Can the Radeon RX 7900 XTX run InternLumina-U2?
InternLumina-U2·16B·Radeon RX 7900 XTX·24 GB VRAM
Yes. The Radeon RX 7900 XTX runs InternLumina-U2 fully in VRAM at Q8_0 (about 19 GB of 24 GB), at roughly 31 tokens/s.
- best quant
- Q8_0
- needs
- 19 GB
- usable
- 24 GB
- est. speed
- ~31 tok/s
Estimates for an 8K context; offload assumes 32 GB of system RAM.Try other contexts on the InternLumina-U2 page →
Every quantisation of InternLumina-U2 on the Radeon RX 7900 XTX
| Quant | Memory | Fit | Speed |
|---|---|---|---|
| Q2_K | 7.1 GB | fits | ~100 tok/s |
| Q3_K_M | 9.6 GB | fits | ~68 tok/s |
| Q4_K_M | 11 GB | fits | ~54 tok/s |
| Q5_K_M | 13 GB | fits | ~46 tok/s |
| Q6_K | 15 GB | fits | ~40 tok/s |
| Q8_0 pick | 19 GB | fits | ~31 tok/s |
| FP16 / BF16 | 33 GB | offload | ~3.2 tok/s |
Speed is a bandwidth-based estimate for generating text (960 GB/s); prompt processing and long contexts are slower.
Measured speed on the Radeon RX 7900 XTX
No community reports for this pair yet.
Report your own speed →Omni models that fit the Radeon RX 7900 XTX comfortably
All that fit →| Model | Params | Best quant | Memory | Speed |
|---|---|---|---|---|
| 12B | Q8_0 | ≈16 GBFits in memory | ~42 tok/s | |
| 14.7B | Q8_0 | ≈17 GBFits in memory | ~34 tok/s | |
| 17.5B | Q8_0 | ≈20 GBFits in memory | ~28 tok/s | |
| 10.7B | Q8_0 | ≈8.9 GBFits in memory | ~46 tok/s | |
| 12B | Q8_0 | ≈16 GBFits in memory | ~42 tok/s | |
| 24.3B | Q6_K | ≈21 GBFits in memory | ~26 tok/s |
Other GPUs that run InternLumina-U2
Frequently asked questions
- Can the Radeon RX 7900 XTX run InternLumina-U2?
- Yes. The Radeon RX 7900 XTX runs InternLumina-U2 fully in VRAM at Q8_0 (about 19 GB of 24 GB), at roughly 31 tokens/s.
- How fast does InternLumina-U2 run on the Radeon RX 7900 XTX?
- An estimate from memory bandwidth gives about 31 tokens/s at Q8_0.
- Which quantisation of InternLumina-U2 should I use on the Radeon RX 7900 XTX?
- Q8_0: the largest one that fits. Expected quality: lossless compared with the full-precision weights.
Memory and speed are estimates from the model's size and the card's memory bandwidth; real numbers depend on the runtime, drivers and context length. Community figures are medians of user reports.