Question 1

How much VRAM does cogvlm2-llama3-chat-19B need?

Accepted Answer

cogvlm2-llama3-chat-19B (19.5B) requires approximately 43.592841911315915 GB of VRAM at full precision (FP16), 21.8 GB at 8-bit quantization (Q8), and 10.9 GB at 4-bit quantization (Q4). The actual VRAM usage may be slightly higher due to KV cache and runtime overhead.

Question 2

What GPUs can run cogvlm2-llama3-chat-19B?

Accepted Answer

We have 10 compatible GPUs indexed for cogvlm2-llama3-chat-19B, including A40, L40S, L40, A100 80GB PCIe, A100 80GB and 5 more. Compatibility is determined by whether the GPU's VRAM can fit the model's memory footprint, potentially using quantization to reduce memory usage.

Question 3

What is the cheapest GPU to run cogvlm2-llama3-chat-19B?

Accepted Answer

The cheapest GPU to run cogvlm2-llama3-chat-19B in the cloud is the A40 at $0.44/hr. Other affordable options include L40S ($0.60/hr) and L40 ($0.99/hr). Prices reflect the lowest on-demand rate across cloud providers.

Question 4

Can cogvlm2-llama3-chat-19B run on consumer hardware?

Accepted Answer

Yes, cogvlm2-llama3-chat-19B can run on consumer GPUs like the RTX Pro 6000 Blackwell (96 GB VRAM) at full precision (FP16).

Question 5

What quantization options are available for cogvlm2-llama3-chat-19B?

Accepted Answer

cogvlm2-llama3-chat-19B supports the following quantization levels: FP16 (full precision): ~43.592841911315915 GB, Q8 (8-bit): ~21.8 GB, Q6 (6-bit): ~16.3 GB, Q4 (4-bit): ~10.9 GB. Lower quantization reduces VRAM usage at the cost of some quality loss. Q8 offers a good balance between quality and memory savings, while Q4 provides maximum compression.

GPU	VRAM	Best Quant	From	Rent
A40	48 GB	Q8	$0.44/hr	Rent
L40S	48 GB	Q8	$0.60/hr	Rent
L40	48 GB	Q8	$0.99/hr	Rent
A100 80GB PCIe	80 GB	FP16	$1.39/hr	Rent
A100 80GB	80 GB	FP16	$1.39/hr	Rent
H100 MEGA	80 GB	FP16	$1.80/hr	Rent
H100 SXM	80 GB	FP16	$1.80/hr	Rent
H100 NVL	94 GB	FP16	$1.80/hr	Rent
H100 PCIe	80 GB	FP16	$1.80/hr	Rent
RTX Pro 6000 Blackwell	96 GB	FP16	$1.89/hr	Rent

cogvlm2-llama3-chat-19B

VRAM Requirements

Compatible GPUs

cogvlm2-llama3-chat-19B FAQ

VRAM requirements for cogvlm2-llama3-chat-19B at different quantization levels
Quantization	VRAM Required
FP16	43.6 GB
Q8	21.8 GB
Q6	16.3 GB
Q4	10.9 GB