New Crucial SSD

[deputc26]

http://www.tomshardware.com/ne...SSD-mlc-nand,8294.html

Any guesses on the controller it uses? I'm guessing Indilinx and that they fudged the write number.

 

[n7]

Sounds about right.

I'd like to see a new controller that actually dominates Intel's...seems nothing can do that so far though.

 

[Rubycon]

Originally posted by: n7
Sounds about right.

I'd like to see a new controller that actually dominates Intel's...seems nothing can do that so far though.

You're not going to see much emphasis on this in the desktop market since it would be of little benefit to desktop usage patterns.

 

[fleshconsumed]

If it can best X25-M Gen2, support trim out of the box, then 128GB for ~$300ish sounds like a bargain.

 

[rcpratt]

If.

I like how every other company overstates their read/write, and Intel actually understates it.

 

[LokutusofBorg]

Crucial is a division of Micron, which is part of the IM Flash joint venture with Intel. Lexar is owned by Micron. Nothing coming out of any of these companies is going to be substantially different from each other, except that they'll be tiered so they don't directly compete as much as possible...

 

[drizek]

The 64GB drive sounds good. It basically has the same cost/gb as the 128GB drive, hopefully it will be as fast. I don't care if they fudge the write speeds, even 120MB/s would be more than good enough. As long as there is no stutter issues and as long as the read speed is 75% as good as they say it is, I will be happy. It is just an issue of deciding whether it is worth buying now or waiting for them to get even cheaper and even faster.

 

[Denithor]

This brings up an issue I've been pondering for a while.

These manufacturers show numbers on their boxes that have little to do with actual product performance - they should be stating 4k random write performance, not sequential read or write numbers (which has very little impact on your daily use unless you just enjoy copying large files back and forth between drives).

 

[taltamir]

Originally posted by: fleshconsumed
If it can best X25-M Gen2, support trim out of the box, then 128GB for ~$300ish sounds like a bargain.

actually... if it can best intel gen2 with it having TRIM and intel NOT having TRIM than this is a strike against it. Because that means the relative performance of intel is gonna go up with trim. while its performance will remain the same.

 

[KeithP]

TweakTown has a review of one of the new Crucial SSD.

-KeithP

 

[rcpratt]

Originally posted by: KeithP
TweakTown has a review of one of the new Crucial SSD.

-KeithP

Looks pretty nice. Too bad they didn't review the X25-M G2, would have liked to compare the reviews. Hopefully Anand will do one.

 

[ss284]

Cliff notes: Indilinx drive, crucial 5 year warranty, decent pricing.

 

[alcoholbob]

The main issue is there is no standard benchmark for performance (compared to say, anechoic chamber for audio). You don't know what block size they used or which application they used to run those benchmarks which can affect the speeds by quite a bit.

The same would go for random performance. What's the queue depth? If you aren't comparing the same thing, one side will look artificially better than the other.

 

[Idontcare]

Originally posted by: Astrallite
The main issue is there is no standard benchmark for performance (compared to say, anechoic chamber for audio). You don't know what block size they used or which application they used to run those benchmarks which can affect the speeds by quite a bit.

The same would go for random performance. What's the queue depth? If you aren't comparing the same thing, one side will look artificially better than the other.

This is true but at the same time there is no excuse for the benchmarkers to be explicit about the specific test setup they used in generating the numbers they report on the side of the box.

Crucial could have just as easily claimed "200 MB/s sustained sequential write speed with 8192KB filesize" or "20MB/s sustained random write speed with 4KB filesize at queue depth of 1".

It is the intentional obfuscation of their test setup conditions that justifiably rankles the crowd that knows of the tricks they are up to.

 

[alcoholbob]

Occasionally some products will actually have an asterisk and at the bottom of the retail box in really tiny font it will specify the test conditions, although typically they are clearly one-sided test conditions.

 

[Rubycon]

Does MT have any influence on this product or is this another re-badged barefoot?

 

[deputc26]

Originally posted by: KeithP
TweakTown has a review of one of the new Crucial SSD.

-KeithP

I was really hoping they'd open it up and give us a look at the flash inside, is it possible that it uses the same 34nm Micron chips as Intel? I think it is possible but unlikely, just have to wait for another review I guess.

 

[Idontcare]

Originally posted by: deputc26

Originally posted by: KeithP
TweakTown has a review of one of the new Crucial SSD.

-KeithP

I was really hoping they'd open it up and give us a look at the flash inside, is it possible that it uses the same 34nm Micron chips as Intel? I think it is possible but unlikely, just have to wait for another review I guess.

Entirely possible, crucial buys their ram on open market though - whatever meets their price/performance design criterion - so if the internal stuff gets billed out to internal customers as being too expensive compared to samsung chips then they will use the samsung chips.

While crucial, via micron, might have access to the same source of underlying 34nm flash chips they do not have access to any of the logic controller IC's that Intel uses in their SSD's...so they might not be able to take advantage of the performance of the NAND chips anyways. Not much point in putting premium gasoline fuel into a ford focus sedan.

 

[jadelee]

Tnx a lot for sharing your thoughts! Now i solved all my questions.

 

[deputc26]

Originally posted by: Idontcare
Entirely possible, crucial buys their ram on open market though - whatever meets their price/performance design criterion - so if the internal stuff gets billed out to internal customers as being too expensive compared to samsung chips then they will use the samsung chips.

While crucial, via micron, might have access to the same source of underlying 34nm flash chips they do not have access to any of the logic controller IC's that Intel uses in their SSD's...so they might not be able to take advantage of the performance of the NAND chips anyways. Not much point in putting premium gasoline fuel into a ford focus sedan.

Would the barefoot controller not show some benefit from higher quality flash? It seems that lower quality flash (OCZ Agility) decreases performance by a small but measurable amount, so I would expect a small improvement in performance by going to better flash (by which I mean to imply that the performance increase would not be proportional to the leap in the flash's potential performance).

Of course my knowledge of a flash controller's interaction with flash chips is, at best, very limited.

 

[taltamir]

it would show a performance increase, early on flash quality was the main determinant, but newer controllers are all about parallelization and increase performance a lot more

 

[fleshconsumed]

http://www.legionhardware.com/document.php?id=850

New review with PCB shots this time. Indilinx controller with Samsung flash.

I kind of hoped this would be basically rebadged intel drive with tweaked firmware for faster sequential writes, but apparently this is simply Crucial version of OCZ Vertex.

 

[taltamir]

what is the point of a 10 page review that can be summed up with one line: "the new crucial SSD is just another indlinx drive"