|
WaveScalar |
Summary
The next two decades will bring us computation resources that seem unimaginable today. Advancements in lithographic technology will make devices of ten billion transistors and beyond. With this opportunity, however, come significant archi tectural challenges that will require a revolution as fundamental as the migration to VLSI. Relative to computation, communication will be orders of magnitude slower. An enormous number of transistors will be available, but there reliability will progressively decrease. Finally, one device that is not changing very quickly is the human mind. This is leading to the ever-expanding productivity-gap between the complexity of modern architectures and the rate at which they can be designed and verified. We propose the Computation Cache, a unique way to construct systems that forges together the concepts of an instruction cache and a microprocessor into a single unified idea.
Students
Steven Swanson
Andrew Schwerin
Andrew Petersen
Ken Michelson
Publications
Steve Swanson, Ken Michelson, Andrew Schwerin and Mark Oskin.
WaveScalar
In the 36th Annual International Symposium on Microarchitecture (MICRO-36), December 2003
Steve Swanson, Ken Michelson, Andrew Schwerin and Mark Oskin.
Dataflow: The Road Less Complex
In the Workshop on Complexity-effective Design (WCED) held in conjunction
with the 30th Annual International Symposium on Computer Architecture (ISCA), June 2003
Andrew Schwerin, Steve Swanson and Mark Oskin
Measuring the Complexity-effectiveness of Future-generation Silicon Architectures using FPGAs: A Status Report
In the Workshop on Complexity-effective Design (WCED) held in conjunction
with the 30th Annual International Symposium on Computer Architecture (ISCA), June 2003
Steven Swanson and Mark Oskin.
Towards a Universal Building Block of Molecular and Silicon Computation
Workhop on Non-Silicon Computing, held in Conjunction with the International
Symposium on High Performance Computer Architecture, Feb. 2002
Related Efforts
- Intelligent Memory Systems
- Substrates
- Data-flow & Hybrids