Question 1

What AI models can run on K80?

Accepted Answer

The K80 (12 GB VRAM) can run models including paligemma-3b-ft-vizwizvqa-448-jax, intent-aware-lfqa-qwen3-4b-intent-implicit, intent-aware-lfqa-qwen3-4b-baseline, Llama-3.2-3B-Instruct-QLORA_INT4_EO8, Llama-3.2-3B-Instruct-SpinQuant_INT4_EO8 and more — 10 compatible models indexed so far. Compatible models are determined by whether the model's memory footprint fits within the GPU's 12 GB VRAM, potentially using quantization (Q8, Q6, or Q4) to reduce memory usage.

Question 2

What is the cheapest cloud option for K80?

Accepted Answer

The cheapest on-demand option for the K80 is Microsoft Azure at $0.18/hr (Standard_NC6_Promo). We track pricing across 4 providers to help you find the best deal.

Question 3

Is K80 good for fine-tuning?

Accepted Answer

The K80 has limited capacity for fine-tuning with only 12 GB of VRAM. It can handle QLoRA fine-tuning of small models (up to ~3B parameters) but is not recommended for larger training workloads. Consider a GPU with more VRAM for serious fine-tuning.

Question 4

How much VRAM does K80 have?

Accepted Answer

The K80 has 12 GB of VRAM (GDDR5). It also offers 240 GB/s memory bandwidth, 8.73 TFLOPS of FP16 compute.

Question 5

What quantization formats work on K80?

Accepted Answer

The K80 supports the following quantization formats: Q8 (8-bit), Q6 (6-bit), Q4 (4-bit). Q4 quantization can fit models requiring up to ~48 GB at FP16 into 12 GB.

K80

Specifications

Cloud pricing

Compatible models

Compare this GPU

Frequently asked questions

VRAM	12 GB
Memory type	GDDR5
Bandwidth	240 GB/s
FP16 TFLOPS	8.73
FP8 support	No

K80