Tesla P4 8GB vs RX 7900 GRE 16GB for Local LLMs

The RX 7900 GRE 16GB has about 200% more memory bandwidth (576 vs 192 GB/s). Batch-1 decode tracks bandwidth closely, so a model that fits both cards should decode roughly 200% faster on the RX 7900 GRE 16GB. The RX 7900 GRE 16GB fits more of the catalog at Q4/8K — 24 vs 16 of 69 models (16 vs 8 GB VRAM). Largest model that fits the RX 7900 GRE 16GB but not the Tesla P4 8GB: gemma-4-26B-A4B-it (25.2B). MSRP is $80 for the Tesla P4 8GB vs $549 for the RX 7900 GRE 16GB. MSRP is a launch list price, not live retail — Amazon prices move; click through for the current price. In the Q4/8K head-to-head below (39 shared models), the RX 7900 GRE 16GB posts the higher tok/s in 39 rows vs 0. Methodology.

Spec comparison

	Tesla P4 8GB	RX 7900 GRE 16GB
Vendor	NVIDIA	AMD
VRAM	8 GB	16 GB
Memory type	GDDR5	GDDR6
Memory bandwidth	192 GB/s	576 GB/s
TDP	75 W	260 W
MSRP (list)	$80	$549
Buy	Search Amazon (affiliate)	Search Amazon (affiliate)

MSRP is the launch list price, not live retail. Buy links are Amazon affiliate links — the price after you click is the live price.

Q4 / 8K head-to-head — top 15 of 39 shared models

Model	Params (B)	Tesla P4 8GB tok/s	Tesla P4 8GB fits	RX 7900 GRE 16GB tok/s	RX 7900 GRE 16GB fits	Winner
MiniCPM5-1B	1.08	113.1 est.	Yes	339.4 est.	Yes	RX 7900 GRE 16GB
Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16	30.0	24.1 est.	No	289.9 est.	No	RX 7900 GRE 16GB
Qwen3-30B-A3B-Instruct-2507	30.5	21.9 est.	No	263.7 est.	No	RX 7900 GRE 16GB
GLM-4.7-Flash	31.2	21.0 est.	No	253.1 est.	No	RX 7900 GRE 16GB
Qwen3.5-35B-A3B	34.7	18.2 est.	No	219.2 est.	No	RX 7900 GRE 16GB
qwen35b-a3b-fable-sft-abliterated	34.7	17.3 est.	No	208.1 est.	No	RX 7900 GRE 16GB
Ornith-1.0-35B	35.0	17.0 est.	No	204.9 est.	No	RX 7900 GRE 16GB
Qwen3.6-35B-A3B	36.0	17.0 est.	No	204.3 est.	No	RX 7900 GRE 16GB
Qwen3.5-2B	2.0	61.1 est.	Yes	183.3 est.	Yes	RX 7900 GRE 16GB
DeepSeek-R1-0528-Qwen3-8B-layer-mix-bpw-3.8-mlx	2.3	53.1 est.	Yes	159.4 est.	Yes	RX 7900 GRE 16GB
Llama-3.2-3B-Instruct	3.2	38.2 est.	Yes	114.5 est.	Yes	RX 7900 GRE 16GB
gpt-oss-20b	21.5	33.6 est.	No	101.8 est.	Yes	RX 7900 GRE 16GB
gemma-4-26B-A4B-it	25.2	23.9 est.	No	96.5 est.	Yes	RX 7900 GRE 16GB
Qwen3-4B-Instruct-2507	4.0	30.5 est.	Yes	91.6 est.	Yes	RX 7900 GRE 16GB
NVIDIA-Nemotron-3-Nano-4B-BF16	4.0	30.5 est.	Yes	91.6 est.	Yes	RX 7900 GRE 16GB

est. = bandwidth-based estimate; measured = curated real benchmark that overrides the estimate. Fits = weights + KV(8K) + 1 GB overhead ≤ 95% of VRAM.

Fit coverage at Q4 / 8K

	Tesla P4 8GB	RX 7900 GRE 16GB
Catalog models that fit	16 of 69	24 of 69
Largest exclusive fit	—	gemma-4-26B-A4B-it (25.2B)

Tesla P4 8GB details → · RX 7900 GRE 16GB details →