-
![]()
12
-
Cherry: Checkpointed Early Resource Recycling in Out-of-order Microprocessors.pdf
- This paper presents CHeckpointed Early Resource RecYcling (Cherry), a hybrid mode of execution based on ROB and checkpoint- ing that decouples resource recycling and instruction retirement. Re- sources are recycled early, resulting in a more efcient utilization. Cherry relies on state checkpointing and rollback to service excep- tions for instructions whose resources have been recycled. Cherry leverages the ROB to (1) not require in-order execution as a fallback mechanism, (2) allow memory replay traps and branch mispredic- tions without rolling back to the Cherry checkpoint, and (3) quickly fall back to conventional out-of-order execution without rolling back to the checkpoint or ushing the pipeline. We present a Cherry implementation with early recycling at three different points of the execution engine: the load queue, the store queue, and the register le. We report average speedups of 1.06 and 1.26 in SPECint and SPECfp applications, respectively, relative to an aggressive con
-
-
![]()
12
-
Fast restore of checkpointed memory using working set….pdf
- Fast restore of checkpointed memory using working set…
-
-
![]()
12
-
cherry checkpointed early resource recycling in out-of-order.pdf
- cherry checkpointed early resource recycling in out-of-order
-
-
![]()
12
-
Cherry-MP Correctly Integrating Checkpointed Early Resource….pdf
- Cherry-MP Correctly Integrating Checkpointed Early Resource…
-
-
![]()
4
-
On the Latency, Energy and Area of Checkpointed, Superscalar….pdf
- On the Latency, Energy and Area of Checkpointed, Superscalar…
-
-
![]()
12
-
Checkpointed Early Load Retirement.pdf
- Long-latency loads are critical in today s processors due to the ever-increasing speed gap with memory. Not only do these loads block the execution of dependent instructions, they also prevent other instructions from moving through the in-order reorder buffer (ROB) and retire. As a result, the processor quickly lls up with uncommitted instructions, and computation ultimately stalls. To attack this problem, we propose checkpointed early load retirement, a mechanism that combines register check- pointing and back-end i.e., at retirement load-value pre- diction. When a long-latency load hits the ROB head un- resolved, the processor enters Clear mode by (1) taking a Checkpoint of the architectural registers, (2) supplying a Load-value prediction to consumers, and (3) EARly-retiring the long-latency load. This unclogs the ROB, thereby clearing the way for subsequent instructions to retire, and also allowing instructions dependent on the long-latency load to execute sooner. When the
-
1篇相似文档
-
![]()
14
-
Cherry Checkpointed Early Resource Recycling in Out-of-order.pdf
- Cherry Checkpointed Early Resource Recycling in Out-of-order.pdfCherry Checkpointed Early Resource Recycling in Out-of-order.pdfCherry Checkpointed Early Resource Recycling in Out-of-order.pdf
-
向豆丁求助：有没有checkpointed?
