Application Highlight: ADLINK AXE-7400SR Drives Advanced AI Edge Solutions

November 04, 2024

Sponsored Story

AI applications for computing at the Edge are at the peak of excitement and interest across every vertical, and enterprises from the industrial to the consumer are looking for ways to incorporate the power of AI.

They are looking particularly for those solutions that can streamline and automate operations, reduce costs, drive efficiency, and even create new profit centers. Sensor fusion is helping with automation and efficiency by making IoT devices smarter and more able to filter out noise in the incoming data, only reporting key information. In robotics, high-powered embedded AI computers are leveraging sensor networks and customized data sets to drive warehouse and manufacturing automation, allowing humans to shift focus to new initiatives and growth.

The rise of Generative AI is causing the AI world to shift its point of view again and everyone is now trying to see if they should also leverage the power of LLM and GenAI strategies to power leading applications and operations. Limiting factors at the Edge have been the power needed, and the memory required, to run GenAI apps on small form factor and low power chips.

Many engineers and developers are working on these problems from both the software and hardware side, but there are early successes happening.

Application Use Case

SimProBot is a US based company that’s creating dedicated on-premise generative AI solutions at the Edge for enterprises. The company’s Tallgeese AI is designed to capitalize on Gen AI shift by offering small and medium-sized enterprises (SMEs) access to GenAI tools without the need for internal coding expertise. The company says it automates repetitive tasks and enhances productivity for traditional manufacturing companies that offer a wide variety of products.

These companies face challenges due to non-standardized processes, so Tallgeese AI integrates product data from internal document management systems to create a comprehensive “Super Product Manager” that can assist sales teams with real-time customer needs and even address growth areas related to knowledge transfers and sales processes.

This kind of desktop computing is the first step in practical GenAI applications for user interface and direct uses for Ai applications that can answer natural language queries. And this solution is powered by compute from a particularly powerful engine. Let’s take a look.

ADLINK AXE-7400SR

The ADLINK AXE-7400SR is one of the AI GPU servers from the company that can function as the hardware platform for Tallgeese AI. Enterprises looking to leverage GenAI tools can choose hardware platforms that match the scale of the deployment, pair them with appropriate GPU cards, and add in various acceleration cards according to application needs to create a full solution that can meet the demands of their on-premise AI applications, ranging from 100 to 6000 TOPS.

The AXE-7400SR series is designed for on-premise AI applications, and it can help enterprises fine-tune AI models across diverse fields and data sets in order to drive the development of applications like chatbots, code writers, and large language models (LLM). The series uses 2U/4U short chassis designs, so the form factor is designed for on-premise environments.

The customizability of the platform gives users flexible and diverse options for application development to meet any smart manufacturing need. ADLINK can add in motion control cards, IO cards, and image capture cards in addition to the GenAI tools. It features a single 4th Gen Intel Xeon Scalable Processor, 8x DDR5 DIMMs at 4800MT/s (8 channels, 1DPC), and supports up to 5x full-height PCIe slots for GPU/FPGA accelerator cards. It also includes a BMC with AST2600, IPMI v2.0/Redfish compliant, a dust-proof design, and a short chassis, making it suitable for factory and automation environments.

Specifications:

• Chipset: Emmitsburg C741 PCH
• Memory: 8x DDR5 DIMMs, 4800MT/s (8 channels, 1DPC)
• BIOS Chip: AMI BIOS on SPI flash Memory
• I/O Interface
  • Expansion: 4x FHFL single-slot PCIe Gen5 x16 slots and 1x FHFL Single-slot PCIe Gen5 x8 slot or 4x FHFL dual-slot PCIe Gen5 x16 slots
• Ethernet:
  • SKU1: 1x GbE port
  • SKU2: 1x GbE + 2x 10GbE ports
• USB
  • 2x USB3.2 Gen1 (rear)
  • 2x USB2.0 (rear)
  • 2x USB2.0 (front)
• Graphics
  • 1x VGA Port
  • 1x DB-9 COM Connector
• Storage:
  • 1x NVMe M.2 2280 slot (PCIe 3.0 x4)
  • 6x SATA-III 3.5“/2.5” disk with RAID 0/ 1/ 10/ 5 (Intel RSTe)
• BMC Controller:
  • Aspeed AST2600 with IPMI 2.0 & Redfish support
  • 1x GbE dedicated for IPMI
• LEDS: HDD, NIC1, NIC2, UID
• Operating System:
  • Windows Server 2022
  • Ubuntu 22.04LTS
• Mechanical and Environmental:
  • Form Factor: 4U 19“ rackmount, 650 x 437.5 x 176.5mm
  • Fans: 3x fans, smart fan speed control
  • Power: 1600W/2000W, AC
• TPM: Plug-in TPM 2.0 module (Optional)
• Temperature and Humidity:
  • Operating: 0°C to +35°C, 5%-95% RH @40°C, non-condensing
  • Storage: -40°C to +70°C, 5%-90% RH, non-condensing
• GPU Support:
  • 4x 2-slots GPU cards
  • 2x 3-slots GPU cards

ADLINK has invested and dedicated itself to making GenAI applications and tools available to enterprises of all sizes, from the data center to the edge and the AXE-7400SR is just one step on that path. The company said it looks forward to accelerating enterprise AI adoption and enhancing the benefits of digital transformation, and we can’t wait to see what’s next.

Additional Resources:

Product Page:  https://www.adlinktech.com/products/edge_server/edge_server/axe-7400sr?lang=en

Product Brief: https://www.adlinktech.com/Products/Download.ashx?type=MDownload&isDatasheet=yes&file=2163%5cDatasheet_AXE-7400SR_20240913.pdf