Corsair Performance Series™ Pro 256GB SSD Review @ [H]

FrgMstr

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,596
Corsair Performance Series™ Pro 256GB SSD Review - Today we review the 256GB Corsair Performance Pro SSD. Corsair provides enthusiasts with both sides of the SSD controller coin by offering both Marvell and SandForce controlled SSDs in its product lines. Today we will take a look at the Marvell powered Performance Series™ Pro 256GB SSD, and the Toshiba Toggle NAND that Corsair has chosen for it.
 
This is our first go at SSDs.....again, this time with a new writer and a new format. Any and all constructive input is welcomed. Help us make the content better by telling us where we blew it and where what we need to be covering to make a better SSD review. Thanks.
 
Well the Crucial M4 256gb is $200 shipped on Amazon to my door. Would I pay $140 more for a bit more performance? No.
 
DJ Big T said:
Well the Crucial M4 256gb is $200 shipped on Amazon to my door. Would I pay $140 more for a bit more performance? No.
Click to expand...

Although this post could have been more tactfully worded, I would tend to agree with it. $140 more for a performance difference that is pretty damned near inconsequential... heck I'm going to go with the M4.

For under $400 I can grab two 256GB M4's and be ahead of the game in both speed and capacity. Now if Intel would just get off their asses and release that magical new RST update that's supposed to get us TRIM in RAID....


Great review, nonetheless!
 
You know you guys make a damn good point and I totally looked up the M4 price incorrectly during the edit. I have made amends to the conclusion and pointed out the price disparity and adjusted the award as well.
 
The only SATA controller chipset I have ever had issues with is Marvell. I just wouldn't trust them to do a quality job.
 
Thanks for the review and thanks for adjusting the rating. An M4 is by far a better deal, but it won't be around for long since they are just cleaning out stock. Not sure what to make of the price right now. Hopefully we'll see price drops in the future.
 
DooKey said:
Thanks for the review and thanks for adjusting the rating. An M4 is by far a better deal, but it won't be around for long since they are just cleaning out stock. Not sure what to make of the price right now. Hopefully we'll see price drops in the future.
Click to expand...

Okay that's great but there is the Plextor for $199 AR: http://www.newegg.com/Product/Product.aspx?Item=N82E16820249015
The Samsung 830 series: http://www.newegg.com/Product/Product.aspx?Item=N82E16820147164

etc. There are plenty of 256GB SSDs for less than $340 shipped. Easily. Hands down. I just can't justify that price.
 
This was a very well-done review. It was in-depth and very informative. For a person relatively new to computer hardware, this review should definitely help when it comes to looking at SSD as you're covering everything about what makes a good SSD a good SSD.

Though it is a little disheartening to see my ol' favorite, the Crucial M4, not handle so well. But oh well.

Oh and congratulations to the new writer!
 
Blue Falcon said:
Now if Intel would just get off their asses and release that magical new RST update that's supposed to get us TRIM in RAID....
Click to expand...

No kidding, where the hell is that? I've been waiting for it for months now. I actually just posted about it here yesterday but the thread fell into obscurity.

Does Kyle or anyone else have a contact at Intel they can ask?
 
Danny Bui said:
This was a very well-done review. It was in-depth and very informative. For a person relatively new to computer hardware, this review should definitely help when it comes to looking at SSD as you're covering everything about what makes a good SSD a good SSD.

Though it is a little disheartening to see my ol' favorite, the Crucial M4, not handle so well. But oh well.

Oh and congratulations to the new writer!
Click to expand...

Huh? The M4 beat out the Performance Pro in a significant number of those benchmarks.
 
Thanks for the interesting review, but I wanted to throw in some info about the user experience like the [H] does with GPUs.

I've used several SSDs and decided on a Crucial M4 for the combination of price and performance. I guarantee that a large majority of people won't notice the difference between these drives outside of a lab. There is absolutely no way to beat the price and performance of the M4 at this point in time.

That said, it will be interesting to see how Marvell's next generation controller performs. But in terms of user experience, it's nothing like GPUs. You will be just as happy you got a Crucial M4 as a Corsair Performance Pro...but you'll also have more money to spend on something else.
 
Great informative article.. thanks! Just confirms my decision for the M4 for my uses.

