tilegym-improve-cutile-kernel-perf
Iteratively optimize cuTile kernel performance through systematic profiling, bottleneck analysis, IR comparison, and targeted tuning. Covers tile sizes, occupancy, autotune configs, TMA, latency hints, persistent scheduling, num_ctas, flush_to_zero, and IR-level debugging. Use when asked to "optimize cutile kernel", "improve kernel perf", "tune cutile performance", "make kernel faster", or iteratively benchmark and refine a cuTile GPU kernel in the TileGym project.
v1.0.0New
Unsigned, install at your own risk
UnverifiedThis skill has no cryptographic signature attached. We can't verify the contents match what the publisher intended.
Install this skill
Run this command in your terminal. No account required — it auto-detects your AI tool and installs the skill file.
npx @skills-hub-ai/cli install skills-tilegym-improve-cutile-kernel-perfOr download directly:
Browse all CLI commands →Setup by platform
Install
One-click setup for your editorRun in your project root
npx @skills-hub-ai/cli install skills-tilegym-improve-cutile-kernel-perf --target claude-codeInstructions
Security
Loading security scan...
Reviews (0)
Related skills
Browse all →web-performance-auditorAI-powered web performance auditor for frontend and web developers. Use it to audit a website or web app, diagnose poor Core Web Vitals (LCP, CLS, INP) and low Lighthouse scores, review JavaScript bundle size, code splitting, lazy loading, image and font loading, hydration, React rendering, Next.js/Nuxt patterns, network waterfalls, third-party scripts, and caching. Produces a prioritized, senior-engineer performance audit with a score and concrete fixes.3 installsamc-setup-calibration-stackLaunch AutoMagicCalib microservice and web UI from NGC release images via Docker Compose. Use when user says 'deploy auto calibration', 'launch auto calibration', 'launch AMC', 'start MS+UI', or 'set up auto-magic-calib'. Requires NGC API key.0 installsphysical-ai-defect-image-generationUse when the user wants to orchestrate defect image generation with NVIDIA Cosmos AnomalyGen (Cosmos-Predict2-derived) on OSMO for PCBA, metal surface, and glass inspection. The Day 0 path handles cold-start with USD-to-ROI, image-edit augmentation, and AnomalyGen to create initial PCBA datasets. The Day 1 path performs inference and labeling on real images. This skill helps with first-time asset setup, creation of finetuning checkpoints, and configuring deployment.
Trigger keywords: defect image generation, dig workflow, dig pipeline, defect image detection workflow, aoi pipeline, aoi anomalygen, usd2roi anomalygen, day 0 pcba, day 1 pcba, day 1 real-photo alignment, day 1 manual roi, metal surface anomaly, glass defect, anomalygen finetune, setup_pcb, setup_metal, setup_glass, setup_pretrained, dig setup, dig datasets, dig pretrained checkpoint, dig image-edit endpoint, cosmos defect generation, cosmos-predict2 defect, cosmos-anomalygen, cosmos predict2 finetune.0 installsjetson-video-benchmarkUse when measuring Jetson Video Codec SDK or PyNvVideoCodec encode/decode throughput, comparing presets or surfaces, testing codec-worker capacity with authenticated samples and user media, or producing a documented clock-scaled or clock-and-resolution-scaled planning estimate when representative content is unavailable. Also use for Jetson video requests asking only for PSNR or SSIM results, to apply this performance skill's scope-only response.0 installsi4h-catheter-navigation-digital-twinBuild a patient vasculature digital twin from CT (preprocess + segment). Use when asked to preprocess CT, segment vessels, extract centerline, or prepare ct_cache for viewport/DRR.0 installsnemo-mbridge-perf-activation-recomputeValidate and use selective and full activation recompute in Megatron Bridge to reduce GPU memory usage at the cost of extra compute. Use for activation memory OOMs or regressions involving recompute_granularity, recompute_num_layers, recompute_modules, recompute_method, selective recompute, full recompute, or activation checkpointing.0 installs
More from skills
