The Software Frontier
Subscribe
Sign in
Home
Notes
Chat
Start Here
Archive
Leaderboard
About
Latest
Top
Discussions
DeepSeek V4-Flash: The Cost of Deciding What to Read
284 billion parameters rebuilt from the published constants, a million-token cache in 3.37 GiB, and the arithmetic showing that 4/5 of the attention…
Aug 10
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
9
2
2
Invite your friends to read The Software Frontier
A warm thank you
Aug 4
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
8
1
How Blackwell’s Tensor Memory Actually Works
Blackwell's largest matrix instruction needs 256 registers per thread. The ceiling is 255.
Aug 3
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
7
2
July 2026
How CUDA Binaries Actually Work
A byte-level surgical dissection of the cubin and fatbin formats, and the second encoding nobody documents.
Jul 31
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
6
3
How the NVIDIA Compiler Moat Actually Works
Inside the NVIDIA Compiler Moat: ptxas, SASS, and the 21 Bits Nobody Else Can Write
Jul 27
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
4
3
Kimi K3: 2.8 Trillion Parameters, Four Bits at a Time
Moonshot just announced the largest open-weight model ever built. The parameter count is the headline. The serving stack is the story.
Jul 22
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
6
1
3
Exploring how MLIR works: the compiler rewiring the AI stack
From tensor graphs to machine code, MLIR is quietly becoming the abstraction layer connecting modern AI frameworks to increasingly specialized hardware.
Jul 15
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
8
1
3
How LLVM Works: The IR That Took Over Modern Computing
A two-person grant at the University of Illinois became the compiler substrate under Apple, Android, the PlayStation, Google and Meta datacenters, and…
Jul 12
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
15
2
The Wafer & the Wallet
Memory costs are inflecting up as enterprise token budgets slam shut. The wafer math, the pass-through proof, and the engineering playbook that defends…
Jul 5
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
8
2
June 2026
The Router and the Wire
Mixture-of-experts promised cheaper inference by doing less arithmetic. The bill did not disappear. It moved into the network, and the entire shape of a…
Jun 29
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
8
2
Decode Is Memory-Bound. Speculation Is the Arbitrage
Speculative decoding is the only inference optimization that turns idle silicon into tokens without changing a single output. Whether that lands on your…
Jun 25
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
4
3
The Split and the Seam
Eighteen months ago, splitting prefill from decode was a contrarian research bet. Today it is the default that every serious stack runs, it helped erase…
Jun 21
•
Lorenzo Bradanini
and
Lorenzo Tettamanti
5
1
2
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts
