Rent an H100 for foundation-model pretraining and FP8 inference. 80 GB HBM3 at 3,350 GB/s, Hopper Transformer Engine, NVLink 4th-gen at 900 GB/s. Pretraining Llama-3 70B-class models hits ~7,200 tok/s/GPU on 8-GPU nodes via FSDP + FP8, single-card 70B FP8 serving at 1,800 tok/s with 32-request KV. The industry-default training silicon. Billed per-minute, paid in BTC, USDT/USDC or CLORE.
H100s are what you rent when minutes matter. FP8 cuts memory and doubles throughput on Transformer workloads; HBM3 keeps the SMs fed; NVLink lets eight cards train as one.
Hopper Transformer Engine roughly doubles A100 80GB pretraining throughput at the same memory footprint — that's why every published 70B+ pretraining recipe in 2025 targets H100. NVLink 4th-gen at 900 GB/s keeps gradient sync from being the bottleneck on 8-GPU pods.
TensorRT-LLM, vLLM, SGLang — all tuned for Hopper. Serve 70B FP8 from a single card with margin to spare for KV-cache.
80 GB HBM3 + FlashAttention-3 means 128k-token contexts run without offload. Ideal for long-doc RAG and agent loops with tool use.
Hopper is the architecture the 70B and 405B foundation models were trained on. Specs from Nvidia's SXM5 datasheet; pricing reflects the lowest live on-demand floor.
// prices are spot-market lows · refreshed every 60 s
Every server is priced by its host. These are the live floors across the marketplace — you'll see hundreds of variants once you're in.
No sales call. No quota request. No three-week procurement. The first four commands are all you need.
Filter the marketplace by H100 80GB, country, GPU count, reliability score, network speed.
Choose a Docker image — PyTorch, vLLM, ComfyUI, Blender — or paste your own.
You get a public endpoint, an SSH key, and Jupyter on port 8888 in under 90 s.
Per-minute billing rounds to the second. Stop the instance and the meter stops with it.
SXM is faster (700 W, NVLink 900 GB/s) and is what you want for distributed training. PCIe (350 W, NVLink Bridge optional) is fine for single-card inference and easier to host. CLORE.AI lists both — filter by NVLink speed in the marketplace.
Single-card, no - 70B pretraining needs an 8-GPU node minimum. CLORE.AI lists 8x H100, 8x H200, and 8x B200 pods with NVLink fabric for exactly this. A100 80GB pods run 70B FSDP training but at lower throughput than Hopper-class. For multi-week training, contact host operators for reserved-instance terms - listed in the marketplace under 'Reserved'.
A100 80GB has no FP8 - peak is BF16/TF32. H100 introduces FP8 with TransformerEngine and roughly 4x the BF16 training throughput at 2x the rental price - so ~2x perf-per-dollar on FP8-eligible workloads. H200 matches H100 compute but adds 141 GB HBM3e. B200 doubles H100 FP8 again with 192 GB HBM3e. Pick by VRAM and bandwidth ceiling, not just sticker FLOPS.
8-GPU H100 SXM, H200 SXM, and B200 nodes ship with NVSwitch fabric - 900 GB/s peer bandwidth on H100/H200, 1.8 TB/s 5th-gen NVLink on B200. PCIe variants (H100 PCIe, A100 PCIe) have NVLink Bridge in pairs only. Multi-node fabric (NVLink-Switch across racks) is available on B200 hyperscale pods - filter by 'NVSwitch' in the marketplace.
Yes. Multi-GPU listings expose all cards in a single rental as a coherent node with NVSwitch (where present), shared NVMe scratch, and InfiniBand or 100 GbE fabric for multi-node training. The standard PyTorch torchrun, DeepSpeed, and Megatron-LM launchers run unmodified. Filter the marketplace by GPU count to find 8x A100, 8x H100, 8x H200 nodes.
V100 (HBM2, 900 GB/s) -> A100 40GB (HBM2e, 1,555 GB/s) -> A100 80GB (HBM2e, 1,935 GB/s) -> H100 (HBM3, 3,350 GB/s) -> H200 (HBM3e, 4,800 GB/s) -> B200 (HBM3e, 8,000 GB/s). Each generation roughly doubles bandwidth or VRAM; KV-cache-bound serving and bandwidth-bound training scale almost linearly with this number.
80 GB HBM3 + 3.35 TB/s + Hopper FP8 — the gold standard for pretraining, FP8 fine-tuning, and 70B+ serving.
FP8 path nearly doubles A100 80GB pretraining throughput — the industry-default 70B pretraining card.
Read the guide →FP8 quant fits 70B + KV cache on a single 80 GB H100 — the canonical serving config.
Read the guide →Code-LLM fine-tuning at 16K context — H100 NVLink + Flash Attn 3 keep gradient sync from being the bottleneck.
Read the guide →Side-by-side specs across the datacenter tier. Click any row to see that GPU.
Step-by-step guides verified on CLORE.AI hardware. Pick a workload, copy the docker image, ship in minutes.
Per-minute payouts in BTC, USDT, USDC or CLORE. No listing fee, no contracts, withdraw any time.
Hosts around the world are accepting workloads right now. Sign up, top up your wallet, and the next hour is yours.