## Slide 1

![The slide features the ADLINK logo at the top left with the tagline 'LEADING EDGE AI COMPUTING'. The main title text reads 'AI Xeleration Edge (AXE) Server Sales Kit'. Below the title, the text 'NCA' is displayed. In the bottom left corner, a white box contains the text 'Q2 2026 v1.2'. On the right side, images of a green motherboard, a white server chassis, and a black server chassis are shown against a blue, mountainous background. To the right of the servers is the logo 'AXE'.](.axe-sales-kit-q2-2026-v1-1-1/slide-001.jpg)

## Slide 2

![The slide is titled 'Opportunity: Real-Time Intelligence at the Edge' and features three blue panels with statistics:\n\n1.  **50+%**\n    Enterprise AI inference will shift from cloud → edge(Gartner) by 2025\n2.  **92%**\n    of users see lower latency with edge platforms for real-time, real-time, mission-critical-critical decisions\n3.  **30%**\n    Local inference network costs reduced & boosts response\n    **2-3x**](.axe-sales-kit-q2-2026-v1-1-1/slide-002.jpg)

## Slide 3

![This slide is titled **'Bridging the Gap: Overcoming Edge Complexity'**.\n\nBelow the title, it states: 'Deploying AI at the edge presents unique challenges that traditional IT infrastructure cannot address.'\n\nThe slide lists three key requirements for edge computing, accompanied by icons on the left and hardware images on the right:\n\n**Low Latency**\n'High-stakes decisions require sub-millisecond responsiveness. We eliminate the delay of cloud round-trips.'\n\n**Scalability**\n'Moving from a single POC to thousands of global nodes requires a unified, scalable hardware-software architecture.'\n\n**Durability**\n'From vibration-heavy factory floors to extreme outdoor temperatures (-40°C to 85°C), the edge demands industrial-grade reliability.'\n\nThe slide footer contains the number **02**.](.axe-sales-kit-q2-2026-v1-1-1/slide-003.jpg)

## Slide 4

![**Title:** Edge AI Groundbreaker\n\n**Section 1: Pioneering Partnerships**\nA leading innovator in Edge AI, collaborating with NVIDIA to deliver industrial-grade solutions\n\n**Section 2: Comprehensive Portfolio**\nFrom ultra-compact platforms to high-density servers & modules to systems, covering the full spectrum of Edge AI computing needs\n\n**Timeline Section:**\n**2020** Support MXM GPU Module Powered by NVIDIA Turing™\n**2019** NVIDIA Jetson Preferred Partner\n\n**Section 3: Proven Global Success**\nDeployed in 300+ projects across 30+ countries, trusted by leading enterprises worldwide](.axe-sales-kit-q2-2026-v1-1-1/slide-004.jpg)

## Slide 5

![The slide presents information about 'Pioneering Edge AI Systems,' subtitled 'Empowering Real-Time and Future-Proof Intelligence at Every Edge.'\n\n**Visual Elements & Labels:**\n*   Several black edge computing devices are displayed on the left and center.\n*   A server chassis and motherboard are shown on the right.\n*   **Component Icons/Labels:** 'SOC', 'MXM', and 'PEG' (appearing twice).\n*   **Vertical Text Overlays:** 'Intel Core' and 'NVIDIA Jetson'.\n*   **Logos:** 'NVIDIA Certified', 'NVIDIA Elite Partner', and 'intel prestige partner'.\n\n**Bottom Feature Boxes:**\n1.  **AI-Integrated Systems**\n    *   'GPU/accelerator-ready with pre-validated I/O for plug-and-play AI integration.'\n2.  **Comprehensive Portfolio**\n    *   'From low-power inference to large model inference— scalable across all edge tiers' (Text is partially cut off at the bottom).\n3.  **GenAI-Ready Design**\n    *   'Optimized for vision AI, predictive analytics, and on-device LLM inference' (Text is partially cut off at the bottom).](.axe-sales-kit-q2-2026-v1-1-1/slide-005.jpg)

