A small Israeli start-up is claiming that its novel signal processing algorithms can enhance the reliability of cheap MLC-based NAND flash memory to the point where it is indistinguishable from the more expensive SLC-based NAND - and it has launched an SSD to prove it.
Multi-Level Cell (MLC) flash memory is significantly cheaper to manufacture than Single-Level Cell (SLC), but comes at a cost: decreased reliability owing to increasing noise levels as the drive ages. As a result, consumer grade SSDs tend to come with MLC flash, while enterprise-grade SSDs get SLC - and a price tag to make your eyes water.
According to
Computerword that could all change thanks to Anobit Technologies, which has launched its first SSD product to feature the Memory Signal Processor. This processor features a new error correction algorithm designed to clean noise from the signals received from MLC flash chips.
The specifications are certainly impressive enough: with traditional MLC-based drives normally wearing out around the 3,000 write cycles point, the Memory Signal Processor allegedly boosts Anobit's Genesis SSD line to 50,000 write cycles - a far better proposition for those wanting to add an SSD to their machine. Performance doesn't seem to suffer, with Anobit claiming a sustained read speed of 220MB/s and write speed of 180MB/s.
The Genesis SSD line with Memory Signal Processor will be available in 200GB or 400GB sizes, both featuring a SATA interface, but the company's mention of an external bridge which allows the drives to connect via SAS or fibre-channel suggests that Anobit is aiming these at the budget-conscious corporate market rather than the high-end consumer market. Its refusal to offer pricing information beyond claims it will be 'very competitive' suggests that home users won't be enjoying increased reliability just yet.
Are you pleased to see companies working on enhanced reliability for SSDs, or will you wait for the technology to trickle down to consumer-grade hardware before you start celebrating? Share your thoughts over in the
forums.
Want to comment? Please log in.