WOLF Advanced Technology, a world leader in SOSA Aligned High Performance Embedded Compute (HPEC) and artificial intelligence (AI) processing solutions for the extreme edge, has announced seven new VPX and XMC modules, built to improve the performance and reliability of synthetic vision, C4ISR, radar processing, autonomous operation, AI, and situational awareness applications.
These new WOLF modules will include one or more of the NVIDIA RTX™ A4500, NVIDIA RTX™ A2000, NVIDIA® ConnectX®, and NVIDIA Jetson AGX Xavier™ Industrial platforms.
System integrators and OEMs are projected to benefit from significant improvements in HPEC and AI performance per slot, up to 100GbE network speeds, PCIe Gen4 rapid data transfers, a new cooling approach, and features that provide system flexibility, scalability, reliability, and security.
Greg Maynard, CTO of WOLF, said, “SOSA is the future, and the WOLF team is enabling this future in novel ways by providing integrators and OEMs with the ability to ingest more than double the data, enjoy more than double the effective performance per slot, operate in hotter environments, and use the latest software libraries to distribute and scale workloads.
“We are adding scalable and flexible HPEC and AI processing modules that are interconnected with the fastest fabrics along with software-enabled scalability and easy-to-deploy technology refreshes. I am incredibly proud of what WOLF has been able to accomplish in this launch.”
Product details will be released beginning in November and will include:
VPX3U-XAVIER-CX6-HPC: This payload module is designed for scalable HPEC and includes an NVIDIA Jetson AGX Xavier Industrial system-on-module for data processing and AI inference. The WOLF module includes support for both PCIe Gen 4 and Ethernet with the integration of an NVIDIA ConnectX-6 SmartNIC for secure, high-speed data transfer, including RDMA over Converged Ethernet (RoCE). Enhanced security features can be enabled for hardware-verified secure boot, hardware-accelerated cryptography, encrypted storage, and more.
VPX3U-XAVIER-CX6-SBC: This SBC module provides autonomous operation for GPU-based data processing applications such as machine vision and sensor processing applications. This flexible module uses the SOSA aligned I/O Intensive SBC Slot Profile, with video I/O, USB, UART, optional NVMe storage, secure high-speed network connections, PCIe Gen 4, and other factory configurable options. For video I/O requirements this module includes the WOLF FGX for SDI video capture and output. Enhanced security features can be enabled for hardware-verified secure boot, hardware-accelerated cryptography, encrypted storage, and more.
VPX3U-A4500E-CX6: A 3U VPX solution for advanced HPEC and AI inference. This design features an NVIDIA RTX A4500 which provides CUDA cores for compute, Tensor Cores for AI, and RT Cores for ray tracing acceleration, and an NVIDIA ConnectX-6 SmartNIC for the high-speed, secure, data transfer required for data-heavy tasks such as sensor data processing and other C4ISR tasks. The ConnectX-6 also provides support for RDMA over Converged Ethernet (RoCE), enabling the fastest method for transferring data across the network to the GPU.
VPX3U-A4500E-VO: A 3U VPX video output and HPC / AI module which includes the NVIDIA RTX A4500 GPU for high performance processing and display, with a PCIe Gen 4 GPU.
VPX6U-A4500E-DUAL-VO: A 6U VPX video output and HPC / AI module which includes two NVIDIA RTX A4500 GPUs for high performance processing and display, with a PCIe Gen 4 GPU and PCIe switch.
XMC-A2000E-VO: A video output and HPC module which features a chip-down NVIDIA RTX A2000 design. This design is suited for lower-power applications which require a small form factor, high-performance-per-watt mezzanine board.
XMC-A2000E-FGX-IO: A video capture, process, and display module suited for lower-power, high-performance-per-watt applications. This module can provide converted video inputs and outputs for virtually any digital or analog signal requirement.