## Slide 6

![**Title:** Define Your Edge AI Scale\n\n**Categories:**\n*   **Light-Weight AI:** City Vehicle Monitoring, Retail Crowd Tracking\n*   **Medium AI:** Manufacturing Scratch Detection, Logistics AMR\n*   **Heavy AI:** Humanoid & Legged Robotics, Large-Scale LLM Inference\n\n**Performance Scale (Left to Right):**\n*   (40 (TOPS)\n*   40-300\n*   300-1500\n*   1500-5000+ (TFLOPS)\n\n**Hardware Models (Top to Bottom, Left to Right):**\n*   10 TOPS (SOC): MXA-312M\n*   20-40 TOPS (SOC): DLAP-211-Orin\n*   40 TOPS (SOC): MXE-330\n*   275 TOPS (SOC): DLAP-411-Orin\n*   310 TOPS (MXM): DLAP-5200\n*   1457 TFLOPS (PEG): DLAP-8100\n*   2070 TFLOPS (SOC): DLAP-701\n*   3000+ TFLOPS (PEG): AXE Series\n\n**Footer:** 10](.axe-sales-kit-q2-2026-v1-1-1/slide-006.jpg)

## Slide 7

![The slide is titled **Decades Evolution in Industrial Servers**. It displays a progression of server hardware solutions labeled as follows:\n\n*   **AdvancedTCA Dual Xeon**\n*   **CompactPCI Dual Xeon**\n*   **Outdoor MEC**\n*   **OCP Telco**\n*   **Server & Server Board**\n\nAdditional logos and text on the slide include **NVIDIA Certified**, **AXE**, **ADLINK | Leading Edge AI Computing**, and **intel prestige partner**.](.axe-sales-kit-q2-2026-v1-1-1/slide-007.jpg)

## Slide 8

![**Title:** 3 Level Approach: Board level + CTOS\n\n**Level 1:**\n*   Level 1\n    Board Level\n*   TA MOQ=100\n\n**Level 6:**\n*   Level 6\n    Barebone\n*   TA MOQ=50\n*   L1 + chassis + PSU\n\n**Level 10:**\n*   Level 10\n    Configured\n    System\n*   TA MOQ=10\n*   L6 + CPU/DIMM\n*   *GPU, SSD, NIC\n\n**Footer:** 8](.axe-sales-kit-q2-2026-v1-1-1/slide-008.jpg)

## Slide 9

![**Title:** From Connection to Cognition\n**Subtitle:** Extending Edge AI capabilities into high-end, real-time vision processing and analytics\n\n**Visuals:**\n*   **Left:** An industrial robotic arm working on a production line.\n*   **Center:** A surgeon in an operating room using robotic assistance.\n*   **Right:** A hand interacting with a digital interface displaying bar graphs and data analytics.\n*   **Bottom Right:** Images of a server chassis and a motherboard.\n\n**Key Features (Bottom Left):**\n*   Edge-Native Design\n*   GPU Optimized Hub\n*   AI Ready Platform\n\n**Logos & Branding:**\n*   AXE\n*   NVIDIA Certified\n*   ADLINK | Leading Edge AI Computing](.axe-sales-kit-q2-2026-v1-1-1/slide-009.jpg)

## Slide 10

![This slide, titled **AXE Highlights**, features three blue information boxes describing key product features above an image of a server chassis and motherboard.\n\n**Left Box:**\n*   **Header:** Edge-Native Design\n*   **Description:** Compact, industrial-grade reliability with an extended product lifecycle\n\n**Center Box:**\n*   **Header:** GPU Optimized Hub\n*   **Description:** Precision-tuned power and thermal design for peak accelerator throughput\n\n**Right Box:**\n*   **Header:** AI Ready Platform\n*   **Description:** Certified integration for GPUs, accelerators, and frame grabbers\n\n**Bottom Section:**\n*   The text **AXE** appears above a motherboard.\n*   The **NVIDIA Certified** logo appears next to a server chassis.](.axe-sales-kit-q2-2026-v1-1-1/slide-010.jpg)

## Slide 11

![The slide is titled 'GenAI Ready Design' and features a digital, wireframe human head on the left. The right side lists three key points:\n\n**1 Proven Throughout**\nValidated tokens/sec performance on mainstream LLMS (Lama, Gemma, DeepSeek, Qwen)\n\n**2 Assured Responsiveness**\nNVIDIA-Certified platform minimizing delay for seamless GenAI interactions\n\n**3 Faster PoC**\nAccelerate model selection with suggested GPU options tailored to workflow performance and budget](.axe-sales-kit-q2-2026-v1-1-1/slide-011.jpg)

