RCX-1540R-PEG: Embedded PC Supports Dual NVIDIA GPUs for AI at the Edge

Author: James Gane
RCX-1540R-PEG: Embedded PC Supports Dual NVIDIA GPUs for AI at the Edge
  • Dual GPU support for dual NVIDIA graphics cards
  • Independent power to support up to two 350W CPU cards
  • Powerful Intel C246 chipset with support for RAID, AMT and TPM2.0
Vecow’s RCX-1540R-PEG is a high-performance dual GPU AI system with a rich selection of I/O and flexible storage options. Featuring the powerful Intel C246 chipset, the RCX-1540R-PEG supports up to 80W Intel Xeon and 9th Gen Core processors and can be combined with up to two NVIDIA GPU cards offering impressive AI compute performance right at the Edge.
The RCX-1540R-PEG has an LGA1151 CPU socket and can support a range of high-performance processors including the 8-core, 16-thread Intel Core i9-9900 whilst the on-board Intel C246 chipset offers additional support for flagship Intel Xeon CPU’s should your application require advanced features such as Error Correction Code (ECC) memory. The RCX-1540R-PEG also features four DDR4 SO-DIMM slots supporting up to 128GB of 2666MHz DDR4 memory as well as M.2 M-key support for SSD’s with four front access SATA trays for HDD/SSD expansion

No embedded computer is complete without a rich selection of I/O and the RCX-1540R-PEG is no different with three serial ports supporting RS-232/422/485, six 10Gbps USB3.2 Gen 2 ports and two Gigabit Ethernet ports powered by Intel-based network controllers, one of which features support for iAMT 12.0. For additional comms expansion, there are two full-size MiniPCIe card slots with SIM support, M.2 E-key for Wi-Fi support and three SMA antenna breakouts.

Dual PCIe x16 support for GPUs

AI inference and GPU usage is increasingly being used at the Edge in industrial applications. In autonomous vehicle applications for example, it would be dangerous for any delay between input from sensors and action by the Edge device. As autonomous vehicles develop, their respective neural networks grow more complex and demand more compute performance to be supplied right at the Edge.

The RCX-1540R-PEG has two PCIe x16 ports that support x8 bandwidth supporting up to two 350W GPUs, this power limit includes the current flagship GPU from NVIDIA the RTX 3090 which boasts a colossal 10496 CUDA cores and 24GB GDDR6X of memory, capable of an astonishing 72 Tera Flops of FP32 performance perfect for deep neural networks. Additionally, the RCX-1540R-PEG can cater for professional cards such as the RTX A4000 which has a single slot footprint allowing for additional PCI Express expansion cards to be added to meet specific application needs.

Our Embedded Systems Capabilities

As with all of our embedded computing range, Impulse can fully configure the RCX-1540R-PEG to your exact specification in our UK based engineering facility with your choice of processor, memory, storage, peripheral cards, and operating system.

With our Embedded Systems capabilities, we can create reliable, repeatable and revision controlled systems aimed at reducing your overall project costs and development time. With our team of in-house engineers and specialists, all with decades of experience, we can offer you fully deployable embedded computing solutions straight out of the box.

For more details regarding our Embedded Systems capabilities please click here.
Get in touch
Our technical sales team are ready to answer your questions.
T: +44 (0)1782 337 800 • E: sales@impulse-embedded.co.uk
+44(0)1782 337 800
WE'RE HIRING!If you are looking for a new and exciting challenge please take a look at our recruitment page
Emily N.
Nov 2021
+44(0)1782 337 800
MediaNewsRCX-1540R-PEG: Embedded PC Supports Dual NVIDIA GP...