The Mobile PCI Express Module standard (MXM) used to be a popular GPU solution for high-end gaming laptops in the early 2000s and 2010s, but now, ever since Nvidia stopped focusing on directly supplying this type of modules to laptop OEMs, the standard is seeing limited use mostly in industrial devices. Even though quite rare these days, we have seen MXM versions up to the Nvidia’s RTX 4000 Ada generation, but the latest RTX Blackwell variants are only now starting to get distributed to select industrial integrators. Aetina is among the first edge AI solution providers to offer Nvidia RTX Pro Blackwell Embedded GPUs which are said to deliver 3x the AI performance over the Ada generation.
Aetina is readying two MXM RTX Pro Blackwell versions. The MX5000B-XA is the high-end solution, while the MX500B-QA is the entry-level one. Each of these versions comes in two variants with different core clocks. The modules are designed to be compact, efficient, and rugged, with use cases such as autonomous mobile robots, unmanned vehicles, smart and automated manufacturing, as well as medical and surgical assistance systems.
Specs for the MX500B-QA include:
- 1792 CUDA cores
- 56 Tensor Cores
- 14 RT Cores
- 1350 MHz base clock / 1792 MHz boost -> 6.42 TFLOPS FPP (35 W TGP)
- 2160 MHz base 2565 MHz boost -> 9.19 TFLOPS FPP (60 W TGP)
- 6 GB 24 Gbps GDDR7 VRAM with 96-bit bus
- PCIe 4.0 X8
- MXM V3.1 Type A
The MX5000B-XA version is substantially more powerful:
- 10496 CUDA Cores
- 320 Tensor Cores
- 80 RT Cores
- 1095 MHz base / 1740 MHz boost -> 36.53 TFLOPS FPP (95 W TGP)
- 1402 MHz base / 1935 MHz boost -> 40.62 TFLOPS FPP (115 W TGP)
- 24 GB 28 Gbps GDDR7 VRAM with 256-bit bus
- PCIe 4.0 X16
- MXM V3.1 Type B
These two MXM modules are available to order now with direct inquiry. Aetina is also planning to launch MX400B and MX2000B modules in Q1 2026.
Source(s)
via TechPowerUp