## Slide 12

![**Title:** Ready for Gen AI\n\n**Subtitle:** Mainstream-LLM benchmark results on AXE-7400SR+ 4 x NVIDIA RTX 6000 Ada GPU\n\n**List:**\n1. Standard: Latency need ( 1,000 ms, Token/s need ) 10\n2. Target 30B ~ 70B models\n\n**Table:**\n\n| | qwen3: 32b | llama4: 16x17b | gemma3: 27b | deepseek-r1: 32b | deepseek-r1: 70b |\n| :--- | :--- | :--- | :--- | :--- | :--- |\n| **Load Duration** | 6,359 ms | 8,251 ms | 36,220 ms | 7,741 ms | 93,495 ms |\n| **Prompt Eval Count** | 19 token(s) | 353 token(s) | 19 token(s) | 14 token(s) | 14 token(s) |\n| **Prompt Eval Duration ((1,000)** | 177 ms | 861 ms | 177 ms | 426 ms | 473 ms |\n| **Eval Count** | 1443 token(s) | 495 token(s) | 1114 token(s) | 967 token(s) | 915 token(s) |\n| **Eval Duration** | 40.593 s | 9.792 s | 29.802 s | 27.820 s | 50.852 s |\n| **Eval Rate ()10)** | 35.550 tokens/s | 50.550 tokens/s | 37.380 tokens/s | 34.760 tokens/s | 17.990 tokens/s |](.axe-sales-kit-q2-2026-v1-1-1/slide-012.jpg)

## Slide 13

![The slide is titled **'Scale in Place — with NVIDIA Confidence'** with the subtitle **'NVIDIA Certified & NVIDIA Qualified GPU'**.\n\nThe visual content displays a combination of hardware:\n*   **Left:** Two server chassis labeled **'Full AXE Series'**.\n*   **Center:** A large white plus sign (**+**).\n*   **Right:** Three graphics cards listed under the header **'NVIDIA Certified'** (next to the NVIDIA logo):\n    *   **'NVIDIA L4'** (top, gold-colored card)\n    *   **'NVIDIA RTX 6000 Ada'** (middle, black card with fan)\n    *   **'NVIDIA RTX PRO 6000 Blackwell'** (bottom, black card with fan), accompanied by a teal tag reading **'Under Planning'**.](.axe-sales-kit-q2-2026-v1-1-1/slide-013.jpg)

## Slide 14

![**GPU Compatibility**\n\n**Headers:** Model Series, GPU Memory, TDP, Form Factor, Thermal, Application\n\n**Row 1:**\n*   **Model Series:** NVIDIA RTX PRO 2000 Blackwell/ 4000 Blackwell SFF/ 4000 Blackwell\n*   **GPU Memory:** 16/ 24/ 24 GB\n*   **TDP:** 70/ 70/ 140 W\n*   **Form Factor:** SFF\n*   **Thermal:** Active\n*   **Application:** Space-constrained systems: Ideal for 2D/3D CAD, medical imaging, and multi-display signage.\n\n**Row 2:**\n*   **Model Series:** Intel® Arc™ Pro B50 Graphics\n*   **GPU Memory:** 16 GB\n*   **TDP:** 70 W\n*   **Form Factor:** SFF\n*   **Thermal:** Active\n\n**Row 3:**\n*   **Model Series:** NVIDIA RTX PRO 4500/ 5000 Blackwell\n*   **GPU Memory:** 32 GB 48/ 72GB\n*   **TDP:** 200/ 300 W\n*   **Form Factor:** Full Height\n*   **Thermal:** Active\n*   **Application:** Professional Design: Suitable for complex BIM modeling, high-quality 3D rendering, and entry-level AI inference.\n\n**Row 4:**\n*   **Model Series:** Intel® Arc™ Pro B60 Graphics\n*   **GPU Memory:** 24 GB\n*   **TDP:** 200 W\n*   **Form Factor:** SFF\n*   **Thermal:** Active\n\n**Row 5:**\n*   **Model Series:** NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation/ Workstation\n*   **GPU Memory:** 96 GB\n*   **TDP:** 300/ 600 W\n*   **Form Factor:** Full Height\n*   **Thermal:** Active/ Double-flow-through\n*   **Application:** AI Development & VFX: Massive VRAM for LLM fine-tuning, high-end video production, and heavy data science workloads.\n\n**Row 6 (Partially Visible):**\n*   **Model Series:** NVIDIA RTX PRO 6000 Blackwell\n*   **GPU Memory:** 96 GB\n*   **TDP:** 400-600 W\n*   **Form Factor:** Full Height\n*   **Thermal:** Passive\n*   **Application:** Scalable](.axe-sales-kit-q2-2026-v1-1-1/slide-014.jpg)

