SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Advanced Micro Devices - Moderated (AMD)
AMD 252.09-0.3%Jan 29 3:59 PM EST
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: etchmeister who wrote (139363)11/10/2004 5:49:12 PM
From: Dan3Read Replies (2) of 275872
 
Re: Placing two 1-Gbit OneNAND flash and one mobile SDRAM in a three-chip MCP

Well, if you need 256MB of storage, then the additional expense of interconnects between the 2 Flash chips and the DRAM chip don't add all that much to costs.

OTOH, there is also testing, and the fact that a problem with any of the 3 chips or any of the interconnects will probably make the whole module unuseable.

AMD's solution is far more elegant and should be less expensive to produce. Flash that acts like memory instead of disk or tape is nice stuff. Note that OneNAND still acts as as serial device and still requires something that can act as RAM.

AMD's solution is RAM that acts as its own storage.

Samsung's solution has some problems since they're going with SDRAM for working storage. SDRAM uses more power and requires a refresh controller. To have 256MB available from OneNAND you need 256MB of NAND, 256MB of SDRAM, and a DRAM controller (likely integrated into the chip) and you have to manage data transfers between the two. SDRAM also uses power while it's holding stored data.

AMD's solution is to just have 256MB of FLASH that functions as both storage and SRAM.

AMD's Mirrorbit is (I believe) more expensive to produce than Samsung's OneNAND but you don't need SDRAM, an SDRAM controller, or to worry about designing management hardware to handle the added complexity. A Mirrorbit solution should use quite a bit less power, too.

There will be a lot of cases where Mirrorbit is a better choice - and AMD will be able to charge more for Mirrorbit since it eliminates DRAM costs, uses less power, and reduces design costs.

Each technology will have its place. I doubt either one will crush the other. Samsung's advantage will be that it's cheap - but it had better be cheap, and there will be cases where a $20 easy-to-design-in low power Mirrorbit chip will win compared to a $5 difficult-to-design-in higher power OneNAND solution.
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  

HomeMailHotSubjectMarksPeopleMarksKeepersSettings
Terms Of UseContact UsCopyright/IP PolicyPrivacy PolicyAbout UsFAQAdvertise on SI
© 2026 Knight Sac Media.  Data provided by Twelve Data, Alpha Vantage, and CityFALCON News