Il suit les mêmes flags que Go, sans dépendance, et mesure les allocations heap (B/op, allocs/op) […]
[Original post on mastodon.social]
Il suit les mêmes flags que Go, sans dépendance, et mesure les allocations heap (B/op, allocs/op) […]
[Original post on mastodon.social]
Il suit les mêmes flags que Go, sans dépendance, et mesure les allocations heap (B/op, allocs/op). ⬇️
news.humancoders.com/t/rust…
Il suit les mêmes flags que Go, sans dépendance, et mesure les allocations heap (B/op, allocs/op). ⬇️
news.humancoders.com/t/rust…
https://www.bwplotka.dev/2026/rust-microbenchmarks-benchstat/
https://www.bwplotka.dev/2026/rust-microbenchmarks-benchstat/
Now models have the entire moat and in fact everyone is trying to get OUT of the way with things like code mode
Now models have the entire moat and in fact everyone is trying to get OUT of the way with things like code mode
https://www.bwplotka.dev/2026/rust-microbenchmarks-benchstat/
[comments] [28 points]
https://www.bwplotka.dev/2026/rust-microbenchmarks-benchstat/
[comments] [28 points]
github.com/torvalds/lin...
github.com/torvalds/lin...
#CUDA #Performance #NVIDIA #Blackwell #Memory
hgpu.org?p=31407
#CUDA #Performance #NVIDIA #Blackwell #Memory
hgpu.org?p=31407
Latest local microbenchmarks:
• yield_switch: 2.33× faster than Go
• idle_spawn: 1.74× faster
• parked task: 765 B vs Go 2,741 B
Still experimental. Repo ↓
github.com/lechatthecat...
#GarbageCollector
Latest local microbenchmarks:
• yield_switch: 2.33× faster than Go
• idle_spawn: 1.74× faster
• parked task: 765 B vs Go 2,741 B
Still experimental. Repo ↓
github.com/lechatthecat...
#GarbageCollector
Also: @gmicloud.bsky.social raised a $223M Series B. And FormEnergy.com is hiring a Manager, Site Reliability Engineering in Berkeley.
theconsensus.dev
Also: @gmicloud.bsky.social raised a $223M Series B. And FormEnergy.com is hiring a Manager, Site Reliability Engineering in Berkeley.
theconsensus.dev
inside.java/2026/10/05/c...
#java #FFM
inside.java/2026/10/05/c...
#java #FFM
z.ai/blog/glm-bui...
z.ai/blog/glm-bui...
LIKWID is a command-line performance monitoring and benchmarking suite for CPU topology, counters, affinity, energy and microbenchmarks. The post LIKWID – performance monitoring and benchmarking su...
📰 Source: LinuxLinks
🔗 Link […]
LIKWID is a command-line performance monitoring and benchmarking suite for CPU topology, counters, affinity, energy and microbenchmarks. The post LIKWID – performance monitoring and benchmarking su...
📰 Source: LinuxLinks
🔗 Link […]
github.com/chrisarg/ben...
of alternative computational for the popcounts under different optimization.
(-O0, -O3, -O3 -flto). Is it the algorithm, the compiler or noise the driver? #hpc
github.com/chrisarg/ben...
of alternative computational for the popcounts under different optimization.
(-O0, -O3, -O3 -flto). Is it the algorithm, the compiler or noise the driver? #hpc
https://github.com/chrisarg/bench_popcount
of alternative computational for the popcounts under different optimization.
(-O0, -O3, -O3 -flto). Is it the algorithm, the compiler or noise the driver? #hpc
https://github.com/chrisarg/bench_popcount
of alternative computational for the popcounts under different optimization.
(-O0, -O3, -O3 -flto). Is it the algorithm, the compiler or noise the driver? #hpc
Microbenchmarks are experiments: mrale.ph/blog/2024/11... by @mrale.ph
Microbenchmarks are experiments: mrale.ph/blog/2024/11... by @mrale.ph
blog.s-schoener.com/2025-02-10-m...
blog.s-schoener.com/2025-02-10-m...