## Slide 15

![**AI Ready Platform**\n\n**Ultra-High Speed AOI**\n*   ISB-W890+ PCIe-CPL64V+ NVIDIA RTX PRO 6000 Blackwell+ Industrial Camera\n\n**AI RAN Ready Platform**\n*   MWC26\n*   SynaXG\n*   AXE-7220 + NVIDIA RTX PRO 6000 Blackwell+ SynaXG SV-1020 vDU\n\n**AI Enablement**\n\n**AI Dev. Stack**\n*   Ubuntu, Windows\n*   Edge AI Vision EVA\n*   Remote Device Management EdgeGO\n\n**Certification & Security**\n*   CE/FCC/UL/CCC/BSMI\n*   IEC 62443\n*   intel ai\n*   NVIDIA Certified\n\n**AI Computing Integration**\n*   NPU / GPU acceleration\n*   M.2 / MXM / PCIe expansion\n\n**Local AI Data Acceleration**\n*   AI SSD with aiDAPTIV+\n*   Offline LLM fine-tuning](.axe-sales-kit-q2-2026-v1-1-1/slide-015.jpg)

## Slide 16

![The slide is titled '**Edge AI PoV Studio**' with the subtitle '**Turning Proof-of-Value into the Right AI Platform Choice**.' It depicts a workflow diagram connecting hardware, applications, and validation.\n\nAt the bottom, a row of hardware devices is labeled '**Edge AI Platforms**' next to the '**NVIDIA RTX**' logo.\n\nOn the left, a panel titled '**Customer AI Applications**' features icons for '**LLM**' and '**VLM**,' followed by a section for '**AI Agents**' which includes an icon labeled '**EAAP**.'\n\nOn the right, a panel titled '**AI Performance Validation**' contains two charts. The upper chart, '**AI Workload Benchmark**,' displays bar graphs for '**Prompt Eval Rate**,' '**Load Duration**,' and '**Eval Rate**,' with a legend listing models such as '**Deepseek-1.1b**' and '**Llama4-16x17b**.' The lower chart, '**AI Performance Evaluation**,' shows line graphs for '**CPU usage**' and '**GPU usage**.'\n\nIn the center, a box labeled '**Customer Success**' displays images of robots and surgical arms. Above this, a box labeled '**Accelerate AI Adoption**' receives arrows from the left and right panels, with a downward arrow pointing to '**Customer Success**.' Arrows also flow from the bottom platforms up toward the applications and validation panels.](.axe-sales-kit-q2-2026-v1-1-1/slide-016.jpg)

## Slide 17

![**Title:** Product Portfolio\n\n**Section: Board**\n*Labels:* ODM, HYGON, AMD EPYC\n*   **IMB-M47:** Pro ATX Motherboard, Intel® Core™ 200S Series & 14/13/12th Gen Core™ (Q670 Chipset)\n*   **ISB-H34:** Pro ATX Motherboard, HYGON C86-4G 3490\n*   **ISB-SR10:** CEB, Intel® 4/5th Gen Xeon® Scalable\n*   **ISB-W890:** CEB, Intel® Xeon® 6 for Workstation\n*   **ISB-TR10:** CEB, AMD EPYC™ 9004/9005 Series\n\n**Section: 2U Server**\n*Labels:* ODM, AMD EPYC\n*   **AXE-5210RL:** Intel® Core™ i9\n*   **AXE-7220SR:** Intel® 4/5th Gen Xeon® Scalable\n*   **AXE-7220GW:** Intel® Xeon® 6 for Workstation\n*   **AXE-7220TR:** AMD EPYC™ 9004/9005 Series\n\n**Section: 4U Server**\n*Labels:* ODM, AMD EPYC\n*   **AXE-5420RL/HG:** Intel® Core™ i9\n*   **AXE-7400SR:** Intel® 4/5th Gen Xeon® Scalable\n*   **AXE-7420GWA:** Intel® Xeon® 6 for Workstation\n*   **AXE-7440GW:** Intel® Xeon® 6 for Workstation\n*   **AXE-7440TR:** AMD EPYC™ 9004/9005 Series](.axe-sales-kit-q2-2026-v1-1-1/slide-017.jpg)

