Can I run Dolphin3-Cyber-8B-GGUF on an RTX 3090?

Yes

Its F16 file (16.1 GB) fits in the 21.6 GB an RTX 3090 can use, with room for ≈ 41K tokens of context.

Full analysis of Dolphin3-Cyber-8B-GGUF Everything that runs on RTX 3090 Get the F16 file on Hugging Face

ollama run hf.co/RavichandranJ/Dolphin3-Cyber-8B-GGUF:F16

Files and context on RTX 3090

Each file of Dolphin3-Cyber-8B-GGUF (and of GGUF versions the site has analyzed), and how much context fits next to it, with the context cache in 16-bit or 8-bit.

File	Size	On RTX 3090	Context (16-bit)	Context (8-bit)
F16	16.1 GB	Fits	41K	82K
Q8_0	8.54 GB	Fits	97K	128K
Q6_K	6.60 GB	Fits	112K	128K
Q5_K_M	5.73 GB	Fits	118K	128K
Q5_0	5.60 GB	Fits	119K	128K
Q5_K_S	5.60 GB	Fits	119K	128K
Q4_K_M	4.92 GB	Fits	124K	128K
Q4_K_S	4.69 GB	Fits	126K	128K
Q4_0	4.66 GB	Fits	126K	128K
Q3_K_M	4.02 GB	Fits	128K	128K
Q2_K	3.18 GB	Fits	128K	128K

Dolphin3-Cyber-8B-GGUF on other hardware

Popular models that run on RTX 3090

Assumes 90% of the card's memory is usable and at least 4K tokens of context for one conversation, with the context cache in 16-bit unless noted. Fits leaves at least 20% of that memory for context; Tight leaves less.