NVIDIA GTC 2025
Location: San Jose Convention Center
Date: March 18th to 21st
Booth # 1023
Meet Our Experts
Highlighted Products
  • Omni Inserter and Integra Gluing

     

    • High-Force Insertion: Programmable modes with force monitoring for precise placement
    • Versatile Dispensing: Offers various glue dispensing applications
    • Flexible Manufacturing: CAD import, board shuffle, and best-fit insertion for adaptability
    • Digital Twin: Simulates processing for investment evaluation & visualized integration
    • User-Friendly: Auto-calibration & modular design for fast line changeover

     

  • D-Bot Collaborative Robot

     

    • NVIDIA Omniverse integration
    • High payloads up to 30 kg, arm reach 800 to 1800 mm
    • Repeatability margin of ±0.02mm with a speed up to 3 m/s
    • Robust design with IP66 rating and easy setup in less than one hour
    • Innovative Reflex Safety for human protection

     

  • DIATwin Virtual Machine Development Platform

     

    • Virtual machine prototyping facilitates the deployment of new machines
    • High-fidelity simulation, accuracy up to 95% ~ 99%
    • Cyber-physical integration & logic process validation
    • Offline/Virtual Commissioning & recipe generation

     

  • DIAEAP+ Platform with Line Manager Application

     

    • Easy setup to get the equipment's data
    • Dashboard & visualization for real-time equipment status
    • Low-Code development enable seamless data integration and applications (for example: Omniverse)
    • Shopfloor Control manages PCB flow, error-proofing, and equipment recipe updates

     

  • 19" 72kW 800V HVDC Power Shelf

    The debuted 72kW Power Shelf integrates two 36kW PSUs with 800V high-voltage output and outstanding 98% power efficiency. The power shelf features 800V high-voltage output distribution, and satisfied the power requirement for next generation AI data centers.

  • Power Capacitance Shelf

    This innovative power solution mitigates the GPU dynamic load reflections to the AC grid, such as AI server EDPP and Idle mode transitions. The PCS is able to perform fast charge and discharge power transitions due to the advantages of the Lithium-Ion Capacitor (LIC). The LIC provides high power density and long hold-up time (15-second/20kW load) to ensure a stable power supply for AI and cloud computing servers and therefore increase the power system reliability. The PCS is now available for 19” (1RU) and 21” (1OU) ORV3 standard racks.

  • 22 kW Battery Backup System

    The 22kW battery backup system with 2RU compact size is an ideal solution for AI data Center. With 97.5% efficiency and UL9540A-certified, it ensures reliability and significant OPEX and carbon savings.

  • Inverter Power System

    The inverter power system is playing the leader role in data center application with 11.4A of the power for 230V system. It is modular, expandable and can be easily installed in 19-inch enclosure, also with high cost-performance ratio.

  • Regulated Bus Converter

     

    • Supports 40~60V input and provides 12V fully regulated output
    • Power up to 2000W in Quarter Brick size; and 4000W in size of 68.0x60.5x15.8mm (follow NVIDIA GB200 PDB power requirement)
    • Peak Efficiency up to 98.5% @54Vin
    • Fully protected: Input UVLO, Output OVP, OCP and OTP
    • Suitable for air cooling and liquid cooling

     

  • High Voltage DCDC Solution

     

    • Horizontal design, suitable for cold plate cooling
    • Input and Ouput: Cable/BusBar connection
    • Complete Protection: Input Hotswap, E-fuse, Input UV/OVP, Output OCP/OVP, etc

     

  • Power Choke

     

    • High power density (Dual / Quad choke)
    • High current, high operating temp. (150°C)
    • Compliant with immersion liquid cooling test conditions
    • Ultra thin package for VPD and SiP module application
    • Ultra low inductance with high performance in high frequency. (10~100MHz)

     

  • 3 in 1 e-Breaker - Circuit Breaker, Fuse and pre-charge bundle

     

    • High current handling – Continuous current 600A
    • Low turn on resistance. Less than 2mΩ
    • Current, voltage, and temperature sensing
    • Pre-charge capability within 100ms (Conditional) – No extra components
    • Ultrafast disconnect time – Less than 5us
    • Compact size and light weight. – 115x65x70 mm3, <700g

     

  • 1.5 MW CDU

    The Liquid-to-Liquid cooling solution provides up to 1.5 MW of cooling capacity. This solution is specifically designed to handle the thermal challenges of multiple high-density racks, each exceeding 100kW, ideal for AI driven workloads. 

