Can the NVIDIA RTX PRO 6000 Blackwell run GLM-5 744B-A40B?

RTX PRO 6000 Blackwell × GLM-5 744B-A40B8K context
NoDoes not fit in memory
Memory at Q4_K_M452.4 GB of 96 GB
Weights 451.1 GBKV cache 0.7 GBRuntime 0.6 GB
On the GPU21%of the Q4_K_M weights
In system RAM356.4 GBthe rest spills over
Offload speed~1.9tokens per second
  1. FP16
  2. Q8_0
  3. Q6_K
  4. Q5_K_M
  5. Q4_K_M
  6. Q3_K_M
  7. Q2_K

No. Even at Q4_K_M, GLM-5 744B-A40B needs about 452.4 GB, but the NVIDIA RTX PRO 6000 Blackwell has 96 GB usable. You could offload some layers to system RAM, but it will be very slow.

Run it in the cloud

Too big for the NVIDIA RTX PRO 6000 Blackwell? Run GLM-5 744B-A40B on a rented GPU instead.

Rent a cloud GPU by the hour. No upfront cost, stop any time. Not sure whether to rent or buy? Compare the cost.

Memory needed by quantization

At 8,192 tokens of context. The black line marks the 96 GB the NVIDIA RTX PRO 6000 Blackwell can use.

QuantQualityMemoryFits?Speed
FP16Full quality1489.3 GBToo big—
Q8_0Near-lossless791.8 GBToo big—
Q6_KExcellent611.4 GBToo big—
Q5_K_MVery good530.5 GBToo big—
Q4_K_MGood, most popular452.4 GBToo big—
Q3_K_MNoticeably degraded365 GBToo big—
Q2_KHeavily degraded280.3 GBToo big—

Measured speed from real RTX PRO 6000 Blackwell owners

Tokens per second that people measured on their own NVIDIA RTX PRO 6000 Blackwell, next to our estimate.

No one has measured GLM-5 744B-A40B on the NVIDIA RTX PRO 6000 Blackwell yet. The speeds above are estimates. Run the benchmark and yours becomes the first real number on this page.

Have an RTX PRO 6000 Blackwell? Add your real speed

One command times GLM-5 744B-A40B in your own Ollama, shows you exactly what it would send, and asks before sending anything. No sign-up, no personal data.

macOS and Linux

curl -fsSL https://gpufit.dev/bench.sh | sh

Windows (PowerShell)

irm https://gpufit.dev/bench.ps1 | iex

How the benchmark works

Frequently asked questions

Can NVIDIA RTX PRO 6000 Blackwell run GLM-5 744B-A40B?

No. Even at Q4_K_M, GLM-5 744B-A40B needs about 452.4 GB, but the NVIDIA RTX PRO 6000 Blackwell has 96 GB usable. You could offload some layers to system RAM, but it will be very slow.

How much VRAM does GLM-5 744B-A40B need?

About 452.4 GB at Q4_K_M with 8,192 tokens of context, and about 1489.3 GB at full FP16 precision.

How fast is GLM-5 744B-A40B on the NVIDIA RTX PRO 6000 Blackwell?

It does not fit in the NVIDIA RTX PRO 6000 Blackwell's memory, so most layers would run from system RAM and generation would be very slow.