On a side note.. just nitpicking... 3 shades of red on the graphs? All out of other colors? :p
 
I didn't like the graphs but otherwise I thought it was a pretty solid article.
 
Nice review. 2 Nits to pick:

1. Several bar graphs say "curcial" instead of "crucial"
2. I don't see anywhere a reference to the height of the drive. You list form factor as 2.5 inch, which is a bit like describing a car's physical dimension as "has 4 wheels". Not quite enough info. With many newer laptops only having room for the smaller 7mm height drives, listing the height and specific part number for that size would be very helpful. Some of the SSDs have 2 versions of each drive with different part numbers, the only difference being 7mm or 9.5mm drive height. So listing this info on each SSD review could be helpful to peopel considering using any particular SSD in a laptop.
 
Last edited:
Another readibility issue:

On the first few bar graphs, the way the drives are listed end up having the Performance pro as the left-most drive, leading to it being the top bar on the graph. A page or two later, the order is reversed so the performance pro is on the bottom. This is confusing. The order should be the same with the reviewed drive either the top or bottom one on all graphs, not the top one on some and the bottom on others.

It's a minor thing but I had to go back and re-read 2 pages because reading through the first time I thought the performance pro was always the top bar, and then the text didn't match what I thought the graphs were saying.
 
Excellent article. I wasn't too keen on the different shades of red. Make them all different colours, maybe make each one the colour that the drive has on it's casing sticker, Intel white, crucial blue, etc.
 
For the record, the Performance Pro is 9mm.

The Performance Pro and the Plextor M3P/M3S/M2Ps are all manufactured by Lite-On, Plextor's parent company. They are all interesting drives, and the M3S 256GB has been $200 AR as well. Still, Samsung, Micron/Crucial, and Intel are really able to put the pedal to the floor when it comes to price. Owning the NAND fab delivers many benefits, but pricing flexibility is certainly a big one.

All the same, many enthusiasts have been digging the Marvell 9174 and Toggle NAND combo, and many longtime SSD users have been buying them despite their price. Most users would be well served by the M4, but those already into SSDs for several generations are looking for something different.
 
On the conclusion page, the icon at the bottom of the page is a Silver Award. On the list of SSD reviews, the icon is for a Gold Award. The text of the article does not state what award it got. I see in this thread that you changed the award, so I am guessing it's Silver now?
 
evilsofa said:
On the conclusion page, the icon at the bottom of the page is a Silver Award. On the list of SSD reviews, the icon is for a Gold Award. The text of the article does not state what award it got. I see in this thread that you changed the award, so I am guessing it's Silver now?
Click to expand...

I will get that fixed. Yes, Silver.

LanPirate1 said:
Excellent article. I wasn't too keen on the different shades of red. Make them all different colours, maybe make each one the colour that the drive has on it's casing sticker, Intel white, crucial blue, etc.
Click to expand...

Tried that, those look like ass. Another suggestion?

flenser said:
Another readibility issue:

On the first few bar graphs, the way the drives are listed end up having the Performance pro as the left-most drive, leading to it being the top bar on the graph. A page or two later, the order is reversed so the performance pro is on the bottom. This is confusing. The order should be the same with the reviewed drive either the top or bottom one on all graphs, not the top one on some and the bottom on others.

It's a minor thing but I had to go back and re-read 2 pages because reading through the first time I thought the performance pro was always the top bar, and then the text didn't match what I thought the graphs were saying.
Click to expand...

We can get that fixed.

flenser said:
Nice review. 2 Nits to pick:

1. Several bar graphs say "curcial" instead of "crucial"
2. I don't see anywhere a reference to the height of the drive. You list form factor as 2.5 inch, which is a bit like describing a car's physical dimension as "has 4 wheels". Not quite enough info. With many newer laptops only having room for the smaller 7mm height drives, listing the height and specific part number for that size would be very helpful. Some of the SSDs have 2 versions of each drive with different part numbers, the only difference being 7mm or 9.5mm drive height. So listing this info on each SSD review could be helpful to peopel considering using any particular SSD in a laptop.
Click to expand...

1. We can get that fixed.

Dan_D said:
I didn't like the graphs but otherwise I thought it was a pretty solid article.
Click to expand...

Suggestions please? Maybe a link to something you like?

Evil Scooter said:
Great informative article.. thanks! Just confirms my decision for the M4 for my uses.

