Meta Releases ZML/LLMD LLM Server With Multi-Architecture Support
The production stack supports NVIDIA, AMD, and TPU hardware.
682313.1K
“Congrats to the @zml_ai team on the release of LLMD One thing I absolutely love is that the model layers are loaded directly from the network (from @huggingface OFC), in streaming mode (no local file) which makes a lot of sense for production deployment https://x.com/zml_ai/status/2074770878458417195 https://x.com/julien_c/status/2074797039473381451/photo/1”
Julien CTECH#335Story Brief
Meta's reported release of ZML/LLMD, an LLM server with multi-architecture support, is circulating on X through @julien_c, which points back to a post from @zml_ai. Related context includes Meta's Llama 4 announcement and the vLLM documentation, but the public claim still needs to stay tied to the source making it. For now, the clearest read is what was claimed, who put it in front of people on X, and what still needs more confirmation before anyone treats it as settled.
“RT @steeve: We're releasing ZML/LLMD, our homegrown LLM server built on top of our homegrown high performance heterogeneous inference stack…”
Ylecun
