Anti-fuse memory provides robust, secure NVM option
Bernd Stamme, Kilopass Technology Inc.
EETimes (7/5/2012 3:14 PM EDT)
Embedded non-volatile memory (NVM) intellectual property (IP) is a requirement for storing data that must be preserved when power to the chip is removed. NVM is found in almost every system on chip (SoC) design today, especially those targeting connected devices accessing content protected by digital rights management and sensitive financial or personal data. As these SoC designs migrate toward 28 nm and lower processes, engineering teams are re-examining the available commercial options. This reappraisal is occurring because of challenges presented by these smaller geometry processes. Suddenly, what was once an insignificant commodity is threatening to become a technology bottleneck.
NVMs can be classified in two groups: those that are programmed once and those that can be reprogrammed. In the former category, the most fundamental form of NVM is the masked ROM, which is programmed during the chip fabrication process. Next, comes the e-Fuse, which gets programmed during final test, followed by, anti-fuse technology, which can be programmed at wafer sort or final test or in the field at later point in time. In the latter category of reprogrammable memory are commercially available embedded Flash and floating gate alternatives. In the lab, new resistive RAM alternatives are being developed to compete in the reprogrammable arena.
To read the full article, click here
Related Semiconductor IP
- NVM OTP XBC TSMC N7 1.8V
- NVM OTP XBC TSMC N6 1.8V
- NVM OTP XBC TSMC N5A 1.2V Automotive Grade 1 with Functional Safety
- NVM OTP XBC TSMC N5 1.2V
- NVM OTP XBC TSMC N4P 1.2V
Related Articles
- Antifuse memory IP fuels low-power designs
- OTP with a ROM Conversion Option Provides Flexibility and Cost Savings for On-Chip Microcode Storage
- Argument for anti-fuse non-volatile memory in 28nm high-K metal gate
- NVM memory: A Critical Design Consideration for IoT Applications
Latest Articles
- Analog Foundation Models
- Modeling and Optimizing Performance Bottlenecks for Neuromorphic Accelerators
- RISC-V Based TinyML Accelerator for Depthwise Separable Convolutions in Edge AI
- Exclude Smart in Functional Coverage
- A 0.32 mm² 100 Mb/s 223 mW ASIC in 22FDX for Joint Jammer Mitigation, Channel Estimation, and SIMO Data Detection