On a side note.. just nitpicking... 3 shades of red on the graphs? All out of other colors? :p
Click to expand...

Please link me to something you think is appropriate?


I appreciate all the feedback guys! We will get this smoothed out as we move forward.
 
I appreciate the feedback as well, and will grow from it!
Pricing is always a concern, and the market right now is explosive. The NAND spot market is more volatile than the gasoline market at times, and I was taking this into consideration on the pricing. The M4 can be cheap today, but higher tomorrow, and will also disappear from the market very soon.
The M4 is a solid bet, I am a big fan of them as well.
Another thing to take into consideration is that a comparable entry from Intel in the form of the 520 is similar in price, but suffers read and write degradation to a higher degree.
Definitely some ups and downs, and a hard one to call, but I do agree that pricing should trump many other factors when dealing with the same generation of SSDs. I am very partial to Toshiba Toggle NAND though!
The graphs are certainly an issue moving forward, and we will work to get this right. I think that the main three issues that I see being commented on are;
1. color
2. 'flipped' results (this is easy to fix, and I should have caught that!)
3. Spelling (OUCH)

Outside of these three issues, as a whole did you find the graphs informative?
 
Last edited:
The graphs are okay, and I don't have a problem with all red graphs, but save the lighter shade of red for the drive being tested as contrast to the comparison drives. For instance, the power consumption testing has the 520 as the highlighted drive, and not the PP. It would help reader comprehension if the current drive was highlighted via the lighter shade of red over the other drives.
 
Nice Review, I bought the Corsair Force GT SSD 120GB based on your review. Love the drive, I really can't see a need to go faster, but a comparison with a the older drives would do wonders to the review,

If I find my uber drive is now 50% slower, maybe I'll pony up some bucks for the next gen, but at this point I keep what i have.

Cheers
 
DJ Big T said:
Well the Crucial M4 256gb is $200 shipped on Amazon to my door. Would I pay $140 more for a bit more performance? No.
Click to expand...

Thanks for the heads-up. Was looking to order one and it saves me around 20$ :D

Review was good, will also echo that the red scheme for the bars is something to change.
 
Since the argument for using ssd is using them in games, what would be the difference in loading in seconds between the fastest and slowest of these SSD's and that of a normal slow 5400 rpm hdd or better?
 
I am surprised no one seems to have run into issues with the M4...

Mine died/having issues less then a year. MANY people using XBMC have had issues/failures. Common to see posts on the M4 on crucials forums. Sometimes issues can be resolved by reseting the hardware (# of powercycles without data usage).

Now some of these issues could be linux related or some other reason, I just know the moment I posted my hardware specs (after replacing almost everything else lol)... number of people immediately jumped in and said they had drive drops, locks ups with the M4...
 
Good review. I concur with the sentiment on SSD pricing being very volatile at the moment. While this is a great SSD, a rebate would help in the pricing department.
 
When you do your test do you turn anything off?
Example
1. Drive indexing DISABLED
2. Prefetch disabled
3. Super fetch disabled
4.Defrag disabled
When I was testing SSD a while ago making sure these were not running helped a lot on some SSDs and some I could not tell a difference. It's just a thought don't know if you do this or not.
 
Also what would be cool for testing SSDs is testing it in single mode and in Raid 0 if you can get two that is. I know a lot has to do with controllers too. I think it would be nice to see how some preform in a raid 0. Maybe a chart on best to worst single drive and raid 0 drives.
 
soulpunisher said:
Also what would be cool for testing SSDs is testing it in single mode and in Raid 0 if you can get two that is. I know a lot has to do with controllers too. I think it would be nice to see how some preform in a raid 0. Maybe a chart on best to worst single drive and raid 0 drives.
Click to expand...

Controller performance is literally all over the map. They are almost as variable as the firmware / controllers on the SSD's themselves.
 
soulpunisher said:
When you do your test do you turn anything off?
Example
1. Drive indexing DISABLED
2. Prefetch disabled
3. Super fetch disabled
4.Defrag disabled
When I was testing SSD a while ago making sure these were not running helped a lot on some SSDs and some I could not tell a difference. It's just a thought don't know if you do this or not.
Click to expand...

