Rent NVIDIA RTX 3080 Ti GPU by the hour
The NVIDIA RTX 3080 Ti is an Ampere-architecture consumer GPU equipped with 12 GB of VRAM and a 350 W TDP. On ONIE, it delivers about 35.1 TFLOPS of standard floating-point compute and a memory bandwidth of about 787 GB/s. With CUDA compute capability 8.6, this GPU handles intermediate machine learning workloads, inference, and creative rendering.
Specifications
VRAM per GPU
Architecture
Compute capability
FP32 compute
Memory bandwidth
Power limit
Values reported by the infrastructure provider for machines currently listed; real performance varies by machine.
Available configurations
Live stock and prices, updated continuously. The final price is confirmed before you order.
Loading live prices…
What the RTX 3080 Ti is good for
Quantized LLM Inference
Run compact and quantized large language models inside the 12 GB VRAM budget using ready templates like Ollama or vLLM.
Image Generation with ComfyUI
Generate images and test diffusion pipelines using the ComfyUI template, leveraging about 787 GB/s of memory bandwidth for prompt processing.
PyTorch Model Prototyping
Develop, debug, and train small-to-medium neural networks in PyTorch and CUDA on an Ampere GPU with compute capability 8.6.
GPU-Accelerated Rendering
Utilize about 35.1 TFLOPS of compute performance for rendering scenes and baking textures that fit within a 12 GB memory footprint.
Is it the right GPU for you?
The RTX 3080 Ti suits developers and creators who need strong Ampere compute performance for workloads that fit comfortably within 12 GB of VRAM. Choose a smaller consumer GPU for basic scripting or lightweight inference, or step up to a GPU with more VRAM if you are fine-tuning large models or loading unquantized weights.
Compare with similar GPUs
Frequently asked questions
Is the RTX 3080 Ti suitable for LLM inference and fine-tuning?
With 12 GB of VRAM and about 787 GB/s of memory bandwidth, it is well suited for running smaller or quantized language models via Ollama or vLLM. For full fine-tuning or running larger unquantized models, a GPU with a larger VRAM capacity is recommended.
How does the RTX 3080 Ti compare to adjacent Ampere consumer GPUs?
The RTX 3080 Ti sits between the standard RTX 3080 and the higher-memory consumer cards in the Ampere lineup, offering 12 GB of VRAM and about 35.1 TFLOPS of compute. If your workload is compute-bound and fits in 12 GB, it offers strong throughput without moving up to a higher-memory tier.
How do I connect to the RTX 3080 Ti instance on ONIE?
You can connect directly via SSH access once your instance is running. ONIE also provides pre-configured templates for PyTorch/CUDA, vLLM, Ollama, and ComfyUI so you can start working immediately.
How does billing work for this GPU?
Instances on ONIE use per-hour billing. You are billed only for the time your RTX 3080 Ti instance is active.
Rent the RTX 3080 Ti by the hour
Create an account, choose a configuration and connect over SSH. Billing stops when you delete the GPU.
Rent RTX 3080 Ti