## Slide 18

![The slide is titled 'New Products' and displays hardware organized into two main sections based on processor type:\n\n**Intel® Xeon® 600 Processor for Workstation Series**\n*   An 'intel prestige partner' logo is shown on the left.\n*   **ISB-W890**: Labeled 'CEB, Up to 4x FHFLDW GPU Gen5 x16'.\n*   **AXE-7440GW**: Labeled 'Up to 4x FHFLDW GPU Gen5 x16'.\n*   **AXE-7420GWA**: Labeled 'Up to 2x FHFLDW GPU Gen5 x16'.\n*   **AXE-7220GW**: Labeled 'Up to 2x FHFLDW GPU Gen5 x16'.\n\n**AMD EPYC™ 9004/9005 Series**\n*   An 'AMD EPYC' logo is shown on the left.\n*   **ISB-TR10**: Labeled 'CEB, 4x FHFLDW GPU Gen5 x16'.\n*   **AXE-7440TR**: Labeled 'Up to 4x FHFLDW GPU Gen5 x16'.\n*   **AXE-7220TR**: Labeled 'Up to 2x FHFLDW GPU Gen5 x16'.](.axe-sales-kit-q2-2026-v1-1-1/slide-018.jpg)

## Slide 19

![The slide displays hardware components on the left and a descriptive text block on the right.\n\n**Left Side (Hardware & Labels):**\n*   **AXE** (Logo)\n*   **intel ai**\n*   **ISB-W890** (Label for the motherboard)\n*   **AXE-7220GW** (Label for the black server chassis)\n*   **AXE-7420GWA** (Label for the white server chassis)\n*   **AXE-7440GW** (Label for the larger white server chassis)\n*   **intel prestige partner** (Logo)\n\n**Right Side (Text Block):**\n**Platforms: Intel® Xeon® 6 for Workstation**\n\n**Powerhouse CPU**\n*   Powered by Intel® Xeon® 600 Processor for Workstation\n*   Delivers up to 1TB DDR5 ECC memory for high throughput and reliability\n\n**Flexible Expansion**\n*   Supports up to 4x FHFLDW GPUs Gen5 x16\n*   Provides up to 12 expansion slots: 3x PCIe x16, 4x PCIe x8 (2 via MICO), 4x PCIe x4 (2 via MICO)\n\n**Engineered for High-Reliability Environments**\n*   Optimized for data-heavy workloads in manufacturing, medical imaging, robotics, and agentic AI applications](.axe-sales-kit-q2-2026-v1-1-1/slide-019.jpg)

## Slide 20

![The slide displays a blue background with network graphics on the left and a large text box on the right.\n\n**Visual Elements (Left Side):**\nThree hardware components are shown in ascending order of size:\n*   A motherboard labeled **ISB-TR10**.\n*   A smaller server chassis labeled **AXE-7220TR**.\n*   A larger server chassis labeled **AXE-7440TR**.\n*   The **AXE** logo appears above the middle chassis.\n*   The **AMD EPYC** logo appears below the largest chassis.\n\n**Text Content (Right Side):**\nThe text box contains the following information verbatim:\n\n**Workstation Platforms: AMD EPYC™ 9004/9005 CPUs**\n\n**High Parallel Performance**\n*   Powered by AMD EPYC™ 9004/9005 Series processors, a data center-class CPU\n*   Supports up to 1TB DDR5 ECC memory for high throughput and reliability\n\n**Flexible Expansion**\n*   Supports up to 4x FHFLDW GPUs (PCIe Gen5 x16)\n*   Provides up to 6 expansion slots, including 4x PCIe Gen5 x16 and 2x PCIe Gen5 x8\n\n**Optimized for Large-Scale LLM Workloads**\n*   Supports highly parallelized workloads such as large-scale LLM inference, databases, and virtualization platforms](.axe-sales-kit-q2-2026-v1-1-1/slide-020.jpg)

