TorchedUp
LearnBetaProblemsSystem DesignSoonPremium
TorchedUp
LearnBetaProblemsSystem DesignSoonPremium
Learn/Vectorized Production Math
📐

Vectorized Production Math

Real production sizing scripts don't loop — they vectorize. This track starts from the four broadcasting primitives every numpy expression decomposes into (bias-add, batched dot, outer product, axis reduce) and then layers them up into the kind of multi-config memory and FLOPs sweeps that capacity-planning teams actually ship. By the end you should be able to take any scalar napkin formula and turn it into a one-line numpy expression that runs across thousands of configs at once.

12 problems · suggested order

  1. ○1#220NumPy Broadcasting: Bias Addeasy
  2. ○2#221NumPy Inner Product (Dot, Batched)easy
  3. ○3#222NumPy Outer Product (Broadcast Grid)easy
  4. ○4#223NumPy Axis Reduceeasy
  5. ○5#224NumPy Einsum Intromedium
  6. ○6#225Weight Memory Sweepeasy
  7. ○7#226KV Cache Size Sweepmedium
  8. ○8#227Min GPUs Grid Searchmedium
  9. ○9#228Inference Memory Sweepmedium
  10. ○10#229Parameter Count Sweepmedium
  11. ○11#230FLOPs per Token (Batched)medium
  12. ○12#231Arithmetic Intensity Gridhard
Tracks are curated by hand. The order above is the suggested learning progression — feel free to skip around if you already know a topic.

© 2026 TorchedUp. All rights reserved.

ChangelogContact UsTerms of ServicePrivacy PolicyRefund Policy