Can 2x RTX 3090 run Llama-3.3-70B-Instruct?
Yes: 2x RTX 3090 runs Llama-3.3 70B at 19.1 tok/s at Q4_K_M, measured on llama.cpp with a 135-token prompt in October 2026.
1 signed run on llm-speed. Single-request decode speed in tokens per second; the short-chat workload unless noted. This run was measured by llm-speed (@llmspeed). Runs measured by llm-speed are credited to @llmspeed.
Measured tok/s by quant and runtime
| Quant | Runtime | Build | Decode tok/s (best run) | Workload | Runs | Source |
|---|---|---|---|---|---|---|
| Q4_K_M | llama.cpp | commit f74e2fa | 19.1 | chat-short | 1 | Measured by llm-speed@llmspeed |
Fastest quant that fits: Q4_K_M on llama.cpp, 19.1 tok/s (measured by llm-speed; it fits because it ran on this hardware). Quants without a run are not listed.
Every measured row
| Run | Date (UTC) | Workload | Quant | Runtime / build | Prompt tokens | Output tokens | Decode tok/s | Prefill tok/s | TTFT ms | VRAM GiB | Source |
|---|---|---|---|---|---|---|---|---|---|---|---|
| r_n93ukk0mkmm | 2026-10-17 | chat-short | Q4_K_M | llama.cpp · commit f74e2fa | 135 | 256 | 19.1 | 415.4 | 325 | – | Measured by llm-speed@llmspeed |
| r_n93ukk0mkmm | 2026-10-17 | chat-long | Q4_K_M | llama.cpp · commit f74e2fa | 3162 | 452 | 18.7 | 939.3 | 3366 | – | Measured by llm-speed@llmspeed |
| r_n93ukk0mkmm | 2026-10-17 | agent-trace | Q4_K_M | llama.cpp · commit f74e2fa | 3235 | 840 | 18.8 | 2028.8 | 1037 | – | Measured by llm-speed@llmspeed |
| r_n93ukk0mkmm | 2026-10-17 | concurrent-decode | Q4_K_M | llama.cpp · commit f74e2fa | 1024 | 256 | 19.0 | – | – | – | Measured by llm-speed@llmspeed |
Questions
Can 2x RTX 3090 run Llama-3.3-70B-Instruct?
Yes: 2x RTX 3090 runs Llama-3.3 70B at 19.1 tok/s at Q4_K_M, measured on llama.cpp with a 135-token prompt in October 2026.
How many tokens per second does Llama-3.3-70B-Instruct get on 2x RTX 3090?
Decode speed measured by llm-speed, by quantization and runtime: Q4_K_M on llama.cpp: 19.1 tok/s. Based on 1 run measured by llm-speed; single-request chat-short unless noted.
What is the fastest quant of Llama-3.3-70B-Instruct that fits on 2x RTX 3090?
Q4_K_M on llama.cpp was the fastest measured setting at 19.1 tok/s; it fit because it ran. Other quants are listed only where a run measured them.
How much VRAM does Llama-3.3-70B-Instruct use on 2x RTX 3090?
Memory use was not recorded for these runs. The VRAM calculator gives an estimate, not a measurement.