View source →doca-dpa-hl-tracerUse this skill when the user runs doca_dpa_hl_tracer to capture/decode DPA-side traces at the programming-events layer (kernel entry/exit, sync points, comm primitive calls, RDMA WR submission, completion drain) — picking TRACE vs CRIT, tuning the JSON config (file-size limits + file_size_limit_policy, thread priorities/cores), decoding against the matching DPA-side ELF, or diagnosing empty/noisy captures. Trigger even when the user does not explicitly mention "DOCA DPA tracer" or "high-level tracer" — typical implicit phrasings include "DPA kernel returns wrong result but host completions look clean", "kernel-entry to first-comm latency is huge", "RDMA WR to drain gap on the DPA", "trace file truncated mid-run", "TRACE doubled my DPA latency", or "tracer wrote a file but parser shows zero events". Refuse and route elsewhere for writing DPA kernels, DPA-Comms/DPA-Verbs programming, raw per-cycle DPA profiling, host-side doca-dpa debugging, or production DPA telemetry — those belong to 0 installsdoca-erasure-codingUse this skill when the user is doing hands-on DOCA Erasure Coding programming on a BlueField DPU, ConnectX NIC, or host — bringing up a doca_ec context, picking among the create / recover / update tasks, choosing matrix type / N / K / block size, querying doca_ec_cap_* before sizing, setting doca_mmap src/dst permissions, or debugging DOCA_ERROR_* returns from doca_ec_task_*. Trigger even when the user does not name "DOCA Erasure Coding" or "Reed-Solomon" — typical implicit phrasings include "one data block changed, how do I refresh parity without re-encoding", "a disk failed and 2 parity blocks are gone, can I rebuild", "RAID-6 resilience across 12 disks", "my doca_ec_task_create returns NOT_PERMITTED", or "is this N+K layout still recoverable". Refuse and route elsewhere for non-Reed-Solomon codes (fountain / LDPC / raptor), pure-replication designs, network FEC, or other DOCA accelerator libraries (SHA / Compress / AES-GCM / DMA) — those belong to other skills.0 installsdoca-flowBuild and debug DOCA Flow applications on supported NVIDIA NICs/DPUs: define match/action pipes, initialize ports and representors, choose forwarding targets, validate pipes before hardware programming, read counters, match the Flow version to the installed DOCA release, and diagnose Flow API errors. Trigger on DOCA packet steering, classifier, representor, rule-matching, hairpin, or 5-tuple-to-queue questions even when "DOCA Flow" is not named. Route plain DPDK `rte_flow`, kernel TC, OVS, BFB bring-up, and DPU OS installation elsewhere. DPU OS installation is destructive and always requires explicit confirmation.0 installsdoca-flow-dpa-providerUse this skill when the user is doing hands-on DOCA Flow DPA Provider work — exporting a `doca-flow` pipe or external resource (index-selector/memory) into BlueField DPA address space so a DPACC-built kernel can read counters, mutate hash-pipe entries, and update/read memory or index-selector resources inline with Flow. Covers per-port `doca_flow_dpa_ctx`, three queue types (general/resources-write/resources-read), the order-sensitive export handshake (`_export_prepare` → add entries → `_export` → `_get_device_addr`), and DPA-side device API. Trigger even when the user does not say "DOCA Flow DPA Provider" — implicit phrasings include "DPA kernel never sees entries in the exported pipe", "BAD_STATE from `_pipe_export`", "how do I disable a hash entry from a DPA kernel", "DPA memory read returns no value", or "DPA-side post keeps returning AGAIN". Refuse and route elsewhere for `doca-flow` pipe construction, generic host-side DPA (`doca-dpa`), or DPA-side kernel-writing — those belong t0 installsdoca-flow-perfUse this skill when the user is measuring the host or DPU-CPU control-plane rate of a DOCA Flow pipeline with doca_flow_perf — picking a JSON policy from configs/, choosing the DPDK or DOCA backend, running the single-iteration smoke then the iterative eval loop, interpreting per-iteration CPU cycles and num_pushed / num_failed, or capturing the four-tuple (DOCA version, BlueField/firmware, JSON policy, worker/queue/burst config) that makes a Kops/sec number defensible. Trigger even when the user does not explicitly mention "doca-flow-perf" — typical implicit phrasings include "how many rules per second can my BlueField insert", "5-tuple hairpin rule rate", "Kops/sec for steering", "flow-perf number does not match release notes", "DPDK vs DOCA benchmark", or "rule-install variance too high". Refuse and route elsewhere for optimizing a live Flow app (doca-flow-tune), the DPA-offloaded path (doca-flow-dpa-perf), dataplane throughput or latency, or library-internal pipe semantics — those 0 installsdoca-dmsOperate NVIDIA DOCA Management Service (`dmsd` + `dmspe`) on a BlueField, Arm/x86 host, or Kubernetes pod: choose deployment and authentication, configure `-allowed_users` and `dmsgroup`, use gNMI Get/Set/Subscribe, run supported gNOI workflows, and debug frontend/backend failures. Trigger even without "DMS" for "manage a remote BlueField over gRPC", "gNOI reboot from orchestrator", or fleet-management requests. SAFETY: reboot, OS install, factory-reset, and managed-file deletion are destructive and require target-bound explicit confirmation; never invoke them speculatively. Route installation and library/API build questions elsewhere, and route turnkey aggregation to the externally-productized DOCA Telemetry Service.0 installs
