Picking statistically valid and early simulation points

Abstract

Abstract

Analysis before Starting an Access: A New Power-Efficient Instruction Fetch Mechanism

“…We use a number of integer benchmarks from the SPEC2000 suites benchmarks. All benchmarks were compiled with highest optimization level by the Alpha compiler [26], and were fast-forwarded pass the first 500 millions instructions to bypass initialization and startup code before measured simulation begins. Then, full-detail simulation is performed for next one billion instructions.…”

Budgeted region sampling (BeeRS): do not separate sampling from warm-up, and then spend wisely your simulation budget

“…While the recent surge of research articles on sampling started with rather large sample sizes (100M in the first SimPoint article [12]), it has later shifted to very small intervals (1,000 in SMARTS [14]), and it is now converging to intermediate sizes (1M and 10M in SimPoint [10,6]), and even to varying sizes in EXPERT [7] and SimPoint VLI [5] (ranging from 52,000 to 6.1M in EXPERT, and 100M to 500M in SimPoint VLI). With 100M samples, warm-up is not an issue, at least with current cache sizes.…”

“…In sampling research works, warm-up has been treated in two different ways: either using functional [14,7] or checkpoint-based warm-up [13], or assuming warm-up is perfect based on the principle of separating warm-up and sampling issues [12,10,6].…”

Pinpointing Representative Portions of Large Intel® Itanium® Programs with Dynamic Instrumentation

“…We want to provide a quick turnaround when looking at a new application or see the effect of a new compiler on an existing application. To speed up the process, we split the validation process into three steps: [16] showing that a single set of simulation points can accurately predict behavior across multiple configurations of the same architecture.…”