Guest Phil

New SSD review Tomshardware (Intel, Sandforce, C300 etc.)

Recommended Posts

Tomshardware published a new SSD review: http://www.tomshardware.com/reviews/windows-7-ssd-trim,2705.html

Unfortunately zero real world benchmarks were included. Still an interesting read though.

edit your link :)

s/b http://www.tomshardware.com/reviews/windows-7-ssd-trim,2705-22.html

^appears the 'comma' after 'trim' in the link, can't be handled by the forum software?

...they left out the best all around performer in the OWC SSD's.

At this time, it makes sense to purchase an SSD if you’ve been waiting for balanced and affordable products to become available. Drives like the OCZ Vertex 2 or the G.Skill don’t outperform the other drives, but they do well in all benchmarks and even deliver good bang for the buck.

We therefore grant our Recommended Buy Award to these two 100 to 120 GB products.

Crucial’s RealSSD C300 remains the very best choice these days after a slow start and with firmware version 002. But the other options are definitely all worth considering.

^bet the 002 firmware doesn't fix the seasoning performance hit:

http://macperformanceguide.com/SSD-RealWorld-Seasoning.html

TRIM

The TRIM feature helps restore an SSD to its original performance. But this requires operating system support, and Mac OS X does not have it. PC users might fare better in keeping their SSD in good working order. Bottom line is that TRIM is a half-assed feature for inferior technology.

The 256GB RealSSD is a SATA 6Gb/sec device, but as of April 2010, no Mac has such capability. It works compatibly with SATA 3Gb/sec however, which is what Macs have.

With active power consumption of 4.3 watts for writes and 2.1 watts for reads, it is a power pig, and thus a dubious choice for a laptop. Power consumption at idle is much lower at 0.094 watts, so it is still a good choice when little disk I/O occurs during normal use of the computer.

Crucial doesn’t given data reliability other than stating “Built-in EDC/ECC”, a dubious “specification”.

Performance brand-new and after seasoning

The Crucial RealSSD offers consistent and fast read performance both before and after seasoning. Top-notch for reads.

After seasoning ,write performance degrades, and develops regular “dropouts”— those steep downward spikes likely indicate performance approaching 0MB/sec for a few seconds, since each data point represents a 2.5GB file (1000 data points). Video capture will not fare well under such a scenario, frames could be dropped.

Edited by udaman

Share this post


Link to post
Share on other sites
TRIM

The TRIM feature helps restore an SSD to its original performance. But this requires operating system support, and Mac OS X does not have it. PC users might fare better in keeping their SSD in good working order. Bottom line is that TRIM is a half-assed feature for inferior technology.

Hmmm......I find it hard to take seriously an article with writing like this.

Given that TRIM is supported by Windows 7, Linux, and BSD I believe, TRIM isn't limited to just PC users. This sounds more like sour grapes from a Apple fanboy.

Share this post


Link to post
Share on other sites
Guest Phil

I'm quite surprised by the Vantage results, especially Intel G2 coming out tops.

pcmark_vantage_score.png

For comparison, Anandtech:

23543.png

Share this post


Link to post
Share on other sites

Which gets to the root as to why we're trying to run more real world tests that mirror use cases. We're aiming for 8-10 by the time we're done.

Share this post


Link to post
Share on other sites
Guest Phil

I'd like to see separate benchmarks measured in seconds, instead of scenario's that blend everything together.

Laptopmag makes nice real world benchmarks, like here:

intel-x25-application-opens-stress.gif

intel-x25-filecopy.gif

Source

I think that's more representative then the way SR is doing it now. For example: going by the SR reviews the Sandforce drives seem almost twice as fast as Intel G2. In real world usage the difference will be much smaller, I expect.

Share this post


Link to post
Share on other sites

Well our opinion is more about testing the drive instead of the system as a whole. Most of the application load tests can vary greatly depending on the system the drive is inside of. Our tests at a basic level try to keep as most of the system out of the equation as possible. You still need to work around the chipset limitations, but our stuff should in theory perform the same with a slow budget processor or a top of the line overclocked Core i7.

Another problem we noticed with application load times and smaller individual application traces was the lack of precision. You can start to see it in some of the tests your linked, where at a certain point all drives will start giving the same score. Toms Hardware or Anandtech has a gaming trace where almost any decently performing SSD gets the same score, with one super performer being on top. Our tests are broad enough where you can see differences between drives in the same class, or in the case of the SandForce models the same hardware but different configurations.

Share this post


Link to post
Share on other sites

Well, and to a certain extent our different methodologies are good...gives buyers a lot of options in terms of making an educated purchasing decision. We don't have any expectation that SR will ever be the only place to read a review of a certain drive...although if you go that route - cool by me ;)

Share this post


Link to post
Share on other sites
Guest Phil

When I see a scenario like this "playing one 720P HD movie in Media Player Classic, one 480P SD movie playing in VLC, three movies downloading simultaneously through iTunes, and one 1080i HDTV stream being recorded through Windows Media Center over a 15 minute period."

My thoughts are:

-very few users would actually run a scenario like that (no one?)

-it's heavy multi tasking, what about performing single tasks.

I'm not surprised the C300 on SATA II comes out so bad on SR. I tested it, it's a good performer with single tasks (as shown by TR). When multi tasking it falls short on the Sandforce drives.

Benchmarks like this one from Techreport are more realistic to me. They show the real differences between the SSDs, in a real world scenario.

time-load-mw2.gif

I wish SR would include benchmarks like this.

Share this post


Link to post
Share on other sites

Why didn't they include an X25-E 64GB on all those results? Just curious, is it that hard to find and test one? They included equally expensive SSDs so I don't think it's the cost that prevented them from having one.

Edited by StamatisX

Share this post


Link to post
Share on other sites

The E doesn't get sampled out to reviewers very much...there's a few others that are enterprise or OEM that see the same fate. SanDisk, Samsung, Seagate, etc.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now