Alex Xi's Notes
Search
Search
Dark mode
Light mode
Tag: inference
26 items with this tag.
Jul 06, 2026
tweet-test-time-scaling-extent
test-time-compute
scaling
inference
Jul 06, 2026
tweet-vibe-coded-cuda-inference
inference
cuda
vibe-coding
Jul 06, 2026
vinegar-dumplings-sglang
sglang
inference
optimization
open-source
Jul 06, 2026
vllm-triton-attention-backend
vllm
triton
attention
inference
gpu-kernels
Jul 06, 2026
tweet-building-spectre
agents
inference
speculative-decoding
Jul 06, 2026
tweet-dwarkesh-frontier-training-lecture
training
inference
frontier-labs
educational
Jul 06, 2026
tweet-hw-sw-codesign-prefill-decode
inference
hardware
prefill-decode
Jul 06, 2026
tweet-kv-cache-nightmares
kv-cache
inference
hardware
Jul 06, 2026
tweet-model-size-from-knowledge
model-size
scaling
knowledge
inference
Jul 06, 2026
next-llm-cost-saving
inference
cost
efficiency
scaling
Jul 06, 2026
prefill-as-a-service
inference
kv-cache
disaggregation
moe
infrastructure
Jul 06, 2026
quantization-from-ground-up
quantization
inference
tutorial
Jul 06, 2026
supporting-mixtral-gpt-fast
inference
moe
torch-compile
gpu-kernels
Jul 06, 2026
turboquant-kv-cache-compression
quantization
kv-cache
inference
compression
efficiency
Jul 06, 2026
cursor-warp-decode
inference
moe
gpu-kernels
blackwell
cursor
Jul 06, 2026
gpu-mode-and-kernel-engineering
gpu-kernels
cuda
inference
ml-engineering
Jul 06, 2026
hitting-1000-tokens-rtx5090
inference
rtx-5090
tokens-per-second
gpu-optimization
Jul 06, 2026
hw-sw-codesign-matx-video
hardware
codesign
inference
efficiency
Jul 06, 2026
lilian-weng-transformer-inference
inference
transformers
optimization
efficiency
llm
Jul 06, 2026
llm-inference-economics
inference
economics
cost
gpu
Jul 06, 2026
a-visual-guide-quantization
quantization
inference
efficiency
llm
Jul 06, 2026
advancing-search-augmented-lms
rag
search
inference
Jul 06, 2026
prefill-decode-disaggregation
inference
infrastructure
kv-cache
Jul 06, 2026
test-time-compute
test-time-compute
chain-of-thought
reasoning
inference
o1
r1
Jul 06, 2026
fireworks-ai
infrastructure
rl
inference
Jul 06, 2026
moonshot-ai
company
inference
china