Proceedings of the 31st annual international symposium on Computer architecture
Instruction packing: reducing power and delay of the dynamic scheduling logic
ISLPED '05 Proceedings of the 2005 international symposium on Low power electronics and design
Tornado warning: the perils of selective replay in multithreaded processors
Proceedings of the 19th annual international conference on Supercomputing
Power-Efficient Wakeup Tag Broadcast
ICCD '05 Proceedings of the 2005 International Conference on Computer Design
Instruction packing: Toward fast and energy-efficient instruction scheduling
ACM Transactions on Architecture and Code Optimization (TACO)
Exploiting Operand Availability for Efficient Simultaneous Multithreading
IEEE Transactions on Computers
Non-uniform instruction scheduling
Euro-Par'05 Proceedings of the 11th international Euro-Par conference on Parallel Processing
Instruction recirculation: eliminating counting logic in wakeup-free schedulers
Euro-Par'05 Proceedings of the 11th international Euro-Par conference on Parallel Processing
PACS'04 Proceedings of the 4th international conference on Power-Aware Computer Systems
| Hi-index | 0.00 |
For highest performance, a modern microprocessor must be able to determine if an instruction is ready in the same cycle in which it is to be selected for execution. This creates a cycle of logic involving wakeup and select. However, the time a static instruction spends waiting for wakeup shows little dynamic variance. This idea is used to build a machine where wakeup times are predicted, and instructions executed too early are replayed. This form of self-scheduling reduces the critical cycle by eliminating the wakeup logic at the expense of additional replays. However, replays and other pipeline effects affect the cost of misprediction. To solve this, an allowance is added to the predicted wakeup time to decrease the probability of a replay. This allowance may be associated with individual instructions or the global state, and is dynamically adjusted by a gradient-descent minimum-searching technique. When processor load is low, prediction may be more aggressive 驴 increasing the chance of replays, but increasing performance, so the aggressiveness of the predictor is dynamically adjusted using processor load as a feedback parameter.