On December 3, 2007, Samsung Electronics and Toshiba Corp. announced that they have licensed to one another, the product specifications and the rights to produce and market Samsung’s OneNAND and Flex-OneNAND, and Toshiba’s LBA-NAND and mobile LBA-NAND memory chips.
Both companies plan to develop compatible products based on the respective specifications to be released next year.
Our Take
After sitting on the sidelines and watching while ONFI amassed a membership roster of 57 companies since its establishment in May 2006, Samsung and Toshiba have finally acted and joined hands to develop products based on common specifications. Both companies are no strangers to collaboration, having shared common NAND specifications for years.
LBA-NAND, Toshiba’s solution for managing the increasing ECC requirements and bad block management of NAND flash by combining NAND flash memory with a controller in a single package, is expected to become mainstream for devices at 40nm and below. ONFI responded with their own version: BA-NAND. With Samsung’s backing, LBA-NAND is expected to be a very viable alternative to BA-NAND.
As for OneNAND, Samsung has been pressured by mobile phone manufacturers to license OneNAND to ensure a second source for the devices. Last year, Samsung licensed OneNAND to ST Microelectronics and now Toshiba. Toshiba and SanDisk (former M-Systems) jointly developed mDOC (mobile Disk-on-Chip) which is similar to OneNAND but based on MLC NAND. Both products integrate a NAND core, SRAM, error correcting engines, and logic circuits in a single chip with a NOR interface. In fact, OneNAND was developed by Samsung based on IP licensed from M-Systems with the licensing agreement subsequently terminated in 2005. mDOC has never been a significant product line for Toshiba and resource-wise, it does not make sense for Toshiba to invest in two similar but incompatible, competing solutions. It appears opting to support OneNAND at the expense of mDOC is the price Toshiba is willing to pay to garner Samsung’s backing for LBA-NAND.
Subscribe to:
Post Comments (Atom)
Posts
5 comments:
Good article. I would love to know your take on the current state of this debate. Do you think ONFI and the S&T standard are both competitive? What do you see as the major advantages and disadvantages of each.
ONFI-compliant chips are not really backward compatible with the standard asynchronous NAND interface because the W/E and R/E pins are replaced with a clock and direction pin.
Samsung has introduced their synchronous toggle-NAND to JEDEC. Toggle-NAND treats the W/E and R/E pins as pseudo clocks so that data can be transferred on the rising and falling signal edges of the clock. Toggle-NAND retains the asynchronous NAND interface.
Intel and Micron are the only suppliers to offer ONFI-compliant parts. Together they account for 10% of the market. Toshiba is likely to back toggle-NAND and Samsung and Toshiba combined have 2/3 of the market.
Hey Greg,
I respect your work very much. Well worded talent goes far in the journalism career. Keep up the good work, so far I've clearly understood and followed up with your writings and I just want to throw some kudos at you, very good to hear people putting their mind to words the clear way :)
Anyways, until the next time I run across your page, c ya' ciao!
Hi Greg,
Good article.
I referred to ONFI2.1 standard, table 12. It mentions that in asynchronous mode (traditional) nand flash mode 'RE and WE' still retain their functionality. It is changed only in source synchronous mode.
In that case it appears that it is back compatible. Please correct if I am missing something.
Pretty cool site you've got here. Thanks for it. I like such topics and everything connected to them. BTW, try to add some photos :).
Post a Comment