Reviews - Crucial CT256M225 256GB SSD review. Keen pricing making sense.

[HEXUS]

£400 for a 256GB high-speed SSD. We investigate its credentials.

Read more.

[kalniel]

They're getting there!

But how can this get such a generally positive review, yet end up with a much lower score than the fairly rubbish corsair case that was recently reviewed?

[Tarinder]

They're getting there!

But how can this get such a generally positive review, yet end up with a much lower score than the fairly rubbish corsair case that was recently reviewed?

A couple of reasons. Firstly, this just costs significantly more and will be come redundant far, far sooner than a chassis will. We see a high-quality case as a 10-year investment - I still have one of the first Cooler Master aluminium chassis going strong - and, as such, £209 isn't a great stretch.

Secondly, it would have scored several points higher had Crucial brought TRIM functionality to the drive as soon as it was released. We're still waiting for that. The whole 'slowing down' issue is yet to be resolved and it wouldn't be right to award a drive 90 per cent when there appears to be a fundametal flaw in the underlying technology.

You may well see higher scores when testing under Windows 7, along with a few firmware releases from all the major vendors.

[pollaxe]

I got one of these recently, despite telling myself I wasn't going to bother.

I got it from Overclockers and it came to around £375 delivered.

I've not had chance to test it much but the limited experience I've had suggests it's a little bit slower than the G.skill Falcon 128 I've got but it's still very nippy.

[kalniel]

Thanks for the reply.

I don't have any concerns about the price of the case you reviewed - I happen to have a £30 Coolermaster case that's also going to last 10 years - but price or longevity don't seem to be a reason to give it such a high score when there are so many negatives in the text.

Wheras aside from lack of trim support (which doesn't seem to be supported by any OS at the moment anyway) there doesn't seem to be a huge reason to mark down this drive - yes it seems expensive compared to mechanical drives, but it's cheaper than the competition and provides way better performance than the hard drive. Whether you include concerns about a company's future support in the product score or not is another matter

I think what I'm getting are two things:
1) It seems strange to mark a product up or down as a function of its comparitive price and longevity across the entire spectrum of hardware products instead of within it's competing market. Eg it seems strange to me to praise (via the score) a case for it's longevity when all cases are long lasting, and to damn an SSD for it's lack of longevity in the face of redundancy when all SSDs and even hard drives, CPUs etc. have the same.

2) The tone of the text doesn't seem to match the final score - if you are able to justify the score in discussion afterwards then you clearly have the reasons, they just need to be made clearer in the review.

[nightkhaos]

If I wasn't planning to buy a DreyTek 2820n, I would probably get one these for my rig instead, as well as the cooling upgrade, the RAID 5 array, and maybe the sound card. It seems in the next major upgrade of my rig, if this tread continues, I'll be getting an SSD.

[GheeTsar]

doesn't seem to be supported by any OS at the moment anyway

I thought W7 supported TRIM already? The OCZ Vertex drives have now implemented it through their latest firmware and I am led to believe that it is functional in W7.

[kalniel]

Okay, any OS released to the public I meant.

[Tarinder]

Thanks for the reply.

I don't have any concerns about the price of the case you reviewed - I happen to have a £30 Coolermaster case that's also going to last 10 years - but price or longevity don't seem to be a reason to give it such a high score when there are so many negatives in the text.

Wheras aside from lack of trim support (which doesn't seem to be supported by any OS at the moment anyway) there doesn't seem to be a huge reason to mark down this drive - yes it seems expensive compared to mechanical drives, but it's cheaper than the competition and provides way better performance than the hard drive. Whether you include concerns about a company's future support in the product score or not is another matter

I think what I'm getting are two things:
1) It seems strange to mark a product up or down as a function of its comparitive price and longevity across the entire spectrum of hardware products instead of within it's competing market. Eg it seems strange to me to praise (via the score) a case for it's longevity when all cases are long lasting, and to damn an SSD for it's lack of longevity in the face of redundancy when all SSDs and even hard drives, CPUs etc. have the same.

2) The tone of the text doesn't seem to match the final score - if you are able to justify the score in discussion afterwards then you clearly have the reasons, they just need to be made clearer in the review.

I see what you're getting at but there are more immediate reasons why the drive doesn't score 90 per cent.

