The SSD Review The Worlds Dedicated SSD Education and Review Resource | 
TESTING SYSTEM AND METHODOLOGY
Since the MyDigitalSSD BP3 is a SATA 3.0 device, Turbo SATA 6GBPS mode was turned on, and Turbo USB 3.0 off in order to maximize performance. Considering the motherboard SATA 3.0 inputs utilize a Marvell controller, the 9128 chipset drivers were upgraded to firmware 1.2.0.1020 and chosen over standard/generic drivers. In addition, only stock firmware was available for the 512GB MyDigitalSSD BP3.
In testing, our main objective is to obtain results as pure and as accurate as possible and we want to ensure that no anomalies slip through. Simply put, we want to provide you with the absolute best results the tested hardware can provide. Repetition in testing is standard and, if necessary, we may conduct specific tests in Windows 7 safe mode to ensure the OS has little to no influence on the end result.
In order to validate and confirm our findings, testing is supported by industry accepted benchmark programs. All results are displayed through capture of the actual benchmark for better understanding of the testing process by the reader.
Real-world benchmarks included a batch of files and folders with capacities of 250MB, 500MB, and 1000MB respectively to test file transfer performance. Windows 7 boot times were also monitored and timed starting from the OS load screen to Steams login window (the last program to load).
For consistency, each test was run three times for the cache to kick in, utilizing AHCI as the storage mode. The average of all three was taken as the final result.
ATTO DISK BENCHMARK VER. 2.47
The ATTO Disk Benchmark measures storage performance with various transfer sizes and test lengths for reads and writes. Several options are available to customize performance measurement including queue depth, overlapped I/O and even a comparison mode with the option to run continuously. ATTO is rated to test RAID controllers, storage controllers, host adapters, hard drives and SSD drives.
As always, ATTO provides crystal-clear and accurate results. The speeds achieved by the BP3 are virtually identical to those in the specifications.
Hi
Marvell 9128 how do 500mbyte/s transfer speed?
This drive just got even better new FW is out sending a copy to Les Now should be up and available for DL later today.
Hi Deepak, Nice review. I’m puzzled about how the 4k write transfer rate in CDM is 49, Anvils its 59, and as ssd is 132? Ive never seen such a thing before.
Thanks Bill,
I just noticed that as well, and yes very odd. We will wait for Deepak to get back on this.
Hey Bill,
Thanks for noticing that. I uploaded the wrong file initially, and re-uploaded the correct one. For some reason, WordPress saves a cache if you have two files with the same name, even if you delete them, and fails to update. So in this case, I deleted the incorrect one, uploaded the correct one with the same filename, but WP overwrote the right one with the wrong one regardless, since the cache has the wrong one stored first.
I’ll rename the correct picture and upload it ASAP. It’s moments like these I’m glad I backup my review files 🙂
Thanks a ton for catching that, and please let me know if you find anything else.
I wan’t an SSD for BF3… But I could also use a new graphics card…
When you figure this out Bill let me know. This has me really intrigued. How about a price for a 512gb flavor?
Amazon has it listed for $419, but I reckon you can find it for $20-$30 cheaper.
the question is really whether this is better than the crucial M$ which is $20-50 less 🙂
Good point. According to the specs, the write speed on the M5 is about 60MB/s slower than the BP3, and roughly the same read speed (I would rate it a tad slower).
Just depends on what you’re looking for the most, and of course, if you can get a great deal on it! 😀
you quote max speeds of sandforce drives (which are based from an intel p67/z68) then use a P55 chipset with the under-performing marvell controller to do this review????
as a reviewer you need some modern hardware that will actually allow the max performance to be shown from this ssd.
Did you run a few passes of ASU endurance test over this drive before you ran any tests.
I found that the Torqx 2 (using Phison controller) performed quite well FOB but write performance fell to crap after all the NAND was written and the drive started to have to perform garbage collection.
Testing such as this is not part of our test protocol.
Unfortunately, FOB performance is not normal for ssds, and phison controllers in the past have shown themselves to be very slow after they have exhausted their FOB state
I’m talking about catastrophic degradation, my torqx-2 went from more then 100meg per sec writes to 25meg per second writes after FOB condition was exhausted. (sequential writes). I couldn’t really recommend a phison controller based drive until I saw this problem fixed.
It isn’t generally a huge concern with marvell and sandforce drives, which generally exhibit a smaller amount of difference between FOB and steady state, but it would be very good to see how this particular drive handles it!
Well aware of the different levels of performance and our testing for the typical consumer drive has been shaped in the direction that we have found the reader enjoys the most. There seems to be a very fine line between providing a package that keeps the reader interested and then one that becomes a bit too technical for most.
Unfortunately, there are so many different experiences by so many different consumers that we cannot conform our testing to that of each. I am familiar with your experience and have to say that it doesn’t speak to the new Phison controller, nor is it the norm with the former. Thank you for the suggestions.
I actually have two OCZ Vertex 2 60GB’s using the SandForce 1200 controller, and they too have degraded an insane amount since their FOB state. There is a massive thread about it at the OCZ forums, but suffice it to say SF isn’t immune to it (which you also mentioned).
I ran the BP3 through MANY tests, and even separate controllers on the motherboard, and got consistent results the entire time. I gotta agree with Les on this one, but you make some solid points, and I feel it’s something that you would only really find out after heavy, prolonged use (sleep, hibernate, reboot, shutdown, gaming, high-performance apps, read/write tests, moving a ton of files, and doing this all in a regular and timely manner).
Thanks for the comment!
Thanks for clarifying. I was finding that on my Torqx 2 64gig, that after a secure erase, once I wrote 64 gig to the device, that performance fell to 25meg per second and would not really recover.
It wasn’t so much that pounding the drive resulted in serious degradation … it was simply that it couldn’t maintain decent performance at all once it had to start garbage collecting and erasing cells.
It is good to hear that phison has solved this problem in their newest controller.
That’s awful to hear, and I share much the same with my Vertex 2’s.
Hopefully a fix comes out for you!
I’m glad you brought it up, and yes Phison did fix this issue. It’s a brand new controller, utilizing SATA 3 for the first time, and works as advertised.
I guess it’s the price we pay as early adopters! 😀