slv-performance-boost
SolidUpdate the kernel (Ubuntu HWE 6.14+), set `amd_pstate=active` (AMD), and apply cpufreq governor=performance, EPP=performance, cpuidle disable, and `scaling_max_freq=cpuinfo_max_freq` so the CPU stays at maximum boost. AMD nodes on kernel < 6.14 require HWE install + reboot.
Install
Quality Score: 83/100
Skill Content
Details
- Author
- ValidatorsDAO
- Repository
- ValidatorsDAO/slv
- Created
- 1 years ago
- Last Updated
- 3 days ago
- Language
- TypeScript
- License
- Apache-2.0
Similar Skills
Semantically similar based on skill content — not just same category
slv-smt-disable
Disable SMT (Hyper-Threading) by adding `nosmt=force` to GRUB. Required for stable single-thread performance on Solana validators / RPCs. Reboot required.
slv-irq-tuning
NIC IRQ 1:1 pinning, NVMe queue limits, RPS/XPS, NIC ring buffer, sysctl (UDP/TCP buffers, swappiness, netdev_max_backlog), THP off, and ulimit nofile=1M for Solana / Shredstream nodes. Reboot is required when NVMe queue parameters change.
vllm-performance-tuning
vLLM performance-tuning operator reference — tuning workflow (baseline → bottleneck → knob → re-bench), fused-MoE kernel autotune (`benchmark_moe.py` generates `E=N,N=M,device_name=X.json` configs), DeepEP all-to-all + expert parallelism + EPLB, CUDA graph modes (FULL_AND_PIECEWISE default), torch.compile AOT + compile cache, scheduler knobs (`--max-num-batched-tokens`, `--max-num-seqs`, `--async-scheduling`), TP/EP/DP/PP decision tree, NCCL/DCGM on H100/H200/B200/GB200, PD disaggregation (Nixl/Mooncake/LMCache), known regressions + vendor quirks (v0.14→0.15.1 MiniMax, MI300X FP8<BF16, DeepGEMM M<128 TTFT).