Can GeForce RTX 4090 run Gemma 2 27B?
✅ Yes, runs great!
Fits in GPU memory (~43 tok/s)
Memory Needed
23.0 GB
Q4_K_M, 8K context
Est. Speed
~43 tok/s
Generation speed
GPU VRAM
24 GB
GDDR6X
💡 Recommended
For best quality that fits: Q4_K_M quantization
All Quantizations
Q8_0
Memory
35.5 GB
Speed
N/A
Bits/weight
8.5
Quality
Highest
Q6_K
Memory
29.1 GB
Speed
6 tok/s
Bits/weight
6.6
Quality
High
Q5_K_M
Memory
25.3 GB
Speed
7 tok/s
Bits/weight
5.7
Quality
High
Q4_K_M
RECOMMENDEDMemory
23.0 GB
Speed
43 tok/s
Bits/weight
4.85
Quality
Good
Q3_K_M
Memory
20.0 GB
Speed
54 tok/s
Bits/weight
3.9
Quality
Lower
Q2_K
Memory
17.3 GB
Speed
70 tok/s
Bits/weight
3
Quality
Lower
Run with Ollama
ollama run gemma2:27bThis will download and run the Q4_K_M quantization