STRATA Reserve capacity
Strata/Use cases/Fine-tuning

GPU servers for fine-tuning

Fine-tuning adapts an open model to your data. The GPU you need depends on the method more than on the model.

Reserve capacityAll GPUs

LoRA and QLoRA

Parameter-efficient methods train small adapters on top of a frozen model. A 7–13B model fine-tunes on one RTX PRO 6000 or one H100; 70B with QLoRA fits on a single 80–96 GB card.

Full fine-tuning

Updating every weight needs memory for gradients and optimizer state, roughly 16 bytes per parameter. A 7B model needs 2–4 H100s; 70B needs a full 8-GPU H100 or H200 node with sharding.

Typical run lengths

Most fine-tunes take hours to a few days. Rent on demand, billed per minute, and stop the server when the run finishes.

Recommended GPUs

NVIDIA H100 80GB
Hopper · 80 GB HBM3

H100

from $3.23/GPU-hr

Specs and pricing →
NVIDIA H200 141GB
Hopper · 141 GB HBM3e

H200

from $4.31/GPU-hr

Specs and pricing →
NVIDIA RTX PRO 6000 Blackwell 96GB
Blackwell · 96 GB GDDR7

RTX PRO 6000

from $2.03/GPU-hr

Specs and pricing →
FAQ
What GPU do I need to fine-tune Llama 8B?

One H100 or RTX PRO 6000 is enough for LoRA and QLoRA. Full fine-tuning is faster on 2–4 H100s.

Do you provide ready environments?

Servers come with PyTorch and CUDA images, and our engineers help with the first deployment.

From the blog