Why the E8 lattice is the perfect quantizer for KV caches
João André Marques
Running 1M-token context on a single GPU (the math)
João André Marques
NexusQuant is now on PyPI, HuggingFace, and 9 awesome lists
João André Marques
Why attention-aware eviction beats random eviction (with data)
João André Marques
The 12 approaches I tested before finding one that works
João André Marques
KV cache memory calculator: how much does your LLM actually use?
João André Marques
How to benchmark NexusQuant on your own model
João André Marques
What I Learned Testing 12 Compression Approaches That Failed
João André Marques
The Math Behind E8 Lattice Quantization (with Code)
João André Marques
How Much GPU Memory Does NexusQuant Actually Save?
João André Marques
How to deploy NexusQuant in production (and what's missing)
João André Marques
NexusQuant vs KVTC vs TurboQuant vs CommVQ — honest comparison
João André Marques
NexusQuant benchmarks: every number, honestly
João André Marques
Longer contexts are easier to compress (not harder)
João André Marques
Why E8 lattice quantization beats scalar quantization for KV caches
João André Marques
Compress your LLM's KV cache 33x with zero training
João André Marques