Intro

We're looking at SSDs for a few reasons

They can provide roughly 2-2.5x the read bandwidth (and sometimes nearly that much for writes, too) of disks.
They have much lower write latency (RAMCloud may be attractive without battery backups/Agigatech stuff/etc, if we use synchronous flash writes to get durability.
We don't need huge capacity, so the cost is reasonable (~$1.6/GB)

Controllers

There appear to be half a dozen different manufacturers of SSD controllers (Marvell, Intel, JMicron, Samsung, Indilinx, Toshiba, Sandforce).

Troublingly, some controllers or associated firmwares appear to be considerably less reliable than others (judging from consumer reviews on, e.g., NewEgg.com). E.g. the Indilinx and JMicron ones seem especially suspect. Sandforce-based drives also look a little sketchy depending on the vendor (perhaps they're making stupid firmware modifications?). Even Intel appears to ship buggy firmware!

Controller Manufacturer	SSDs
Marvell 88SS9174	Intel 510 series, Plextor M2, Crucial M4
Intel	All Intel SSDs aside from the 510 series
Samsung	Apparently Samsung drives, only
Toshiba	Only consumer drive appears to be Kingston SNVP325 series (though also used by OEMS, e.g. Apple)
Indilinx	Tons of random consumer drives (e.g. OCZ)
JMicron	Tons of random consumer drives (e.g. Kingston, OCZ)
Sandforce	Lots of random consumer drives

Drives We're Evaluating

We looked at 4 drives with chips from Marvell, Intel, Samsung, and Toshiba:

SSD	Sizes	Controller (purported)	Tested Firmware	Latest Firmware	Flash Chips (*)	Part Numbers	Link
Intel 320 Series	80GB, 120GB	Intel PC29AS21BA0	0302	??	Intel/Micron 25nm 29F16B08CCME1	SSDSA2CW080G3B5, SSDSA2CW120G3B5	Product, Firmware Bug Discussion
Samsung 470 Series	64GB, 128GB	Samsung S3C29MAX01	AXM09B1Q	AXM09B1Q (5/5/2011)	Samsung 32nm(?) K9HDGD8U5M	MZ-5PA064, MZ-5PA128	64GB, 128GB
Crucial M4	64GB, 128GB	Marvell 88SS9174	0002	0002 (6/20/2011)	Intel/Micron 25nm 29F128G08CFAAB	CT064M4SSD2, CT128M4SSD2	64GB, 128GB, Firmware
Mushkin Calisto Deluxe	60GB, 120GB	Sandforce SF1200	3.6.1	3.6.1 (4/28/2011)	Intel/Micron 34nm 29F64G08CAMDB	MKNSSDCL60GB-DX, MKNSSDCL120GB-DX	60GB, 120GB

(*) Guess based on various hardware review site tear-downs.

Read Performance: Random 8MB reads

This experiment targets recovery -- how quickly can we read 8MB segments from the disk?

The result is boring -- all disks can average between 260 and 270 MB/s (which is apparently about all we can expect of our 3Gbps SATA ports).

-TODO- 8MB latency may be interesting. If we have lots of nodes, we may only read a few segments. Hopefully that offers just as good bandwidth as reading a lot of them.

Write Performance: Random 8MB writes to some fraction of the disk

This experiment writes 8MB segments randomly to the first x% of the disk, for varying values of x. The total data written in each test is twice the drive's reported capacity.

The idea is to test how well write performance holds up under increased space utilisation. For example, if we write 120GB to the first 20% of the disk's logical blocks (randomly overwriting), how does the drive fare? We'd hope that when utilisation isn't too high, the drive's FTL would be able to efficiently erase recently overwritten LBAs and keep up with the write load.

Before each run, the disks were trimmed using hdparm. Perhaps a secure erase would have been better, but most drives appeared to regain full write performance this way. Notably, the Mushkin one did not (the 100% test was first, followed by 20,40,60%...).

Samsung 470 Series (Samsung Controller)

Crucial M4 (Marvell Controller)

Intel 320 Series (Intel Controller)

Mushkin Callisto Deluxe (Sandforce Controller)

Browser not supported