The Soviet microelectronics of the 1980s presented significant challenges for implementing non-volatile memory. The K155RE3 PROM, a 256-bit chip, required burning internal fuses with voltages ranging from 12.5 V to 14 V. This process was unreliable, with a programming coefficient of 0.3, and risked destroying the chip or previously stored data.
Reliability issues extended to physical connections. Soviet-era connectors were prone to oxidation and poor contact, which could lead to data corruption. Furthermore, the K573RF1 EPROM, introduced around 1982, offered 1 KB of memory and was erasable via ultraviolet light. While more functional, it was expensive and required a 15–30 minute erasure process.
The KR1601RR1 chip, used in the Elektronika MK-52 calculator, featured 4 kilobits of memory but required complex power management. It needed three supply voltages, including -30 V for programming, which posed a risk to other components. Additionally, the manufacturer guaranteed data retention for only 3,000 hours to 5 years.
Implementing a 64 KB storage device using these components would require 128 chips, resulting in a bulky, power-hungry, and fragile assembly. The need for external voltage multipliers and the sensitivity of MNOS structures to power-up sequences made these chips impractical for modern-style flash storage applications.