WEI will disable most relevant things automatically when it detects a SSD. You also want to fully disable all power-saving features (use ThrottleStop to achieve this firmly) as some can impact 4k performance.

As for the benchmarks, CrystalDiskMark and Anvil are both also commonly used, although the latter is more reputable these days for accurate results. Generally speaking, the baseline metric is incompressible and non-cached data, so that's the place to start with any results, although Atto and HDTune are useful for sequential. Performance should be tested at 0%, 25%, 50%, and 75% usage.

It's nice to see SSD reviews on this site. I've done boatloads of research on it but it would be nice not to have to always do that with new hardware generations in the future. As for the person who mentioned prior Corsair GT review, it greatly influenced me towards a buying decision that after research I know was only partially valid, so there's room for improvement in the SSD scene.
 
Nice to see some SSD love on hardocp - for the price though there are many better options. Sure this thing hauls ass, but not $140 dollars worth of ass.

Considering I just say a 256GB SanDisk Extreme for 179.99 on amazon/slickdeals...

Love the extreme drops that SSD's are seeing - I'm in the market for one, but can skate by on my 64gb model for another month or two. If the trend continues if I can get a 256GB SSD for 150-160 I'm on board.

If I had money falling out of my butt I'd love to pick up a 256 OS Drive and a 256 for my Steam games. Truth be told though - even if I get into a battlefield game 15 seconds quicker - it doesnt matter anyway because I'll be sitting there waiting for the actual game to start on the server. Single player though - I can see where this would be awesome -- too bad I still havent run through the campaign yet. lol
 
Great review, very informative. Do not like the graph colors. Question though, Since the M4 is on the way out what does Crucial have in line to replace it?
 
Maxx said:
WEI will disable most relevant things automatically when it detects a SSD. You also want to fully disable all power-saving features (use ThrottleStop to achieve this firmly) as some can impact 4k performance.

As for the benchmarks, CrystalDiskMark and Anvil are both also commonly used, although the latter is more reputable these days for accurate results. Generally speaking, the baseline metric is incompressible and non-cached data, so that's the place to start with any results, although Atto and HDTune are useful for sequential. Performance should be tested at 0%, 25%, 50%, and 75% usage.

It's nice to see SSD reviews on this site. I've done boatloads of research on it but it would be nice not to have to always do that with new hardware generations in the future. As for the person who mentioned prior Corsair GT review, it greatly influenced me towards a buying decision that after research I know was only partially valid, so there's room for improvement in the SSD scene.
Click to expand...

Great INFO TYVM I will try that software ThrottleStop.
 
Wumpus said:
[notes some nit picking above]

"This leads to a floating point of performance depending upon the compressibility of the data that is presented to the SSD during use"

I thought this was a word filter change at first, but I suppose it sort of makes sense. Kind of jarring to any geek who knows what "floating point" is and how compression works.
Click to expand...

Thanks Wumpus, you are right that it might sound like some funny wordplay :)
"This leads to a highly variable point of performance depending upon the compressibility of the data that is presented to the SSD during use"?
Not sure if that sounds better, a little partial to the first version lol, especially since its slightly clever...

ClearM4 said:
Good review. I concur with the sentiment on SSD pricing being very volatile at the moment. While this is a great SSD, a rebate would help in the pricing department.
Click to expand...

The pricing is something to take into consideration for sure, but it would be a very boring test and review process if the lowest price SSD always wins!
I am thinking of checking price histories and taking that into account for comparing prices, as they shoot up and down so dramatically. With value being such an important metric, I think we must find a better way of comparing prices than taking the 'right this moment' approach. Of course it is impossible to predict the future, but taking the past price histories of SSDs could be a viable approach? not sure, but very open to others' thoughts on this!

soulpunisher said:
When you do your test do you turn anything off?
Click to expand...

Yup, but nothing near the ridiculous amount of 'tweaks' that we see on many guides. I take a very minimalistic approach to SSD tweaking, as most really isnt needed. The majority of recommended 'tweaks' are akin to voodoo imo, mostly aimed at fear-mongering in regards to endurance concerns.
This would actually make a great topic for a future article! (noted)


soulpunisher said:
Also what would be cool for testing SSDs is testing it in single mode and in Raid 0 if you can get two that is. I know a lot has to do with controllers too. I think it would be nice to see how some preform in a raid 0. Maybe a chart on best to worst single drive and raid 0 drives.
Click to expand...

