better suited to testing mechanical devices and it's for this reason that we always over-test something to ensure we get a good feel for the speeds available.
As always all our tests are run five times with the highest and lowest being removed and an average taken of the rest.
ATTO Disk Benchmark
The ATTO benchmark is a good test of a drives sustained transfer rates both in read and write conditions. It is able to test in both small and large blocks, which is especially good with SSDs as they definitely perform better with large file chunks. Although that isn't to suggest they aren't good with small ones, just the difference is better on the larger items.
Both drives provided exceptional read speeds. Any worries that the inability to get a good look at either the controller or cache chips might mean something being hidden were quickly dispelled. The E10 surpassed the magic 200MB/s barrier on all of our tests. The Samsung based M10 provided phenomenal consistency with 150MB/s seen in everything but the 8192 test.
Read speeds, like clock speeds, tend to be the big number than gets heavily advertised and doesn't tell the full story. After all, a drive that you don't write to is almost entirely pointless, and one that only writes very slowly is almost equally without use. Thankfully both drives keep up the read performance.
The E10 consistently around 150MB/s, only dropping below 100MB/s in the small 128 test.
The M10 once again shows the consistency achieved by using the popular Samsung NAND MLC, with around 95MB/s seen across all the tests.
Crystal Disk Mark 2.2
Crystal Disk Mark is able to test tiny, medium and sequential reading and writing with a variety of sizes. To ensure compatibility we use the default settings. The read tests are indicative of the performance difference between the two drives, with the Sequential test backing up the ATTO results, and the 4K and 512K random tests providing a sterner test. Naturally the 4K result shows the poor performance of SSDs with exceptionally small files, but once we reach a more realistic size the speed differential between solid state and mechanical is there for all to see.
The results for the write test are very similar to the read test. Small sizes are bad, but otherwise the transfer rate is around the ATTO results and also around the manufacturers specifications.
PC Mark Vantage
PC Mark Vantage uses a variety of in-built applications and testing situations to provide a more real-world scenario and a greater approximation of the speeds that should be achieved when the entire system is in use.
Rather than run the entire Vantage suite of tests we've concentrated on the HDD Suite. The faster E10 giving a result 1000 points higher than the M10, but neither are slouches.
Viewing the breakdown of the results obtained we can see how the entire subsystem can slow down the transfer rates. Of course these are still vastly above what you'd expect with a mechanical drive, but we can't help to be slightly disappointed by some of the scores when the system is really working hard. Especially strange is how the M10 was faster at application loading and Windows Media Player than the E10.
Does this translate into extra speed loading games?