Do this only if the format matches
Confirm the round with recruiting first. This is an original practice exercise, not a claim that a specific company uses this exact prompt.
Protocol
- Clone the official public challenge and run its submission tests unchanged.
- Record baseline cycles and map execution units, dependencies, and scratchpad limits.
- Use the trace to choose one bottleneck hypothesis.
- Predict the effect before changing code.
- Measure, verify, and either keep or revert the change.
- Review agent changes for weakened tests or machine constraints.
Deliverables
- A verified cycle count and unchanged official tests.
- An experiment ledger with predictions and measurements.
- One failed optimization and what it taught you.
- A short explanation of simulator-to-hardware limits.
Graduation gates
The official verification command passes.
Every kept change has a measured effect.
Scratchpad and dependency constraints remain valid.
The candidate can distinguish instruction count, memory traffic, and critical-path depth.
Starter files
External practice artifact
Anthropic original performance take-home ↗
Official public challenge. Follow its current verification instructions.