Rent an NVIDIA RTX 2060 GPU with 12 GB VRAM
Rent RTX 2060 cloud GPU servers by the hour on Cloudj: 12 GB VRAM, NVIDIA Turing with Tensor Cores and about 6.5 TFLOPS FP32 at a 130 W power limit. This 12 GB version is an older but roomy way to fit SDXL, 8B LLMs at 8-bit and small fine-tunes in memory, for learners and indie developers. Pick a live configuration, start from the Ubuntu 22.04 + CUDA or PyTorch + JupyterLab template and connect over SSH with your key.
RTX 2060 specifications
VRAM per GPU
12 GB
Architecture
NVIDIA Turing
Compute capability
7.5
FP32 compute
≈ 6.5 TFLOPS
Memory bandwidth
≈ 268 GB/s
Power limit
130 W
Values reported by the infrastructure provider for machines currently listed; real performance varies by machine.
Live RTX 2060 configurations and prices
Out of stock right now
Browse available GPUsWhat you can run on the RTX 2060 cloud GPU
SDXL on a budget card
12 GB fits SDXL in ComfyUI with a LoRA or two; generation is slower than on Ampere or Ada cards.
8B LLMs at 8-bit
Run Llama-class 8B models at 8-bit or 13B at 4-bit through Ollama for private chat.
FP16 mixed-precision learning
Turing Tensor Cores speed up FP16 training of small networks, so you can practice modern recipes.
Speech and vision inference
Run Whisper-style transcription or detection models that need more than 8 GB.
Is the RTX 2060 the right GPU for your workload?
Choose the 12 GB RTX 2060 when your workload needs more memory than entry-level consumer cards offer, without requiring the higher compute throughput of newer architectures. Upgrade to a newer generation or workstation GPU for heavy training, or pick a smaller GPU if your task requires minimal VRAM.
Guides for RTX 2060
RTX 2060 cloud GPU: frequently asked questions
Rent the RTX 2060 cloud GPU by the hour
Pick a live configuration, choose a template and connect over SSH with your key. Billing stops when you destroy the instance.