MXM development kits pave the way for AI
General purpose graphics processor units (GPU) and edge AI computing platform provider, Aetina, has announced MXM development kits for embedded artificial intelligence (AI) and vision computing.
The development tools include an MXM motherboard and an MXM to PCIe carrier that can help the developers to build a test system to confirm the performance specifications and verify peripheral devices early in the development stage.
The kit includes the MXM motherboard, MH110, for evaluation. It supports Aetina MXM 3.0/3.1 type A/B modules and Intel’s Core i7 CPUs, up to 7th Generation. A DDR4 SO-DIMM 2133MHz dual channel is expandable up to 32Gbyte, sufficient for entry-level AI, says the company.
For display, there is also a HDMI 1.4 and four DisplayPorts for multi-display use.
For peripheral I/O extension, the MH110 has a Mini-PCIe slot, a M.2 B-key and M-key for Wi-Fi, Bluetooth, 4G LTE, storage or capture card options.
Aetina also provides the RTM-M3C-6DP carrier board that can convert the Aetina MXM 3.1 type A/B module into a PCI Express x16 Gen3 interface. This can be used by designers with little embedded MXM experience to test the performance and computing capability of a graphics card, shortening the develop cycle.
This Aetina MXM development kit series can be used with Aetina MXM modules, claimed to be the most compact, thinnest graphics modules based on the industrial standard Mobile PCI Express Module (MXM) Version 3.0/3.1 specification, powered by an Nvidia Quadro GPU. The graphics performance, GPU computing and video capabilities are particularly suitable for embedded AI and vision computing application such as medical imaging, defence, gaming and deep learning applications, says Aetina.
Aetina was founded in 2012 and is based in Taiwan. It is a long-term support provider of high-performance GPGPU and Jetson edge AI computing solutions for embedded applications. Aetina focuses on the industrial market, providing industrial components and service and is also an integrator of artificial intelligence of things (AIoT).