ScalablyAI — European B300 inference provider integration
#91
by pavle-scalably - opened
Hi Hugging Face team — cc @julien-c @Wauplin @SBrandeis @hanouticelina
I’m Pavle Lazić, founder of ScalablyAI. We are preparing a European 8× NVIDIA B300 / 2.304 TB HBM3e inference node and would like to build it from day one to become a Hugging Face Inference Provider.
We’re deliberately keeping the first B300 deployment flexible. If Hugging Face has an underserved model, task, region, latency tier, or privacy/ZDR requirement, we’d rather configure the node around that real demand than choose a model arbitrarily. We can adapt the serving stack, model mix, pricing and deployment accordingly.