Crucial M4 256GB SSD Review (C400)

[alfaunits]

Tablets aren't using these SSDs, they use (cr)Apple style stuff (i.e. slow ) (remember the MacBook Air SSDs?)
After prices settle a little, I believe we will see a considerable price drop per GB, prolly in the are of 1.3$ and very likely 1.5$ very fast. One important thing ti consider is that the 25nm NAND requires more ECC area and a lot more spare area in order to have at least the same endurance. That adds cost.

BTW, someone mentioned these SSDs are meant for servers/enterprise? How so? Their endurance is not acceptable for server environments.

[Anvil]

There are a lot of implications.

Not only is the page size doubled, so has the block size, so 8KB * 256 blocks = 2MB vs 512KB on the 34nm plattform. (a huge shift)

As Johnw was saying there are other reasons for the 8KB move.

"As flash geometries get smaller, manufacturers are also moving to larger page sizes in order to
minimize the wiring overhead required to erase, program, and read the bit cells. Current flash devices
typically employ a 4KB page size, but manufacturers are likely to move to an 8KB page size in the next
generation of devices that will arrive in 2010." Link

The move to 8KB has been a known neccessity for years of course, a lot of documentation floating around.

As to the importance of 4KB block size, nothing changes the way we work, the 4KB is just as important as it used to be, however the 8KB page size can't be neglected.
It will make a difference for small files and just as important, it will make larger files (blocks) faster.

I'm trying to find some utility that counts/catalogues the files on my C drive (any drive) by file size, I wrote one such utility years ago (pre Windows) but it's not that easy anymore.
(Hard Links is making it harder to get 100% correct results)

edit:
Reviewed at TechReport as well, Link

Small file performance seems to be the recurring issue.
StorageReview have made the same comments about small files, their review is not up yet but preliminary AS SSD, CMD scores are available.

[Computurd]

You are NOT seeing ANY 8K. Those are 8 SECTORS, so 4K! Did it never occur to you there are no 0.5 lengths ?

thank you for pointing that out again after i had already acknowledged my mistake. read post #67. yeah, im not familiar with diskmon, so i made a dumb mistake.
your first post in your little four post tirade was at 12:44, the second was at 12:45, the third was at 12:54 then the last at 1:06...all in one line. quadruple post L:OL
you mean to tell me in the 22 minutes that you were reading this thread you didnt read my reply to overthere?

[mbreslin]

Can we stop the bickering now, and get technical?

[Ao1]

JDEC define an SSD Enterprise Endurance Workload as consisting of random data with the following payload size distribution:

512 bytes (0.5k) 4%
1024 bytes (1k) 1%
1536 bytes (1.5k) 1%
2048 bytes (2k) 1%
2560 bytes (2.5k) 1%
3072 bytes (3k) 1%
3584 bytes (3.5k) 1%
4096 bytes (4k) 67%
8192 bytes (8k) 10%
16,384 bytes (16k) 7%
32,768 bytes (32k) 3%
65,536 bytes (64k) 3%

http://www.jedec.org/standards-documents/docs/jesd219

[F@32]

JDEC define an SSD Enterprise Endurance Workload as consisting of random data with the following payload size distribution:

512 bytes (0.5k) 4%
1024 bytes (1k) 1%
1536 bytes (1.5k) 1%
2048 bytes (2k) 1%
2560 bytes (2.5k) 1%
3072 bytes (3k) 1%
3584 bytes (3.5k) 1%
4096 bytes (4k) 67%
8192 bytes (8k) 10%
16,384 bytes (16k) 7%
32,768 bytes (32k) 3%
65,536 bytes (64k) 3%

http://www.jedec.org/standards-documents/docs/jesd219

Ao1, please stop clouding the issue with facts

[billdavis]

Hey, can we now argue dependibility of the drive options out there?
Intel vs Crucial vs OCZ

[Computurd]

the issue there bill is the controller. i have found that when speaking reliability, you arent speaking to the manufacturer as much as to the underlying controller. for instance, kingston also uses the intel controller for ssd-now series. and many many use sandforce just the same as OCZ. sandforce is the most widely used controller.

so lets not talk manufacturers i guess cause that isnt fair....however two on your list do mfr their own controllers.

Intel is gold standard in reliability. period.

Crucial is damn near as good, the only thing they lack is the amount of time that the intels have on the market to 'prove' stability. intels have been out longer and have a bit more 'track record', but the crucials are just as reliable imo.

SandForce big problem is that the mfr's who sell them cannot give out destructive firmwares to fix 'panic-locked' drives. SF is tight as a virgin when it comes to their 'secrets', probably makes them look more unreliable than they truly are, however, it means if they get in a jam, you gotta RMA. SF will not allow destructive firmware flashes that are usually curative of the underlying issues 'out in the wild'.