As a harbinger of things to come, the pricing of this will drop, very quickly, in the next few weeks and months, so as much as it's a good deal now, the deal will become a lot sweeter very soon. It's also not as fast as the Intel drive and not as good at random reading as all-Samsung drives.

We're also slightly concerned that the competition is doling out newer firmwares when Crucial tends to be conservative. I guess enthusiasts want the option as soon as it can be made available.

However, the very fact that it has an award attached to it denotes that we do like it enough to recommend, and that says it all. The actual percentage score is made up of weighted categories based on technology, implementation, value, coolness/desirability, customer care, stability, and performance.

Seeing as the percentage scoring has been called into question, I'm going to revist the whole scoring metric and tweak it for following reviews, with a better appreciation of how we get to the final score.

[Faiakes]

I would like to see a greater and explicit emphasis put in real world performance, which is mostly connected to random (sustained) read/write speeds.

As The TechReport articles have shown, the Indilinix controller is great in the synthetic benchmarks used for PR but not that good in real life operations. Hence, I'd like to see drives evaluated on that accord.

[kalniel]

I would like to see a greater and explicit emphasis put in real world performance, which is mostly connected to random (sustained) read/write speeds.

As The TechReport articles have shown, the Indilinix controller is great in the synthetic benchmarks used for PR but not that good in real life operations. Hence, I'd like to see drives evaluated on that accord.

That's interesting - the Anandtech articles show the opposite, that the Indilinx is poor in synthetic benchmarks but good in real life operations.

[drunkenmaster]

The thing for me is, pcmark vantage get WAY to much exposure. According to that benchmark, the samsung based drives "appear" clearly ahead in gaming and vista loading, I mean their scores are what 20-30% ahead, yet in reality, the real world tests, the other drives beat the samsung based drives.

Well theres not much in the vista start up you can't argue that PCMark Vantage is right in any way shape or form. Gaming, well you did load times, didn't do min/max/average fps(which some benchmarks seem to suggest higher min framerate with ssd's) and only one game. PCMark could argue in this case their benchmark tries to simulate multiple different games. But again the INdilinx drives are faster than the Samsung while Vantage suggests it couldn't possibly be.

Basically sythetic benchmarks in this case, shows nothing.

The Samsung's show a clear advantage in small reads and a massive problem 4kb writes over the Indilinx. But what does that tell us, not much at all without knowing what you need. Is 1mb/s 4kb random reads enough, is 10mb/s, is 30mb/s where does that performance translate into real world performance, we have no clue.

With hard drives you really have to benchmark in real world, show for instance a Crysis load time, and min/max/average fps vs a normal hdd with the same setup. But also do this for another fps, a RTS, a racing game and, etc, etc, etc.

The other issue would be this, different applications work differently, different coders/teams write their games in different ways. The problem I have with PCMark Vantage is it tries to give an idea of real world performance, but its all writen by the same team with their own personal little ways of coding that will be carried across all their benchmarks. IE They might favour a coding structure that favours small reads, and this is carried across all their benchmarks, when in reality in 10 different real games small reads aren't very widly used.

Multiple benchmarks writen by multiple different people gives a better idea of relative performance than a suite of benchmarks all writen by one group of people.

We can see Vantage is fairly useless as, in lots of reviews of SSD's, infact reviews for months on end include Vantage which gives one impression of what drives are best, and the real world benchmarks in all these reviews, disagree entirely with the Vantage results, as yours clearly do also.

Its also very easy to find these drives listed at £370, making it massively cheaper than competing 256gb's, and suggesting the prices will drop massively is, well quite possible, but not all that likely. The 256gb is already quite a lot cheaper £/gb than the 128/64gb versions, which are all very much cheaper than competing versions. You could get £99 64gb Crucials last week, while the Vertex 60GB is between £180-200. The price drop, has realistically already happened, and while OCZ/Corsair/Intel may be dropping prices very soon to compete, Crucial's have no reason to.


Basically, I would absolutely kill to see the 64gb, the 128gb, and 2x64/2x128gb in raid 0 to see scaling, all in multiple real world tests, from gaming to vista loading to using a batch file to open 20 different apps and testing the time and some other things.

[GaryRW]

AHHHHHHH!!!!!

Apologies for repeating comments I've made in other SSD threads, but can we please have reviews of the lower capacity drives, just once in a while. As you repeatedly point out in the review, the drive is near identical to the OCZ Vertex drive. I can see the advantage of comparing to confirm Crucial haven't somehow dropped the ball, but I don't think many people would be surprised by the results.

