Model comparison

NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16 vs GLM-5.2

These two publish different quantization sets; the table below has the exact sizes.

From the file· summed bytes, KV per layer

Side by side

NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16GLM-5.2
Parameters561B753B
Architecturenemotron_h_moeglm-dsa
Layers78
Native context262,1441,048,576
Mixture of expertsyes, 512 expertsyes, 256 experts
Quantizations published1520
Smallest quantization175.12 GiB169.33 GiB
Q4_K_M
Licenceothermit

KV cache by context

the term that decides long-context viability
ContextNVIDIA-Nemotron-3-Ultra-550B-A55B-BF16GLM-5.2Ratio
4,0960.34 GiB
8,1920.69 GiB
16,3841.37 GiB
32,7682.74 GiB
65,5365.48 GiB
131,07210.97 GiB