[billdavis]

Cool!!
thanks

I like the crucial's personally but if intel came out with another E series SLC based ssd I'd prob buy that
but truth is that I plan on using my 1880 with many-ssd-raid0 array for a mac! :0

[cezar]

Crucial m4 vs. G3 (vs. V3, vs. C300):
http://www.tomshardware.com/reviews/...0,2908-11.html

m4 is the new king of the PCMark Vantage

[Brahmzy]

Crucial m4 vs. G3 (vs. V3, vs. C300):
http://www.tomshardware.com/reviews/...0,2908-11.html

m4 is the new king of the PCMark Vantage

Yet the Vertex owns the m4 in Vantage from other reviewers .... who's correct, and how can the results be so different?

[johnw]

Yet the Vertex owns the m4 in Vantage from other reviewers .... who's correct, and how can the results be so different?

There's a guy that works for Micron who goes by "adamantine piggy" who says that there is some sort of bug or compatibility issue with the m4 and Windows and some motherboards. Apparently if you revert from the Intel driver to MSAHCI and then reinstall the Intel driver, the m4 performance can be much improved. Weird and annoying, but it sounded credible. That guy posts in anandtech forums.

http://www.anandtech.com/show/4244/i...eview?all=true

Hey Anand, your 256GB Micron/Crucial M4 results on the Vantage HDD test seem to reflect an issue I have had with several P67 motherboards. On the 6G/s ports, on certain MB's, with the Intel SATA driver, it's not running at full speed. The only way I fixed it was to revert the the Intel driver back to the Win7 MSAHCI driver and then update again back to the Intel SATA driver. Why? I have no idea, but that drive should bounce up to the 60,000+ in the Vantage HDD test.

I noticed this issue with both a MSI and a ASUS P67 chipset MB. The 45K HHD is the exact same I got here in the Micron R&D lab before I did the MSAHCI swap and the revert back to Intel driver. The MSAHCI driver does about 55K with that drive. then the change back to the Intel driver suddenly bumps the core to 62Kish. This also affects all the other Vantage scores, but is most significantly seen with the HDD test.

Also, Anand's Intel 320 review had the m4 results included in the charts. The m4 came in second in Anand's PCMV:

http://www.anandtech.com/show/4244/i...d-320-review/8

[Highendtoys]

TweakTown's came out today as well.

http://www.tweaktown.com/articles/39...out/index.html

[johnw]

TweakTown's came out today as well.

http://www.tweaktown.com/articles/39...out/index.html

Look how much the m4 performance falls off when 25%, 50%, or 75% full:

http://www.tweaktown.com/articles/39...t/index10.html

This looks to me like it may be a firmware bug. I hope it is something that Micron can fix quickly. I certainly would not buy an m4 with such a performance drop off as it fills up.

[Machinus]

superbiiz is now sold out of all M4 drives at all capacities.

is that all until april 26?

[alfaunits]

Weren't those pre-orders anyway?

[Brahmzy]

Weren't those pre-orders anyway?

A few got sent out (accidentally I believe) - I don't think SuperBiz was supposed to let 'em go, but they let a few of the first ones get out.

[alfaunits]

Strange, I think those were pre-releases Get ready for firmware woes?

[Anvil]

Anand has published their review of the m4

Link
(about to read it now)

[hksnyper]

A few got sent out (accidentally I believe) - I don't think SuperBiz was supposed to let 'em go, but they let a few of the first ones get out.

I purchased one of the 256gb C400 drives from SuperBiiz. I've had it for a week. After I installed (on my sata2 mobo - Gigabyte G33m-ds2r), I ran CrystalMark and found the benchmarks for the 4k random read/write to be inline with the 3 or 4 reviews that I've seen posted. My sequential read/write were slower, but I attribute that to the sata2 mode.

The drive was oem without any cables or manuals. The drive only had a crucial label, no micron fancy blue label.

If anyone has some specific questions, I'll be glad to help.

-Rob

EDIT: I'm a little annoyed that the random read/write are slower than the C300, but price per gig is a little cheaper so that helps calm my nerves. Also, Keeping my fingers crossed that a future firmware will help the speed a little (and let's hope that we don't get another 006-like firmware)!

[alfaunits]

If I can get them 512GBs at $900, I'll bite

[WangChung]

Sigh... why don't they ever review the 64GB ones? More than a few people use the 64GB for their OS and use standard platter drives for their storage. Those of us that are less "Xtreme"

And to that end, why are the 64GB drives' performance so sub-par to the 128GB and 256GB?

[alfaunits]

The 64GB versions have less NAND chips, hence less possible NAND channels, which is why they can't do the same sustained performance. That's in general - in case of Intel it's intentional