Can GeForce RTX 4070 run Gemma 2 9B?

✅ Yes, runs great!

Fits in GPU memory (~80 tok/s)

Memory Needed
10.8 GB
Q3_K_M, 8K context
Est. Speed
~80 tok/s
Generation speed
GPU VRAM
12 GB
GDDR6X

💡 Recommended

For best quality that fits: Q3_K_M quantization

All Quantizations

Q8_0
Runs slowly
Memory
15.8 GB
Speed
7 tok/s
Bits/weight
8.5
Quality
Highest
Q6_K
Runs slowly
Memory
13.7 GB
Speed
10 tok/s
Bits/weight
6.6
Quality
High
Q5_K_M
Runs slowly
Memory
12.6 GB
Speed
11 tok/s
Bits/weight
5.7
Quality
High
Q4_K_M
Runs slowly
Memory
11.8 GB
Speed
13 tok/s
Bits/weight
4.85
Quality
Good
Q3_K_M
RECOMMENDED
Runs great
Memory
10.8 GB
Speed
80 tok/s
Bits/weight
3.9
Quality
Lower
Q2_K
Runs great
Memory
9.9 GB
Speed
105 tok/s
Bits/weight
3
Quality
Lower

Run with Ollama

ollama run gemma2:9b-q3_k_m

This will download and run the Q3_K_M quantization