Trade Shows

News

Electronic Engineering Journal reports on SXVPU
New Partnership: Solectrix & Image Engineering

Events

MedtecLIVE, Stuttgart, June 18–20, 2024

iVT Expo Cologne, June 26–27, 2024

> alle News und Termine

Solution

SXVPU AI Model

An AI-accelerated model that offers seamless integration of neural networks for AI vision applications.

SXVPU Datasheet (PDF)

sxvpu_ai_hailo_amd

Specifications

AI Accelerator
Hailo-8 Edge AI Processor with 26 TOPS
CPU
Quad-core Arm® Cortex®-A53 MPCore™ up to 1.5 GHz
GPU
Mali™-400 MP2 up to 667 MHz
Chipset
SoC integrated
Trusted Platform Module (TPM)
Infineon 2.0, Security TPM2.0 supporting IEC62443
System Memory
DDR4 4 GB onboard 64-bit
Storage
eMMC 16 GB, 64 MB QSPI FLASH

Expansion
M.2 Key M, 22 mm × 42 mm (PCIe Gen2 x4)
Construction
Universal aluminium profile enclosure
Pressure compensation element
Protection Class
IP65 (IP67 optional)
Operating Temperature
−40 °C to +75 °C (65 °C w/ max. performance, 85 °C w/ min. performance)
Weight (net/gross)
1.2 kg (2.65 lb) / 1.5 kg (3.31 lb)
OS Support
PetaLinux (Yocto-based)
Mounting
Wall brackets

Interfaces

Front

sxvpu_ai_front

Camera Input
4× GMSL2 via SMA connectors (FAKRA connectors optional)
Ethernet/PoE
1× 10/100/1000 Mbps Ethernet with hardware timestamping (IEEE 1588)
Digital I/O
1× CAN 2.0
1× General Purpose Input 12 V / 24 V
1× Status LED (RGB)
Power Supply
9 to 36 VDC + 1× Power Enable Input

Rear (during development phase only)

sxvpu_ai_rear

Display
1× Mini-DisplayPort Connector (DP++)
USB
2× USB 3.2 Gen1
Expansion
1× OCuLink 2 (PCIe Gen3 x4)
Debug
1× Debug USB with 3× UART + JTAG
Power Supply
1× Battery Socket
Storage
1× MicroSD Card Slot

SXVPU is a registered trademark of Solectrix GmbH. All other trademarks are the property of their respective owners.

Contact