Intel and Micron, who as partners of the joint venture IMFT, have nand flash memory, today announced that 128Gb is large, i.e. 16GB. The memory is on 20nm produced, makes higher speeds possible and is, in theory, cheaper.
Intel and Micron have produced previously, 20nm mlc nand flash memory, but with a size of 64Gb per chip. The new chips are available with 128Gb is twice as big and that was made possible by a new cell architecture in which the so-called ‘floating gate’ thinner. There is also use of the high-K+-technology, which means that the endurance of the memory chips remain the same as with 25nm chips. So they should be 3000-5000 schrijfcycli can endure.
It is striking that the pagesize of the memory is increased to 16KB. That has the consequence that the current ssd controllers new firmware need to with the memory handle. Also, the ONFi-interface has been updated, from version 2,x to 3, more bandwidth brings with it. That is now up to 333MT/s, while the current 25nm memory from IMFT, with 166MT/s. That will have repercussions on the read speeds of ssds, which are even higher.
Another advantage of the large chips is price. From a single wafer, more chips can be achieved, making it ultimately cheaper to take the chips to produce. Also, there may less surface, more memory can be placed. The ssd’s that Apple in its MacBook Air’s used can then also in regard to storage space increase. The mass production of 128Gb-flashchips only in the second quarter of 2012. The moment that the first ssds with such chips popping up will certainly take some time to emerge.
