Research project for high performance, low cost computing. Uses stack-based instructions for a 4-way VLIW processor. Implemented in current technology, it would outrun high-end DSPs (TMS 320C6x, TigerSHARC), and let the full program run on one process...
Benefits: simple, easily pipelined, useful in self-clocked systems, very flexible and optimizable, good memory access, highly. Problems: awkward to program. [cowlark.com]
Class project to design and implement One Instruction Set Computer; using instruction: SUBLEQ A B C. Meaning: subtract value in M(A) from M(B) and store it in M(B), if result is not positive, go to instruction C. Descriptions, diagrams, code, tables.
Goal: dominate very low cost, low power, single core CPU market, and have the only ultra-high-end supercomputer on-a-chip system. Because the CPU core is so small, it is possible to put 1000s of cores and memory in one IC chip package.
Simple one instruction language; type of OISC; specifications from Clive Gifford eigenratios self-interpreter page. Each subleq instruction has 3 operands which are memory addresses. Oleg Mazonka.
By Henk Corporaal; John Wiley, 1998, ISBN 047197157X. Introduces Transport Triggered Architectures, TTAs. In standard architectures, programmed operations trigger internal data transports. TTAs work by programming data transports themselves. This remo...
Automatic Synthesis of Application Specific Processors. Goal: design, build high-performance processors via a new class of transport-triggered architectures, TTAs, programmed by specifying data-transports not operations; well suited for application sp...