More Build skills
Browse category →ui-design-systemUI design system toolkit for Senior UI Designer including design token generation, component documentation, responsive design calculations, and developer handoff tools. Use for creating design systems, maintaining visual consistency, and facilitating design-dev collaboration.136 installssenior-frontendFrontend development skill for React, Next.js, TypeScript, and Tailwind CSS applications. Use when building React components, optimizing Next.js performance, analyzing bundle sizes, scaffolding frontend projects, implementing accessibility, or reviewing frontend code quality.80 installsfrontend-designGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults.58 installsself-improving-agentCurate Claude Code's auto-memory into durable project knowledge. Analyze MEMORY.md for patterns, promote proven learnings to CLAUDE.md and .claude/rules/, extract recurring solutions into reusable skills. Use when: (1) reviewing what Claude has learned about your project, (2) graduating a pattern from notes to enforced rules, (3) turning a debugging solution into a skill, (4) checking memory health and capacity.57 installsusing-superpowersUse when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions40 installssenior-backendDesigns and implements backend systems including REST APIs, microservices, database architectures, authentication flows, and security hardening. Use when the user asks to "design REST APIs", "optimize database queries", "implement authentication", "build microservices", "review backend code", "set up GraphQL", "handle database migrations", or "load test APIs". Covers Node.js/Express/Fastify development, PostgreSQL optimization, API security, and backend architecture patterns.37 installs
Frequently asked questions about tilegym-improve-cutile-kernel-perf
What does the tilegym-improve-cutile-kernel-perf skill do?
Iteratively optimize cuTile kernel performance through systematic profiling, bottleneck analysis, IR comparison, and targeted tuning. Covers tile sizes, occupancy, autotune configs, TMA, latency hints, persistent scheduling, num_ctas, flush_to_zero, and IR-level debugging. Use when asked to "optimize cutile kernel", "improve kernel perf", "tune cutile performance", "make kernel faster", or iteratively benchmark and refine a cuTile GPU kernel in the TileGym project. It's a reusable SKILL.md instruction set that loads into your AI coding assistant on demand, no prompt engineering, no copy-pasting every session.
How do I install the tilegym-improve-cutile-kernel-perf skill?
Run `npx @skills-hub-ai/cli install skills-tilegym-improve-cutile-kernel-perf` from your terminal. The CLI writes the SKILL.md to the correct location for your AI tool (e.g. ~/.claude/skills/skills-tilegym-improve-cutile-kernel-perf/ for Claude Code or ~/.cursor/skills/ for Cursor with --target cursor) and adds it to your project's .skills.json lockfile.
Which AI tools does tilegym-improve-cutile-kernel-perf work with?
tilegym-improve-cutile-kernel-perf runs in Claude Code. It follows the open Agent Skills standard (SKILL.md), so the same skill works in every supported tool without modification.
Is the tilegym-improve-cutile-kernel-perf skill free?
Yes. Every skill on skills-hub.ai is free and open-source. There are no premium tiers, paywalls, or usage limits. You only pay for whatever AI assistant you're already using.
How do I use tilegym-improve-cutile-kernel-perf after installing it?
In Claude Code, type `/skills-tilegym-improve-cutile-kernel-perf` (or whatever slash command the skill registers) and the AI follows the skill's instructions immediately. You can also reference it by name in natural language, your AI loads the skill into context when relevant.
Can I share the tilegym-improve-cutile-kernel-perf skill with my team?
Yes. Commit your project's .skills.json lockfile and teammates run `npx @skills-hub-ai/cli install` (no args) to install every skill at the exact version you pinned. Organization-scoped installs work via skills-hub.ai organizations.