Skip to content
← CUDA, Triton & GPU Performance / 12
expertNewNVIDIAOpenAIAnthropic

Blackwell added tensor memory and 2-CTA MMA. What problem does each one solve, and what changes in a GEMM kernel written for it?

Hopper's biggest GEMM tiles spend half the register file on the accumulator and feed every operand byte through one SM's shared memory. Blackwell attacks both limits, and the answer counts the bytes each change gives back.

Updated Oct 2026 · Worked answers on inference mechanisms, with calculations, implementation details, and follow-up questions.

Hopper's biggest GEMM tiles spend half the register file on the accumulator and feed every operand byte through one SM's shared memory. Blackwell attacks both limits, and the answer counts the bytes each change gives back.

10 answers per topic instead of 5, plus saved progress and bookmarks · costs nothingor unlock all 136 remaining answers · ₹2,000 / $25

Every worked answer and derivation in all 12 topics, every concept page, and the full practice-test bank.

6 months · One payment · No auto-renewal

UP NEXT ON YOUR JOURNEY
Next in this trackCould you keep the KV cache of hot sequences resident in the H100's 50 MB L2 during decode, and would it be worth it?Next in this trackYou are about to report one number for your new kernel in a design review. How do you measure it so the number survives scrutiny?Next in this trackA kernel runs 20% slower on an H200 than on an H100, even though the H200 has 43% more memory bandwidth. How can that happen?
DISCUSSION · 0

Nobody has commented on this answer. Add a correction or another way to work it.