Investigating the impact of code generation on performance characteristics of integer programs
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
As the complexity of interactions among microprocessors, platforms, and system software increases, compilers play a greater role in extracting performance for applications. Different compiler technologies have contributed significantly in improving the designs of various processor architectures and in delivering end-user performance. Workload analysis of benchmark suites has traditionally focused on understanding the behaviors of the suites under different system configurations and studying the sensitivity of the suites to different system parameters. In this paper, we study the effect of compiler technology and implementation on workload behaviors. This study aims at understanding how performance and its leading metrics behave differently with different compiler implementations targeting the same architecture and programs. We use the quad-core AMD Opteron processors and the SPEC CPU 2006 Integer benchmark to evaluate how various micro-architecture performance metrics are sensitive to three top-performing compilers. Even though all three compilers produce overall good performance, our analysis shows that different compilers may vary widely on individual program performances. Binaries from different compilers vary both in terms of architecture metrics, such as instruction counts and instruction mix distribution, and micro-architecture metrics, such as cache miss rates and branch mis-predictions. For programs with large deviations in some of the metrics, we attribute the differences to the choices or strengths in a particular compiler implementation, for example on function inlining, 32-bit versus 64-bit compilation, software prefetching, vectorization, register allocation, etc.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it