MIT Department of Electrical Engineering & Computer Science

May 9, 1995
3:30 Reception, 4:00 Lecture
Room 34-101, 50 Vassar Street

Title: An Overview of Flash Memory Technology

Chi Chang
Advanced Micro Devices, Inc.
Sunnyvale, California

Flash memories have experienced rapid market growth in the last five years, starting from slightly less than $40 million annual sales in 1990 to a projected over $1 billion sales this year. Flash was first introduced primarily as an EPROM replacement for its in-system reprogrammable capability while enjoying the high density/low cost advantages over the byte-alterable EEPROM devices.

In 1994, we witnessed the emergence of another market for Flash, namely, the memory card and solid state drives. Although for the time being, the Flash based solid state mass storage is still in its infancy and is much more expensive relative to the rotating magnetic disk for obvious reasons: First, the chip density available now is mostly in the 2 Mb to 4 Mb range where the cost per bit is still very high. Secondly, the Flash Manufacturing is still climbing up on its learning curve and very few semiconductor houses have come close to mastering this expertise.

However, with the 16-Mbit and higher densities becoming available, this situation can change rather quickly. At a lower cost-per-bit level, Flash can compete with the low-density small-form-factor hard disk drive for applications such as the (battery powered) handheld computers and communication devices.

In this overview talk, common Flash memory cell types and associated program & erase operations will be discussed, along with cell degradation issues and endurance limitations. Some of the key process challenges and technology scaling trend will be addressed as well.