  • 6U L2L In-Rack CDU - 200kW

     

    • Cooling capacity: 200kW
    • Flow rate: 200LPM @20 psi
    • Higher pumping power with low approach temp
    • Pump Redundant 2+1
    • HMI screen with touch functionality
    • Leakage, dew point, coolant level protections
    • Dimension: 445(W) x 815(D) x 264(H) mm

     

  • 4U L2L In-Rack CDU - 135kW

     

    • Cooling capacity: 135kW
    • Flow rate: 138LPM @20 psi
    • Higher pumping power with low approach temp
    • Pump Redundant 2+1
    • HMI screen with touch functionality
    • Leakage, dew point, coolant level protections
    • Dimension: 445(W) x 815(D) x 175(H) mm

     

  • 20U L2A In-Rack CDU - 24kW

     

    • Cooling capacity: 24kW
    • Flow rate: 30LPM @10 psi
    • Pump & PSU redundant design
    • Hot swap FAN design
    • Leakage, coolant level protections
    • Dimension: 450(W) x 880(D) x 890(H) mm

     

  • Cold Plate Loop for GPUs

     

    • Cooling capacity: 6200 W (4*AI GPU+2*CPU)
    • Microchannel & impingement cooling design
    • Precision liquid flow distribution design
    • Hot swappable / dry disconnect capability
    • Flexible copper tube design
    • Dimension: 547(L) x 190(W) x 29(H) mm

     

  • 4U 3D Vapor Chamber

     

    • Cooling capacity (single 3D VC): over 1000W
    • Cooling capacity (total): 8750W
    • 4U high performance thermal solution
    • Efficient heat transfer and cooling
    • Dimension: 486(L) x 416(W) x 151.3(H) mm

     

  • Server Fans

     

    • Server Fan Array- Delta’s latest 40/60/80/92mm server fans boast of industry-leading performance
    • Up to 60% fan efficiency, provide powerful fans design for efficient AI compute cooling

     

  • DC Rack Fans

     

    • Industry leading high-performance efficiency
    • Multiple section blade design for low noise
    • PWM speed control function
    • Operating voltage from 40V to 60V
    • IP55 ingress protection

     

  • EC Rack Fans (HVDC & AC)

     

    • Support 400 HVDC power requirement
    • Option to support HVDC (400V- 800V)
    • High efficiency metal 3D Curve blade
    • Spider bracket for structure enhancement
    • New generation EC motor for lightweight and motor temperature reduction
    • Low THDi build in for good power quality
    • IP54 protection

     

  • MGX Rack

    NVIDIA MGX rack features a 1,400A busbar and a 42RU manifold design

  • Core Shell Liquid Cooled Busbar (1/4 Length)

     

    • Power Dissipation: 400 KW+
    • Application: Power and IT Shelf for High Current (4000 A~8000 A) or High Voltage (+/-400 V vs. 800 V)
    • High Flexibility: Optimized Fin Channel and Flow Rate for Excellent Heat-Exchange
    • High Usability: Facile Assembly and Installation (MGX and ORV3 Compatible)

     

  • Proprietary Pin-Type HV Busbar (1/4 Demo Sample)

     

    • Power Dissipation: 140 KW+
    • Application: Miniaturized Power and IT Shelf (32.5 mm x 50 mm) with High Current Capacity (2500 A+)
    • High Flexibility: Optimized Fin Channel and Flow Rate for Excellent Heat-Exchange
    • High Usability: Facile Assembly and Installation (Customized Design Applicable)

     

AI Partners with
AI Partners with
Contact Us Today
Want to schedule a meeting or can’t make it to the show? Let’s get in touch now.