## Slide 21

![**Title:** Edge AI Server Motherboard\n\n**Logo:** AMD EPYC\n\n**Table Content:**\n\n*   **Model:** ISB-W890 | ISB-TR10\n*   **Support CPU:** Intel® Xeon® 600 Processor for Workstation | AMD EPYC™ 9004/9005 Series\n*   **Max GPU:** 4x FHFLDW GPU Gen5 x16 | 4x FHFLDW GPU Gen5 x16\n*   **Memory:** 8 DIMM DDR5 Slot | 8 DIMM DDR5 Slot\n*   **Storage:**\n    *   (Left): 8x 2.5' SATA III, 2x M.2 NVMe, 2x Slim SAS, 4x MCIO\n    *   (Right): 2x 2.5' SATA III, 6x 2.5' NVMe, 2x M.2 NVMe, 5x MCIO\n*   **Networking:** 1x GbE, 1x 2.5 GbE | 2x 1 GbE, 2x 10 GbE](.axe-sales-kit-q2-2026-v1-1-1/slide-021.jpg)

## Slide 22

![**Title:** Edge AI Server\n\n**Categories:**\n*   **4U Server:** AMD EPYC GenAI GPU Server, intel ai Automation GPU Server\n*   **2U Server:** AMD EPYC SFF GPU Server\n\n**Table Specifications:**\n\n| | AXE-7440TR | AXE-7440GW | AXE-7420GWA | AXE-7220TR | AXE-7220GW |\n| :--- | :--- | :--- | :--- | :--- | :--- |\n| **Model** | AXE-7440TR | AXE-7440GW | AXE-7420GWA | AXE-7220TR | AXE-7220GW |\n| **Support CPU** | AMD EPYC™ 9004/9005 Series | Intel® Xeon® 600 Processor for Workstation | Intel® Xeon® 600 Processor for Workstation | AMD EPYC™ 9004/9005 Series | Intel® Xeon® 600 Processor for Workstation |\n| **U / Depth** | 4U / 650 mm | 4U / 650mm | 4U / 550mm | 2U / 450 mm | 2U / 450 mm |\n| **Max GPU** | 4x FHFLDW GPU Gen5 x16 | 4x FHFLDW GPU Gen5 x16 | 2x FHFLDW GPU Gen5 x16 | 2x FHFLDW GPU Gen5 x16 | 2x FHFLDW GPU Gen5 x16 |\n| **Storage** | 8x 2.5” SATA III 2x M.2 NVMe | 8x 2.5” SATA III 2x M.2 NVMe | 8x 2.5” SATA III 2x M.2 NVMe | 4x 2.5” SATA III 2x M.2 NVMe | 4x 2.5” SATA III 2x M.2 NVMe |\n| **Redundant Power** | 2x 1300W/1600W/2000 W, 1+1 | 2x 800W/1300W/1600W , 1+1 | 1x 1650W | 2x 800W/1300W/1600W , 1+1 | 2x 800W/1300W/1600W , 1+1 |](.axe-sales-kit-q2-2026-v1-1-1/slide-022.jpg)

## Slide 23

![**Title:** ISB-W890 (CEB Server Motherboard)\n\n**Key Specifications:**\n*   CEB form factor: 12' x 10.5' (305 mm x 267 mm)\n*   Intel® Xeon® 600 Processor for Workstation\n*   4-channel DDR5 memory, 8× DDR5 RDIMM slots, 6400 MT/s\n*   High PCIe flexibility: 7x PCIe slots, including 3x PCIe x16 slots\n*   2x Slim SAS slots, 2x M.2 NVMe, 8x SATA III slots\n\n**Rear I/O Panel Labels:**\n*   DP\n*   Clear CMOS Reset Hole\n*   COM\n*   VGA\n*   1Gb LAN\n*   USB 3.2 Gen 2\n*   2.5Gb AUDIO LAN\n*   USB 3.2 Type C\n\n**Footer:**\nADLINK | Leading Edge AI Computing](.axe-sales-kit-q2-2026-v1-1-1/slide-023.jpg)