I think the lower capacity drives are the ones a lot of people will be considering for system drives. I for one want to get on the SSD bandwagon as soon as possible, and see the best way of doing that is to buy the cheapest decent drive that's big enough for the OS, apps and a couple of games. Then in say a year's time, I'd plan on upgrading the main rig to have a more comfortable sized OS drive or to a faster drive (when the prices have gone down and/or performance has gone up).

At the moment, a 64 or 80gb would do it, but although they are officially rated with lower specs, there's very few reviews of the smaller drives out there. I'd like to know how much actual difference a drop of 40-50mbs in the rated sequential speeds would make to real world and synthetic benchmarks?

(I know you review what you're sent for the most part, but you must be able to apply a little pressure...)

Edit: & what drunkenmaster said while I was typing - reviews of 2 & (just to be silly) 4 of the 64gb drives in RAID 0 would be interesting. An article comparing 64gb & 256gb single drives against motherboard and/or add-in card RAID with multiple 64gb drives would be great

presumably proper add in hardware RAID cards could get round the 300gb/s sata 2 limit??

[drunkenmaster]

The very key question is how do 2x64gb perform compared to 1x128gb in things like game loading, obviously sequential transfers will be far higher, for a similar cost and a little inconvinience, but how does the bottom end performance in raid translate into real world performance.

I've caught between getting 2x64 and a single 128gb, a review with a 128gb raid 0 would also help, do 2x64's scale better than 2x128gb's, if so it would suggest that random read/write numbers obviously benefit being higher than mechanical hd's, but realistically theres only so much you need. IE, Intel's random write just trashes everything else, yet, its not signficantly faster in real world performance. As I was getting at whats enough, 2mb/s, 5mb/s 10/20/50 where does the real world performance stop showing any benefit.


I'm almost tempted to buy 2 of each just to test them myself Just waiting on that lotto to come up with the right numbers.

[Faiakes]

The very key question is how do 2x64gb perform compared to 1x128gb in things like game loading, obviously sequential transfers will be far higher, for a similar cost and a little inconvinience, but how does the bottom end performance in raid translate into real world performance.

Yes, that would have been interesting to see.

I've caught between getting 2x64 and a single 128gb, a review with a 128gb raid 0 would also help, do 2x64's scale better than 2x128gb's, if so it would suggest that random read/write numbers obviously benefit being higher than mechanical hd's, but realistically theres only so much you need. IE, Intel's random write just trashes everything else, yet, its not signficantly faster in real world performance. As I was getting at whats enough, 2mb/s, 5mb/s 10/20/50 where does the real world performance stop showing any benefit.

Oh, yes, good point.
A comparison focused on that would come very close to accurately assessing any SSD.

[Tarinder]

AHHHHHHH!!!!!

Apologies for repeating comments I've made in other SSD threads, but can we please have reviews of the lower capacity drives, just once in a while. As you repeatedly point out in the review, the drive is near identical to the OCZ Vertex drive. I can see the advantage of comparing to confirm Crucial haven't somehow dropped the ball, but I don't think many people would be surprised by the results.

I think the lower capacity drives are the ones a lot of people will be considering for system drives. I for one want to get on the SSD bandwagon as soon as possible, and see the best way of doing that is to buy the cheapest decent drive that's big enough for the OS, apps and a couple of games. Then in say a year's time, I'd plan on upgrading the main rig to have a more comfortable sized OS drive or to a faster drive (when the prices have gone down and/or performance has gone up).

At the moment, a 64 or 80gb would do it, but although they are officially rated with lower specs, there's very few reviews of the smaller drives out there. I'd like to know how much actual difference a drop of 40-50mbs in the rated sequential speeds would make to real world and synthetic benchmarks?

(I know you review what you're sent for the most part, but you must be able to apply a little pressure...)

Edit: & what drunkenmaster said while I was typing - reviews of 2 & (just to be silly) 4 of the 64gb drives in RAID 0 would be interesting. An article comparing 64gb & 256gb single drives against motherboard and/or add-in card RAID with multiple 64gb drives would be great

presumably proper add in hardware RAID cards could get round the 300gb/s sata 2 limit??

And we have. Crucial is sending us the 64GB model in the next few days. Review coming next week.