| ??? 11/19/03 17:25 Read: times |
#59039 - RE: Failure rate estimation Responding to: ???'s previous message |
Werner wrote:
ATMEL will confirm, that their FLASH technology is the same in all devices. Erik wrote: I doubt it, the trend is towards smaller geometries and simpler flash cells. I agree, Erik. This degradation with higher integration can even be seen from raliability data: AT89C55 has 5 times bigger memory space than AT89C51, but data retention failure rate is 15.4FITS / 2.2FITS = 7 times higher. By the way, I was asked, why MTTF of 7,413years for AT89C55 is so much higher than specified data retention time of 10years. Well, ATMEL specifies 10years data retention time for all their memory and microcontroller products, when I remember correctly. If we now assume Tu = 125°C and a confidence level of 99% we get: Af = exp [Ea / k x (1 / Tu - 1 / Tt)] = exp [0.5 / 8.6171D-5 x (1 / (125 + 273.16) - 1 / (150 + 273.16)] = 2.4 lambda = chi-square / 2 / Af / t x 1D+9 = 9.216 / 2 / 2.4 / 480,000 x 1D+9 = 4000FITS Which means a MTTF of 28.5years. This is in the same order of magnitude as 10years. And we have still some headroom for endurance induced increase of data retention failure rate. Kai |
| Topic | Author | Date |
| Data Retention in Flash 10years ? | 01/01/70 00:00 | |
| RE: Data Retention in Flash 10years ? | 01/01/70 00:00 | |
| RE: Data Retention in Flash 10years ? | 01/01/70 00:00 | |
| RE: Data Retention in Flash 10years ? | 01/01/70 00:00 | |
| Soooooo many times discussed here! | 01/01/70 00:00 | |
| RE: Soooooo many times discussed here! | 01/01/70 00:00 | |
| RE: Soooooo many times discussed here! | 01/01/70 00:00 | |
| RE: Soooooo many times discussed here! | 01/01/70 00:00 | |
| RE: Soooooo many times discussed here! | 01/01/70 00:00 | |
| RE: Soooooo many times discussed here! | 01/01/70 00:00 | |
| Failure rate estimation | 01/01/70 00:00 | |
| RE: Failure rate estimation | 01/01/70 00:00 | |
| RE: Failure rate estimation | 01/01/70 00:00 | |
RE: Failure rate estimation | 01/01/70 00:00 |



