readygpu

GMKtec EVO-X2 Ryzen AI Max+ 395 128GB vs NVIDIA GeForce RTX 3090 Ti

For running models locally, priced in Canada. At 128k context the first runs 65 models fully in memory, the second 25.

In short

The GMKtec EVO-X2 Ryzen AI Max+ 395 128GB has more memory for models: 128 GB against 24 GB. At 128k context the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB runs 65 of the models we track fully in memory and the NVIDIA GeForce RTX 3090 Ti 25; only the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB holds Qwen 3.8 Flash Next. On the models both run fully in memory, the NVIDIA GeForce RTX 3090 Ti is about 4.4× as fast as the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB (estimated, median across models). In Canada the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB costs $5,099 at Amazon and the NVIDIA GeForce RTX 3090 Ti $3,999 at Amazon: the NVIDIA GeForce RTX 3090 Ti is $1,100 cheaper.

Updated . Speeds are estimates; how they are worked out.

GMKtec EVO-X2 Ryzen AI Max+ 395 128GB

$5,099

Amazon

Add to build
NVIDIA GeForce RTX 3090 Ti

$3,999

Amazon

Add to build

Specifications

SpecGMKtec EVO-X2 Ryzen AI Max+ 395 128GBNVIDIA GeForce RTX 3090 Ti
ChipRyzen AI Max+ 395 (Radeon 8060S)–
Memory bus256-bit384-bit
Memory bandwidth256 GB/s1,008 GB/s
FP16 compute59.4 TFLOPS80 TFLOPS
SoftwareROCmCUDA
Unified memory128 GB–
GPU can use by default96 GB–
GPU can use, tuned120 GB–
Operating systemwindows–
ProcessorRyzen AI Max+ 395–
Storage2000 GB–
GPU–GeForce RTX 3090 Ti
Memory–24 GB GDDR6X
Board power–450 W
PCIe generation–Gen 4
PCIe lanes–x16

What each runs

Estimated decode speed with the most precise quant that fits each one est.

