FRX INIT READY·command frx init·evidence NCU + traces·model auto-detected·loop init -> profile -> brief

Guided first run

Find GPU mistakes.
Fix them with proof.

Run frx init to check your machine, detect the GPU model, and get the right command for CUDA kernels or PyTorch training. Add --explain to write the LLM brief in the same run.

Opportunity-ranked kernels, framework tax signals, and architecture-aware thresholds for H100, A100, L4, RTX4090, and RTX5060.

first run

init

arch profile

auto GPU

LLM brief

1 file

Start with frx init View on GitHub

Works with

PyTorch·JAX·TensorFlow·NVIDIA

frx profile --ncu report.csv --explain

BRIEF READY

GPU

rtx5060

PRIMARY

uncoalesced

BRIEF

written

Measured evidencethreshold: high > 4

loadL1L2SMissue

Recommended next actions

Top fix

Coalesce loads

Reduce working set

Check

Tune occupancy

Ready

Validate rerun

CUDA streams — 12.4 msHotspot

ENG

gemm_fwd

attn_fwd

ln_fwd

gemm_bwd

attn_bwd

MEM

H2D

ckpt

D2H

adam

ELW

relu

add

gelu_bwd

bias_bwd

04ms8ms12ms

FwdBwdOpt

SM occupancy × timePeak 98%

SM0

SM1

SM2

SM3

SM4

SM5

SM6

SM7

IdleForwardBackwardOpt

idleactivepeak

frx_llm_prompt.txt - paste into your LLM with measured evidence

Trusted by teams shipping the world's most compute-intensive workloads

AI21 labs

Perplexity

character.ai

Midjourney

Cohere

Anyscale

What we detect

Complex GPU mistakes. Named, ranked, and backed by evidence.

Fournex turns NCU counters, PTX structure, and runtime traces into the mistake that matters: the threshold crossed, the kernel worth fixing first, and the validation command to prove the change.

9.7 sectors/request, high > 4

Uncoalesced global loads

Catch memory access patterns that spray global-load transactions. Get coalescing actions and a re-profile command to confirm the fix.

View sample trace→

L1 hit rate below 40%

L1 cache thrashing

Separate L1-only locality problems from broader memory pressure. Prioritize shared-memory tiling or working-set reduction based on evidence.

View sample trace→

L2 hit rate below 50%

L2 cache thrashing

Identify L2-only pressure without confusing it with L1 misses. Fournex points engineers toward reducing the active working set first.

View sample trace→

Low occupancy, register cap

Occupancy limited by registers

Use measured occupancy and launch-resource data to spot register pressure that blocks residency, then reduce live ranges or split kernels.

View sample trace→

Too few eligible warps

Low scheduler utilization

Surface kernels that keep schedulers waiting even when occupancy looks acceptable. Separate ILP, block-size, and warp-eligibility issues.

View sample trace→

GPU idle, CPU or copy bound

Host and input stalls

Connect runtime traces to practical fixes for dataloaders, pinned memory, transfer pressure, and hidden synchronization points.

View sample trace→

Runtime share x MFU gap

Kernel opportunity ranking

Rank kernels by runtime share, roofline region, MFU gap, and severity so the highest-impact work rises above small noisy offenders.

View sample trace→

Launch fragmentation, idle gaps

Framework abstraction tax

Surface framework overhead that is not explained by input, copy, or sync stalls, with inferred graph-capture and fusion opportunities clearly marked.

View sample trace→

Evidence to action

From profiler output to ranked fix list. Under a minute.

Every finding comes with the exact signal that triggered it, the threshold it crossed, confidence level, actions to take, validation steps, and caveats before you change code.

Measured metrics with [!!], [ !], [ok], and threshold hints
NCU, PTX, and runtime trace signals routed to one report
Recommendations include why, actions, validation, and risks
Next steps include the exact re-run command

Read profile docs Product roadmap

frx-profile / attention_kernel

Analyzed

Primary

uncoalesced

L1 hit rate

31%

L2 hit rate

71%

#FIX / TITLEEVIDENCERISK

Coalesce global loads in attention_kernel

effort · Low · confidence · High

9.7 -> 2.1Low

Reduce active working set before touching tiling

effort · Low · confidence · High

L2-onlyLow

Lower register pressure to recover occupancy

effort · Medium · confidence · Medium

+18 ptsMedium

Remove host sync points in the training loop

effort · Low · confidence · High

low riskLow

Profile report with why, actions, validation, and caveats

Product evolution

Profiling and validation today. Guarded automation next.

We don't ask engineers to trust blind automation. The product starts with evidence, then moves through ranked recommendations, before/after validation, benchmark proof, and guarded automation.

