English
Using sub-RISC processors in next generation customizable multi-core designs: Part 1
By Andrew Mihal, Scott Weber, and Kurt Keutzer, University of California, Berkeley
Mar 28 2007 (23:15 PM), Embedded.com
The now-common phrase "the processor is the NAND gate of the future" begs the questions: "What kind of processor?" and "How to program them?" When this is discussed, the focus is usually placed on RISC-based processors augmented with instruction extensions as the natural building block. The presumption is that programming will be done in the C language.
Challenging this viewpoint, our opinion is than even a RISC processor is too coarse-grained for typical embedded applications, and that C is insufficient for programming a multiprocessor architecture. As an alternative, we explore the design and deployment of tiny sub-RISC processors as the "NAND gate of the future."
With Tiny Instruction-set Processors and Interconnect (TIPI) processing elements, we can achieve better performance than RISC elements with fewer resource requirements. Also, we can deploy concurrent applications with a programming methodology more productive than C.
Mar 28 2007 (23:15 PM), Embedded.com
The now-common phrase "the processor is the NAND gate of the future" begs the questions: "What kind of processor?" and "How to program them?" When this is discussed, the focus is usually placed on RISC-based processors augmented with instruction extensions as the natural building block. The presumption is that programming will be done in the C language.
Challenging this viewpoint, our opinion is than even a RISC processor is too coarse-grained for typical embedded applications, and that C is insufficient for programming a multiprocessor architecture. As an alternative, we explore the design and deployment of tiny sub-RISC processors as the "NAND gate of the future."
With Tiny Instruction-set Processors and Interconnect (TIPI) processing elements, we can achieve better performance than RISC elements with fewer resource requirements. Also, we can deploy concurrent applications with a programming methodology more productive than C.
|
E-mail This Article | 
|
|
Printer-Friendly Page |
Related Articles
- Evaluating the performance of multi-core processors - Part 1
- Evaluating the performance of multi-core processors - Part 2
- Multi-Core Processors: Driving the Evolution of Automotive Electronics Architectures
- Techniques for debugging an asymmetric multi-core application: Part 2
- Achieving multicore performance in a single core SoC using a multi-threaded virtual multiprocessor: Part 1
New Articles
- From a Lossless (~1.5:1) Compression Algorithm for Llama2 7B Weights to Variable Precision, Variable Range, Compressed Numeric Data Types for CNNs and LLMs
- Embracing a More Secure Era with TLS 1.3
- New PCIe Gen6 CXL3.0 retimer: a small chip for big next-gen AI
- Maximizing ESD protection for automotive Ethernet applications