Intel® Xeon® W-1290TE Processor / 10th Generation Intel® Core™ i9 Processor / NVIDIA GPU
Edge AI Computing System
Intel® Xeon® W-1290TE (20M Cache 1.8GHz up to 4.6GHz)
*Use only with Intel W480E
10th Generation Intel® Core™ i9-10900TE (20M Cache 2.0GHz up to 4.5GHz)
10th Generation Intel® Core™ i7-10700TE (16M Cache 2.0GHz up to 4.4GHz)
10th Generation Intel® Core™ i5-10500TE (12M Cache 2.3GHz up to 3.7GHz)
10th Generation Intel® Core™ i3-10100TE (6M Cache 2.3GHz up to 3.6GHz)
Intel® Celeron® Processor G5900TE (2M Cache 3.0GHz)
Q470E / W480E
2 x DDR4 2666/2933 MHz SO-DIMM up to 64GB (Optional ECC support with Xeon W-1290TE)
2 ~ 255 Level Reset
2.0
N/A
2 (MXM Graphics)
3 (Intel built-in GPU)
N/A
N/A
Max Resolution (DP 1.2) : 4096 x 2340 @ 60Hz
N/A
1
1
1
9 x Intel i210-AT and 1 x i219LM (iAMT9.5 w/ i5, i7 only) Gb/s Ethernet Controllers Onboard Support PXE and WOL
10 x RJ45 Ports for GbE （Optional 8 with M12 X Coded Connectors and ）
Optional 8 x PoE Total Max.120W
3 x RS 232/422/485 (option additional 1 x RS 232/422/485)
4 x USB 3.2 5Gbps Ports
8 x DI (5~48VDC) and 4 x DO (5VDC, 100mA)
N/A
N/A
N/A
2 x 2.5″ drive bay for SATA TypemPCIe
1 x M.2 2280 Key M slot supports NVMe and SATA SSD
*Thermal heatsink is required for M.2 storage. Please contact sales representative for more information.
0, 1, 5
2 x Full Mini-PCle Slots and 1 x Full Mini-PCIe Slot w/ USB interface only for WWAN sharing 2 x SIM Card Sockets with M.2 B Key 3042 Slot
1x M.2 A-E Key 2230 slot, 1 x M.2 B Key 3042 Slot w/ 2 x SIM Card Sockets for WWAN
Patented Automatic SIM Card Detection
9V – 48V DC Power Input
Automatics Recovery Short Circuit Protection
Vehicle Power Ignition for Variety Vehicle
Power off Delay Time Setting by BIOS and Software
Internal Battery Kit for 10 Mins Operating
(Charging Temp. 0~45oC, Discharging Temp. -10~60oC)
*UPS backup time varies depending on actual overall system power consumption and battery is required to be charged before using.
Patent No. : M447854 – Build-in Battery
260(L) x 250(W) x 95(H) mm
5300g (bare-bone)
Aluminum Alloy
Wall-mount
10% RH – 90% RH (non-condensing)
-40°C ~ 70°C w/0.6 m/s airflow (for barebone system)
*The operating temp. varies by accessories installed. Patent No. : M565941 – Thermal Cooling
-40°C ~ 80°C
N/A
N/A
CE, FCC Class A, E-Mark, ECE R118, EN 50155, EN 45545-2 R25
N/A
N/A
Windows 10 64bit
Ubuntu 20.04 64bit
N/A
ABOX-5210xyy-zz (x= P= w/PoE, x= Empty= w/o PoE, yy=G6=RTX 3000, yy=G7=T1000, zz=Z9= W-1290TE, zz= i9= i9-10900TE, zz= i7= i7-10700TE, zz= i5= i5-10500TE, zz= i3= i3-10100TE, zz= C1= G5900TE)
Intel® Xeon® W Processor / 10th Generation Intel® Core™ Processors w/ NVIDIA® GPU/ 10 x GbE (Optional 8 x PoE)/ 3 x COM/ 3 x DP/ 2 x HDMI®/ 4 x USB 3.2 Gen 5Gbps/ 9-48V Input AI GPU Computing
Made in Taiwan
| Model | ABOX-5210G6 | ABOX-5210G7 | ABOX-5210G8 | ABOX-5210G9 |
|---|---|---|---|---|
| MXM Graphics | NVIDIA® Quadro® RTX 3000 | NVIDIA® Quadro® T1000 | NVIDIA® Quadro® RTX A4500 | NVIDIA® Quadro® RTX A2000 |
| Graphic Memory | 6GB GDDR6 | 4GB GDDR6 | 16GB GDDR6 | 8GB GDDR6 |
| CUDA Cores | 1920 | 896 | 5888 | 2560 |
| Tensor Cores | 240 | N/A | 184 | 80 |
| RT Cores | 30 | N/A | 46 | 20 |
| System Operating Temp. | 0ºC ~ 60ºC w/0.6 m/s airflow | 0ºC ~ 70ºC w/0.6 m/s airflow | 0ºC ~ 60ºC w/0.6 m/s airflow | 0ºC ~ 70ºC w/0.6 m/s airflow |
A major smart city initiative sought to modernize urban tram networks to meet the increasing safety demands of shared road environments. To protect passengers and citizens, the transport authority required a sophisticated Open Road Tram Collision Prevention system. The project aimed to create a future-proof safety backbone capable of real-time tram obstacle detection and precise classification across a busy metropolitan tram fleet. Navigating Shared Right-of-Way: The Need for Ultra-Low Latency Operating trams on open roads shared with vehicles and pedestrians presents unique technical hurdles. Unlike closed-track systems, trams must navigate “urban canyons” where unpredictable obstacles frequently emerge. Ultra-low-latency collision prediction is required to provide the braking system or driver with sufficient reaction time. The system must support long-range obstacle detection using radar, cameras, and LiDAR, while maintaining stable operation under harsh rail vibration, power fluctuation, and extreme outdoor conditions. Accurate classification of obstacles is critical to avoid false alarms and ensure a timely, reliable response. Data-Driven Safety for Complex Transit Environments Beyond basic sensing, the city needed a collision avoidance solution that could bridge the gap between high-speed data acquisition and immediate mechanical action. The goal was to move toward a proactive safety model, using real-time telemetry to identify risks before they escalate. Given the high stakes of urban transit, the reliability of this edge AI technology was paramount to prevent accidents and service gaps. The system serves as a critical guardian, ensuring that trams can coexist safely with diverse road users in a dense metropolitan setting. The Solution: ABOX-5210 with High-Bandwidth Sensor Fusion The ABOX-5210 was implemented as the core computing unit for the collision prevention system. Powered by Intel® processors and NVIDIA® GPUs, it provides high-bandwidth multi-sensor fusion...
Read more