ModelGMKtec EVO-X2 Ryzen AI Max+ 395 128GBNVIDIA GeForce RTX 3090 Ti
Qwen 3.8 Flash Next3tok/sUD-Q3_K_XLwon't run
Qwen 3.8 27B4tok/sQ8_0offload
Qwen 3.5 122B-A10B8tok/sQ5_K_Moffload
Qwen 3.6 27B4tok/sQ8_0offload
Qwen3 Coder Next11tok/sQ8_0won't run
gpt-oss 120B11tok/sMXFP4won't run
Nemotron 3 Super 120B-A12B8tok/sQ5_K_Mwon't run
Gemma 4 31B3tok/sQ8_0offload
Mistral Small 4 119B10tok/sQ6_Koffload
Qwen 3.6 35B-A3B15tok/sQ8_0offload
Qwen 3.5 27B4tok/sQ8_0offload
Gemma 4 26B-A4B12tok/sQ8_064tok/sQ5_K_M
GLM 4.5 Air2tok/sQ4_K_Mwon't run
Qwen 3.5 35B-A3B15tok/sQ8_0offload
Qwen3 Next 80B-A3B13tok/sQ8_0won't run
GLM 4.7 Flash7tok/sQ8_0offload
Devstral Small 2 24B3tok/sQ8_0offload
Nemotron 3.5 Lightning 30B-A3B20tok/sUD-Q6_K_XLoffload
Kimi Linear 48B-A3B20tok/sQ8_0offload
gpt-oss 20B15tok/sMXFP466tok/sMXFP4
Qwen3 Coder 30B-A3B5tok/sQ8_0offload
Qwen3 32B2tok/sQ8_0won't run
Gemma 4 12B9tok/sQ8_041tok/sQ8_0
Qwen 3.5 9B10tok/sQ8_045tok/sQ8_0
Gemma 3 27B4tok/sQ8_0offload
Mistral Small 3.2 24B3tok/sQ8_0offload
DeepSeek R1 Distill Qwen 32B2tok/sQ8_0won't run
Llama 3.3 70B2tok/sQ5_K_Mwon't run
Ministral 3 14B4tok/sQ8_0offload
Qwen3 14B4tok/sQ8_0offload
DeepSeek R1 0528 Qwen3 8B5tok/sQ8_0offload
Qwen3 8B5tok/sQ8_0offload
Phi-4 14B3tok/sQ8_0offload
Qwen 3.5 4B16tok/sQ8_071tok/sQ8_0
Ministral 3 8B5tok/sQ8_027tok/sQ4_K_M
Granite 4.2 8B5tok/sQ8_0offload
Gemma 4 E4B12tok/sQ8_052tok/sQ8_0
LFM2.5 8B-A1B11tok/sQ8_050tok/sQ8_0
Llama 3.1 8B6tok/sQ6_K27tok/sQ5_K_M
Qwen2.5 0.5B66tok/sQ8_0292tok/sQ8_0
Qwen3.5 0.8B58tok/sQ8_0256tok/sQ8_0
Qwen3.5 2B46tok/sQ5_K_M204tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B25tok/sQ8_0111tok/sQ8_0
Mellum2.1 12B A2.5B Thinking18tok/sQ8_082tok/sQ8_0
MiniCPM5 2B17tok/sQ8_075tok/sQ8_0
LFM2.5 2.6B13tok/sQ8_057tok/sQ8_0
LFM2.5 VL 3B13tok/sQ8_057tok/sQ8_0
Kimi VL A3B12tok/smlx-8bit51tok/smlx-8bit
granite vision 4.1 4b10tok/sQ8_044tok/sQ8_0
granite 4.2 3b10tok/sQ8_043tok/sQ8_0
GLM 4.6V Flash9tok/sQ8_041tok/sQ8_0
DeepSeek R1 Distill Qwen 7B9tok/sQ8_040tok/sQ8_0
GLM 4.1V 9B Thinking8tok/smlx-8bit37tok/smlx-8bit
LFM2.5 24B A2B7tok/sQ8_0offload
DeepSeek R1 Distill Llama 8B6tok/sQ8_027tok/sQ5_K_M
Ministral 3 8B Reasoning 25125tok/sQ8_027tok/sQ4_K_M
DeepSeek R1 Distill Qwen 14B3tok/sQ8_0offload
Devstral Small 25073tok/sQ8_0offload
Magistral Small 25063tok/sQ8_0offload
Magistral Small 25073tok/sQ8_0offload
Magistral Small 25093tok/sQ8_0offload
GLM 4.5V3tok/sQ4_K_Mwon't run
GLM 4.6V3tok/sQ4_K_Mwon't run
granite 4.2 30b2tok/sQ8_0won't run
DeepSeek R1 Distill Llama 70B2tok/sQ5_K_Mwon't run
DeepSeek V4 Flashwon't runwon't run
GLM 5.3 Flashwon't runwon't run
MiniMax M2.7won't runwon't run
DeepSeek R1won't runwon't run
DeepSeek R1 0528won't runwon't run
DeepSeek V3 0324won't runwon't run
DeepSeek V3.1won't runwon't run
DeepSeek V3.2won't runwon't run
DeepSeek V4 Flash 0731won't runwon't run
DeepSeek V4 Flash Vision Expwon't runwon't run
Devstral 2 123B Instruct 2512won't runwon't run
GLM 4.7won't runwon't run
GLM 5.1won't runwon't run
GLM 5.2won't runwon't run
GLM 5.3won't runwon't run
Kimi K2won't runwon't run
Kimi K2 Instruct 0905won't runwon't run
Kimi K2 Thinkingwon't runwon't run
Kimi K2.5won't runwon't run
MiniMax M2won't runwon't run
MiniMax M2.1won't runwon't run
MiniMax M2.5won't runwon't run
MiniMax M3won't runwon't run
Mistral Medium 3.5 128Bwon't runwon't run

Questions

Which is better for local AI, the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB or the NVIDIA GeForce RTX 3090 Ti?
Choose the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB to run bigger models. Choose the NVIDIA GeForce RTX 3090 Ti for speed. Choose the NVIDIA GeForce RTX 3090 Ti to spend less.
Which can run bigger models, the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB or the NVIDIA GeForce RTX 3090 Ti?
At 128k context the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB runs 65 of the models we track fully in memory and the NVIDIA GeForce RTX 3090 Ti 25; only the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB holds Qwen 3.8 Flash Next.
Which is faster for local AI, the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB or the NVIDIA GeForce RTX 3090 Ti?
On the models both run fully in memory, the NVIDIA GeForce RTX 3090 Ti is about 4.4× as fast as the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB (estimated, median across models).
What is the price difference between the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB and the NVIDIA GeForce RTX 3090 Ti?
In Canada the GMKtec EVO-X2 Ryzen AI Max+ 395 128GB costs $5,099 at Amazon and the NVIDIA GeForce RTX 3090 Ti $3,999 at Amazon: the NVIDIA GeForce RTX 3090 Ti is $1,100 cheaper.

More comparisons