RAID is definitely in our future if we can get the samples.

Maxx said:
WEI will disable most relevant things automatically when it detects a SSD. You also want to fully disable all power-saving features (use ThrottleStop to achieve this firmly) as some can impact 4k performance.

As for the benchmarks, CrystalDiskMark and Anvil are both also commonly used, although the latter is more reputable these days for accurate results. Generally speaking, the baseline metric is incompressible and non-cached data, so that's the place to start with any results, although Atto and HDTune are useful for sequential. Performance should be tested at 0%, 25%, 50%, and 75% usage.

It's nice to see SSD reviews on this site. I've done boatloads of research on it but it would be nice not to have to always do that with new hardware generations in the future. As for the person who mentioned prior Corsair GT review, it greatly influenced me towards a buying decision that after research I know was only partially valid, so there's room for improvement in the SSD scene.
Click to expand...

by "performance should be tested at 0%, 25%, 50% and 75% usage, i believe you are referring to fill capacity?
if so, yes these are interesting numbers. As we get further along with firmware and controller maturation's we are seeing less and less performance variability in regards to fill level testing. My bet is the next gen there will be even more minute variability once you get up over 25% fill. Pretty much all testing should be conducted at 25% fill imo, and that is something we will take into consideration during testing. 25% fill with static data would make for an interesting angle.
The parameters of this testing regimen aren't entirely complete yet, so there will be some evolution as we move forward.

cfdfireman said:
Great review, very informative. Do not like the graph colors. Question though, Since the M4 is on the way out what does Crucial have in line to replace it?
Click to expand...

The M5 should be announced sometime in the near future, but do not quote me on that, honestly I am not privy to any insider info on that one. that is just the 'buzz'.
 
HughBriggs said:
by "performance should be tested at 0%, 25%, 50% and 75% usage, i believe you are referring to fill capacity?
if so, yes these are interesting numbers. As we get further along with firmware and controller maturation's we are seeing less and less performance variability in regards to fill level testing. My bet is the next gen there will be even more minute variability once you get up over 25% fill. Pretty much all testing should be conducted at 25% fill imo, and that is something we will take into consideration during testing. 25% fill with static data would make for an interesting angle.
The parameters of this testing regimen aren't entirely complete yet, so there will be some evolution as we move forward.
Click to expand...

Thanks for posting back. I was speaking in quick and general terms, but yeah, I meant fill capacity. The reason I use % specifically is because some drives over-provision in a way that reduces capacity (240GB vs. 256GB, for example). This can be mitigated by superior garbage collection. In other words, while it may seem the 256GB drives are larger than the 240GB ones, in reality they use the same memory, often in both type and size. The difference is in the controller and how it handles garbage collection with the free and (in some cases) provisioned space. This also impacts methodology in "cleaning" a drive as you can't fill SandForce ones with 0's or 1's to refresh it, for example...

Additionally, as far as filling a drive for testing, the consensus is that you rely on purely random (incompressible + non-cached, as I mentioned) data to do so. You can do this over an entire free area and issue Trim commands (newest Anvil benchmark enables sending of Trim) to work with this. I own SandForce myself so I lean towards "real-world" data (46% incompressible), but for a baseline benchmark you do want to take the approach I mention. For more information on this methodology, please see:

http://www.snia.org/forums/sssi/pts

As for the people asking about RAID-0, keep in mind that there are Intel drivers and OROM's available to enable Trim to pass-through. However, this makes the RAID a SCSI device, and only Windows 8 can pass Trim through SCSI at this time.
 
The SNIA methodology is the basis of our steady state testing, and several concepts and approaches are actually used from the spec to conduct the tests. (128 k fill twice, measurement of steady state)
Incompressible data is key for conditioning and static data as well, i definitely agree there!
there are currently two 'classes' for SNIA testing, which is enterprise and consumer. Enterprise being much more rigorous. The SNIA specs are unfortunately not 'complete' yet.

Unfortunately with the last two versions of firmwares from SF, TRIM is in fact not working, leading some reviewers to have stacks of SSDs waiting to be tested.

The building blocks for TRIM in RAID have been laid, now just for the final implementation :)
 
You must log in or register to reply here.
Back
Top