Memory devices become increasingly susceptible to reflow-induced memory loss as process geometries shrink and reflow temperatures increase. Data I/O previously established that programming was safe with MLC devices. Does that change with automated grade TLC?
- As vendors move from MLC to TLC devices, will there still be an acceptable margin through reflow?
- Is there an impact on cell lifetime if devices are pre-programmed?
- What are the best practices for the newest generation of devices?
To better understand the thermal impact of oven reflow on automotive-grade 3D UFS, Data I/O partnered with a leading UFS memory vendor to study the impact.
The experiment showed :
-reflow slows down read performance but does not cause reliability failures
-More reflows cause read performance to slow down but do not cause loss of data.
Up to 3 reflows possible
Refreshing the device after reflow fixes read performance issues completely.
Oven reflow test was performed using automotive-grade 3D TLC UFS devices from multiple semiconductor vendors. Our study found that pre-programming is safe and full device performance is maintained with a post reflow data refresh. Data I/O has published a set of recommended best practices and guidelines for oven reflow
.{{cta('6cb678ef-4935-4a2b-bb53-2c366b3227e9')}}