The New (Profitable) Ramtron Message Board

STOCKTALK

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 : The New (Profitable) Ramtron

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next

To: poster1234 who wrote (108)	10/28/2007 3:22:03 PM
From: sleupendriewer	Read Replies (1) of 647

yes that is possible - but (!!!) at least on the 8MBit-chip in the mentioned paper they have also additional logic to replace defect rows columns or even up to 128 single biterrors. This is quite evolved with repsect to DFM (design for manufacturing ;) - I assume you know this) - I have seen this in no other MBit-chip in any paper with such gimmicks - everyone with only a slight tech baxckground should realy read the paper if you haven't - it's realy a good one ... (p.s.: what I also like about this circuitry is that it might be possible by that in a mcu lateron to find in a fram-program (!!!)- not just data - memory, the cells which are mostly used - a short example : if you have program memory and a program (that is totaly sequential) that executes on each cell every 1000 cycle then you will have even at a cycle frequency of 40 MHz no problem to reach a lifetime of 40 years - BUT if you have a loop or just a short program that executes mostly on some or just one cell - then it's easy to calculate that you get into trouble after some time - BUT if you can find this cell/row/or column and "repair"(substitute) it with a new one when the mostly used cell gets "tired" - then you have no problem anymore - and afaik this is what they implemented (without the internal testing routine as far as I got it - at least they haven't described it in the paper) in the 8MBit - so this is imo not only good for configuration for increasing the yield during production-testing - this can also lead to totaly new devices - like for example enhanced versas with just fram-inside for program and data memory ... my opinion - if I got something wrong - let me know ...) Because of that effort I assume - if your assumption on the selection and configuration of devices during testing is right - it's just a matter of time - and we will see a 8MBit-device - just by part selection during testing (the good ones get 8MBit and the other 4MBit - if both blocks are crap - what a pity ;)) and we might also see (maybe) some kind of mixed calculation for the 4MBit device - because if things are equal for the 2MBit - than this might work with 2 blocks enabled also as a 4MBit dvice - and by that they would get a mixed calculation of good 2x2Mbit- and 1x4Mbit-devices - this might decrease pricing then for the 4MBit ... just speculation based on some information from the web, but realy funny, isn't it ...

Report TOU Violation

Share This Post

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next