Xiphos Q9 Versal Computer-on-Module
Xiphos Systems Corporation
High-performance AI/ML-capable Versal-based COTS computer-on-module in a tiny form factor, built for intensive onboard processing.
Technical specifications
- Operating temperature
- -40 to +60 °C
- Size
- 90 mm x 110 mm x 20 mm
- Mass
- 56 g without connectors
- Power
- 30-50 W typical; 8 to 14 VDC; over-current detection & protection (global and local) and brownout protection
- Memory
- 16 GB LPDDR4 DRAM (2x8GB); 1GB (4x 256 MB) QSPI Flash (NOR); 2x NVMe SSD, 1,000 GB each, on independent buses/firecode power control; 2x LPDDR4 interfaces PS and PL-accessible, 8GB each; 1MB MRAM for critical telemetry/state data
- MPSoC model
- AMD Versal System-on-Chip
- Application Processing Unit
- Dual core ARM Cortex A72 Application Processing Unit at up to 1.7 GHz
- Real-Time Processing Unit
- Dual core ARM Cortex R5F Real-Time Processing Unit at up to 750 MHz
- AI Engines
- 152 AI Engine-ML Tiles
- System Logic Cells
- 820,000
- Flip Flops
- 750,000 FF
- Look-up Tables
- 375,000 LUT
- DSP Slices
- 984
- Control FPGA
- Microchip PolarFire
- Operating system
- Yocto Linux BSP (LTS distribution)
- Real Time Clock
- RTC with sleep & wake up on alarm/interrupt; dedicated power pin for external battery
- Interfaces
- 2 mezzanine connectors providing 1.8V/3.3V GPIO, 1.8V MIO, LVDS; 26.5625 Gbps GTY transceivers supporting PCIe Gen 4, Ethernet, JESD204b, SSD, SERDES, etc.; 2 multi-rate MACs (each 1x100G or 2x50G/40G or 4x10G)
About
The Q9 is the first time a Versal-based platform has been available in such a small form-factor that can fit almost anywhere. It is also a true commercial grade platform with all of the benefits of COTS but with the tools and systems in place to make development, demonstration and launch fast and reliable. It is ideal for development of the next generation of intensive AI applications. The card extends the capability of the Xiphos Q-Card processor family, adding support for high-speed JESD204B interfaces and PCIe, as well as access to 2TB of high speed data storage. The Q9 is ideal for advanced payload applications, requiring onboard AI/ML processing, I/O-intensive applications, and high-speed C&DH systems. At its core is a hybrid environment of powerful multi-core CPUs and reprogrammable logic, providing flexible and scalable performance, primarily aimed at missions that need intensive processing capabilities such as real-time sensor processing using AI/ML and can afford the power budget of 30-50 W.
Documentation
Need the full ICD, test reports or a specific revision? Ask the supplier directly.
