RunsOnMy ModelsGPUs & MacsBy VRAMBest GPUCan I run…?OllamaStart hereAPI pricesVRAM calculatorNew

Can ERNIE 5.5 Flash Base run on the H100 SXM?

Not fully in VRAM. Even at 2-bit with 8K context, ERNIE 5.5 Flash Base needs about 101 GB, more than the 76.0 GB usable on the H100 SXM. It can still run with offload (slower): Q4_K_M with part of the model in system RAM. Details and command.

Runs with offload (slower)

Runs with offload 4-bit (Q4_K_M) at 8K context: all attention layers plus the experts of 19 of 43 MoE layers on the H100 SXM (about 74.3 GB), and the experts of the other 24 layers in system RAM (about 89.1 GB). Needs a PC with at least 128 GB of system RAM. Speed is not estimated for this model (its active-parameter count is uncertain).

llama-server -m ernie-5-5-flash-base-Q4_K_M.gguf -c 8192 -ngl 999 --n-cpu-moe 24

Estimate, not a benchmark. No GGUF repo was found on Hugging Face yet; point -m at your own GGUF file. The speed range spans dual-channel DDR4-3200 to DDR5-6000 system RAM; your CPU, RAM speed and llama.cpp version move it. If llama.cpp runs out of GPU memory, raise --n-cpu-moe by 1-2; if VRAM is left over, lower it for more speed. Recent llama.cpp builds can also pick the split automatically. How we estimate offload.

Newer compressed-attention design: long-context figures are an upper bound.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K)101 GB No101 GB No103 GB No108 GB No—
3-bit (Q3_K_M)131 GB No132 GB No133 GB No138 GB No—
4-bit (Q4_K_M)164 GB No164 GB No166 GB No171 GB No—
5-bit (Q5_K_M)191 GB No191 GB No193 GB No198 GB No—
6-bit (Q6_K)220 GB No220 GB No221 GB No227 GB No—
8-bit (Q8_0)284 GB No284 GB No286 GB No291 GB No—
16-bit (F16/BF16)534 GB No534 GB No536 GB No541 GB No—

• real GGUF file size. Speed is a bandwidth ceiling (3350 GB/s), not a benchmark. Methodology.

Hardware that does fit it

DGX SparkRyzen AI Max+ 395 128GBH200Mac M2 Ultra 192GBMac M3 Ultra 256GBMac M3 Ultra 512GB

What fits instead

"Runs" and "fits" mean the whole model in GPU memory at 8K context (4-bit or better for the model pick). Prices are US launch MSRPs, not today's street prices. "Popular" means the 60 most-downloaded models we track.

More

ERNIE 5.5 Flash Base on every GPU and Mac · Every model that runs on the H100 SXM · Other models that fit 64 GB VRAM · Try other settings in the calculator

Guides: Best GPU for local LLMs by budget · Can I run gpt-oss, Llama 70B or Qwen3? · Mac vs PC for local LLMs

Data: Hugging Face Hub (baidu/ERNIE-5.5-Flash-Base), NVIDIA specs. Model data last checked 2026-11-18; refreshed hourly.