01 · Phase 1Shipping now

Guided profiling

First-run environment checks
Low-overhead trace collection
Deterministic bottleneck classifier

02 · Phase 2Shipping now

Ranked recommendations

Map bottlenecks to concrete fixes
Score by impact, effort, and risk
Explainable, repeatable output

03 · Phase 3Shipping now

Compare and bench validation

Before/after evidence diffs
Warmup-discarded kernel bench
One-command LLM briefs

04 · Phase 4On the roadmap

Policy-driven autopilot

Auto-apply within your guardrails
Continuous adaptation
Learned optimization policies

Capabilities

Built for engineers. Readable in the terminal.

Start with a concrete diagnosis, not a wall of counters. Keep the full evidence trail, from kernel opportunity ranking to benchmark proof, when you need to defend the change in review.

Guided setup

frx init checks Python, PyTorch, CUDA, Nsight Compute, and the active GPU model.

Evidence-backed fixes

Each recommendation names the metric, threshold, rule, and validation step.

Memory diagnostics

Separate L1 thrashing, L2 thrashing, bandwidth pressure, and uncoalesced loads.

Occupancy diagnosis

Tie low occupancy to registers, shared memory, block size, or scheduler pressure.

Before/after validation

Compare source, PTX, NCU, and bench evidence before trusting a change.

Kernel opportunity scoring

Rank kernels by runtime share, roofline region, MFU gap, and severity.

Framework tax signals

Spot launch fragmentation and inferred graph-capture or fusion opportunities.

LLM-ready briefs

Use --explain on profile or collect to write the measured prompt in the same workflow.

Simple workflow

Start with init. Leave with a brief.

Let Fournex check your environment, recognize the GPU model, and route you to the right workflow. For NCU reports and PyTorch training runs, add --explain to generate the LLM-ready brief without running a second analysis command.

Installpip install fournex

First runfrx init

Patch trainingfrx init --patch train.py

Training brieffrx collect --explain -- python train.py

NCU commandfrx ncu-command -- ./my_binary

Live profilefrx profile --explain -- python train.py

Existing NCUfrx profile --ncu ncu_report.csv --explain

Static PTXfrx profile --ptx kernel.ptx

Compare fixfrx compare baseline.cu optimized.cu

Bench kernelsfrx bench bad.cu good.cu

frx - profile.sh

$ frx profile --ncu report.csv --explain

GPU: NVIDIA RTX 5060 -> rtx5060

VERDICT: uncoalesced_access, confidence=high

MEASURED: global load sectors/request 9.7[!!] high > 4

Next steps

Try rec_ncu_improve_coalescing, then re-run the same profile command.

[!!] L1 hit rate 31% low < 40%

[ok] L2 hit rate 71% healthy

[ !] occupancy limited by registers secondary

LLM briefwritten

frx_llm_prompt.txt - paste into your LLM

Engineer loop

Short path from suspicion to fix. No profiler archaeology.

Fournex keeps the path small: capture the evidence, name the mistake, apply the recommendation, and rerun the same command to verify the change.

init

guided first run

auto

GPU model detection

3 files

summary, prompt, evidence

0 guesswork

threshold evidence

Engineers do not need to interpret every counter by hand. The report explains why a metric is bad, which rule fired, what to change, and how to confirm the result.

View the docs

Moat

A workload-performance dataset that compounds every week.

We're not another rules engine. Every analyzed workload expands the mapping from trace patterns to validated fixes — turning usage into defensibility.

Proprietary trace → fix → outcome dataset

Validated optimization deltas across hardware

Policies that improve as more teams onboard

Trust layer: every change auditable and reversible

Compounding flywheel

More workloads analyzed

Richer optimization traces

Better policy learning

Stronger recommendations

More workloads onboard

Better data → better policies → better outcomes → more workloads. That loop is the product.

Find GPU mistakes.Fix them with proof.

Complex GPU mistakes. Named, ranked, and backed by evidence.

Uncoalesced global loads

L1 cache thrashing

L2 cache thrashing

Occupancy limited by registers

Low scheduler utilization

Host and input stalls

Kernel opportunity ranking

Framework abstraction tax

From profiler output to ranked fix list. Under a minute.

Profiling and validation today. Guarded automation next.

Guided profiling

Ranked recommendations

Compare and bench validation

Policy-driven autopilot

Built for engineers. Readable in the terminal.

Guided setup

Evidence-backed fixes

Memory diagnostics

Occupancy diagnosis

Before/after validation

Kernel opportunity scoring

Framework tax signals

LLM-ready briefs

Start with init. Leave with a brief.

Short path from suspicion to fix. No profiler archaeology.

A workload-performance dataset that compounds every week.

Find GPU mistakes.
Fix them with proof.