## Slide 24

![**Title:** ISB-TR10 (CEB Server Motherboard)\n\n**Specifications List:**\n*   CEB form factor: 12” x 10.5” (305 mm x 267 mm)\n*   Single socket SP5 (LGA 6096), AMD EPYC™ 9005/9004 Series processors\n*   8x DIMM slots (1 DPC), DDR5 RDIMM 6400 MT/s\n*   4x PCIe 5.0 x16, 2x PCIe 5.0 x8, 1x MCIO (PCIe 5.0 x8 or 8x SATA 6 Gb/s), 4x MCIO (PCIe 5.0 x8)\n*   2x M.2 NVMe, 8x SATA III slots\n\n**Image Labels (Rear I/O Panel):**\n*   IPMI LAN Header\n*   VGA Port\n*   1G LAN RJ-45\n*   UID Switch\n*   USB 3.2 Gen 1 10G LAN RJ-45\n\n**Footer/Logo:**\n*   ADLINK\n*   K\n*   Leading Edge AI Computing](.axe-sales-kit-q2-2026-v1-1-1/slide-024.jpg)

## Slide 25

![**AXE-7440GW (GenAI GPU Server)**\n\n*   Intel® Xeon® 600 Processor for Workstation\n*   8 expansion slots: 4x PCIe x16 (1 via MICO), 2x PCIe x8, 2x PCIe x4\n*   Up to 4x FHFLDW GPUs Gen5 x16\n*   8x 2.5”/3.5” SATA HDDs\n*   3x internal fan modules, 2x optional external fan modules\n\n**Image Labels:**\n*   2x USB 2.0\n*   6x HDD\n*   2.5Gb LAN\n*   1Gb LAN\n*   USB 3.2 Type C\n*   USB 3.2 Gen 2\n*   Clear CMOS reset hole\n\n**Footer:**\nADLINK | Leading Edge AI Computing](.axe-sales-kit-q2-2026-v1-1-1/slide-025.jpg)

## Slide 26

![The slide presents the specifications and visuals for the **AXE-7420GWA (Automation GPU Server)**.\n\n**Key Specifications (Left Column):**\n*   Intel® Xeon® 600 Processor for Workstation\n*   11-slot expansion: 3x PCIe x16, 4x PCIe x8 (2 via MICO), 4x PCIe x4 (2 via MICO)\n*   Up to 2x FHFLDW GPUs Gen5 x16\n*   Short-depth chassis (550 mm) for maximum rack space efficiency\n*   Up to 8x 2.5'/3.5' SATA HDDs (or 1x CD-ROM), 2x SlimSAS SSDs\n\n**Visuals (Right Side):**\nThree images of the server chassis are displayed:\n1.  A vertical side view.\n2.  A horizontal front/rear view.\n3.  A horizontal rear view highlighting connectivity ports.\n\n**Connectivity Labels (Pointing to rear ports):**\n*   2.5Gb LAN\n*   1Gb LAN\n*   USB 3.2 Type C\n*   USB 3.2 Gen 2\n*   Clear CMOS Reset Hole\n\n**Footer:**\nThe bottom left corner features the logo 'ADLINK' with the tagline 'Leading Edge AI Computing'.](.axe-sales-kit-q2-2026-v1-1-1/slide-026.jpg)

## Slide 27

![The slide displays specifications for a server unit against a blue background.\n\n**Header:**\nAXE-7220GW (SFF GPU Server)\n\n**Top Right:**\nStylized logo text: AXE\nStatus text: preliminary\n\n**Bullet Points (Left):**\n*   Intel® Xeon® 600 Processor for Workstation\n*   2U, 450 mm short-depth chassis for space-efficient deployment\n*   8x DDR5 6400 MT/s RDIMM slots\n*   Up to 2x FHFLDW GPUs (Gen5 x16)\n*   4x fan modules and SSD drive bays\n\n**Labels (Right, pointing to the server image):**\n*   4x SSD\n*   4x Fan Module\n\n**Footer (Bottom Left):**\nADLIN\nK\nLeading Edge AI Computing](.axe-sales-kit-q2-2026-v1-1-1/slide-027.jpg)

