-
Most fundamentally, there is a gap between what can be done in a datacenter context today, and what can be done in the cloud, because some information is only known by the cloud provider. So one use c…
-
Recently, there is a problem revealed by a series of fuzz test cases, like #2555, #2560, #2562, #2563, ~~#2650~~, #2728, #2732, #2755. All of them share the same pattern in their Wasm modules.
``` …
-
Consider implementing the following time-series optimizations in Vinyl:
- If inserted keys are monotonically growing, split ranges on dump by the max key to avoid compaction.
- Force compaction of r…
-
Hi!
Recently I checked Profile-Guided Optimization (PGO) improvements on multiple projects. The results are [here](https://github.com/zamazan4ik/awesome-pgo/). E.g. PGO [helps](https://github.com/e…
-
Raising this issue to track performance optimization for bert training via the torch-mlir path.
Due to how torch-mlir currently lowers training workloads, the input for an iree benchmark command in…
-
Hi!
rsyslog right now does not support building with more advanced optimization techniques like [PGO](https://en.wikipedia.org/wiki/Profile-guided_optimization) and [BOLT](https://github.com/llvm/l…
-
Hi!
ArangoDB right now does not support building with more advanced optimization techniques like [PGO](https://en.wikipedia.org/wiki/Profile-guided_optimization) and [BOLT](https://github.com/llvm/…
-
Investigate the impact of using different span sizes on benchmark workloads.
-
**Is your feature request related to a problem? Please describe.**
libcudf aggregations show lower throughput when data cardinality is less than ~1000 distinct values. This is due to serializing atom…
-
Hi!
Recently I did many Profile-Guided Optimization (PGO) benchmarks on multiple projects - the results are available [here](https://github.com/zamazan4ik/awesome-pgo). I think since the project is…