Reports
Measured runs of Glyd against bf16 and the other lossless formats, each with its method, the command to repeat it and a link to the raw logs.
- Report ·
Qwen2.5-72B on two H100s
Qwen2.5-72B on two 80 GB H100s in vLLM: Glyd's KV cache holds 111,328 tokens against bf16's 8,800, room for about 27 conversations at once, not 2.
- Report ·
Nineteen open models: every matrix 32 to 33% smaller, bit for bit
Every Linear layer's matrix of nineteen open models, SmolLM3-3B to GLM-4.5-Air, packed and unpacked bit for bit: 31.9 to 33.0% smaller, quality on ten.
- Report ·
Qwen3.8-27B, the top open model for one GPU, in 41 GB
Qwen3.8-27B, the top-scoring open model that fits one GPU, in 41,071 MiB of GPU memory with Glyd against bf16's 51,771: under a 48 GB card, bit for bit.
- Report ·
Against DFloat11 and ZipServ on an RTX 4080 SUPER
Qwen3-8B on a 16 GB RTX 4080 SUPER: Glyd generates 2.5 to 3.4 times DFloat11's tokens a second at the same size, and its kernels keep pace with ZipServ's.
- Report ·
Qwen3-32B on one 48 GB GPU, bit for bit
Glyd holds Qwen3-32B's weights in 44.5 GB instead of 65.5, so it runs on one 48 GB RTX A6000 instead of two, every weight exact, faster to 32 sequences.
- Report ·
Ten popular open models: every matrix 32 to 33% smaller, bit for bit
Every Linear layer's matrix of ten popular open models, from SmolLM3-3B to Llama-3.3-70B-Instruct, packed and unpacked bit for bit: 31.9 to 33.0% smaller.