## Slide 28

![AXE-7440TR (GenAI GPU Server)\npreliminary\n• AMD EPYC™ 9005/9004 Series processors\n• Optional PCIe slots: Up to 4x PCIe Gen5 x16, 3x PCIe Gen5 x8 via MCIO\n• Up to 4x FHFLDW GPUs Gen5 x16\n• 8x 3.5'/2.5' SATA SSDs or HDDs\n• 3x internal fan modules, 2x optional external fan modules\n2x USB 2.0\n6x HDD\nADLIN K | Leading Edge AI Computing](.axe-sales-kit-q2-2026-v1-1-1/slide-028.jpg)

## Slide 29

![Slide 29](.axe-sales-kit-q2-2026-v1-1-1/slide-029.jpg)

## Slide 30

![This slide features a background image of an industrial welding setup with robotic arms. The text overlay on the left reads:\n\n**Welding System Robotics**\n\n**Customer Demands**\n*   64core+ CPU, 256GB DDR\n*   2x RTX-5070\n*   Local RMA and service\n\n**Values to Customers**\n*   AMD EPYC 64C/96C\n*   Up to 3x NVIDIA RTX 5070/5080\n*   CPU liquid cooling option\n\nIn the bottom left corner, there is an image of a server labeled **AXE-7440TR** next to a logo reading **USBiGiGE VISION VISION**.](.axe-sales-kit-q2-2026-v1-1-1/slide-030.jpg)

## Slide 31

![The slide presents information about 'Micro LED AOI' (Automated Optical Inspection) set against a background image of an industrial inspection arm hovering over a circular wafer.\n\n**Text Content:**\n\n**Micro LED AOI**\n\n**Customer Demands**\n*   Real-time AOI: ≤10 ms end-to-end\n*   Mixed I/O & growing model complexity strain compute\n*   Consolidate 3 IPC into one Server\n\n**Values to Customers**\n*   Space, function consolidation\n*   NVIDIA-Certified System with NVIDIA RTX 6000 Ada GPU\n*   CoaXPress\n*   intel\n\n**Hardware Labels:**\n*   AXE-7440GW\n*   Euresys](.axe-sales-kit-q2-2026-v1-1-1/slide-031.jpg)

## Slide 32

![This slide presents information about a 'KMS' (Knowledge Management System).\n\n**Left Column:**\n*   **Header:** KMS (Knowledge Management System)\n*   **Customer Demands:**\n    *   On-premise\n    *   Limited power, cooling, space\n    *   Up to 4x double deck GPU\n    *   (20sec latency per question\n    *   5% lower hallucination\n*   **Values to Customers:**\n    *   Short depth 4U\n    *   4x RTX6000 Ada, Pre-certified\n    *   Optimized for LLaMA4/70B\n*   **Image:** A server chassis labeled 'AXE-7440GW/' (with a partially obscured line below it starting with 'AXE-7440').\n\n**Right Column (Illustration):**\n*   **Interface Tabs:** SOP, QC Records, BOM\n*   **Section Title:** Supplier Capability Q&A\n*   **Chat Box:**\n    *   **Q:** What is your SMT capacity per day?\n    *   **A:** Our SMT line A-B supports 120k CPH, with a monthly capacity 3,2 M boards. We are ISO 13485 certified.\n*   **Buttons:** Export to, Export to Excel, Cite\n*   **Visual:** Two illustrated figures looking at the interface.](.axe-sales-kit-q2-2026-v1-1-1/slide-032.jpg)

## Slide 33

![Contact Information\n\nProduct Sales\nManager\nRay Wang\nray.wang@adlinktech.com\n+866-3-2165088 ext.7953\n\nMarketing\nPartner\nVictor Chan\nvictor.chan@adlinktech.co\nm\n+866-3-2165088 ext.8413](.axe-sales-kit-q2-2026-v1-1-1/slide-033.jpg)

[🔗 Link to the original document](.axe-sales-kit-q2-2026-v1-1-1/axe-sales-kit-q2-2026-v1-1-1.pptx)