skip to main content

ThinkEdge SE100 with NVIDIA RTX 2000E Ada GPU Delivers Exceptional MLPerf 5.1 Results for Edge AI in Retail

Article

Home
Top
Published
29 Oct 2025
Form Number
LP2325
PDF size
5 pages, 221 KB

Introduction

Imagine walking into a modern retail store where AI seamlessly enhances the shopping experience. Smart cameras count foot traffic, monitor shelf inventory, and analyze customer behavior—all in real time. Behind the scenes, this is powered by Lenovo’sThinkEdgeSE100 equipped with the NVIDIA RTX 2000E Ada GPU, delivering industry-leading performance as demonstrated in the latest MLPerf v5.1 Inference benchmark.

ThinkEdge SE100
Figure 1. ThinkEdge SE100 edge server

Real-World Retail Use Case: AI at the Edge

Retailers are increasingly deploying edge AI to optimize operations and improve customer engagement. Use cases include:

  • People counting for traffic analysis and staffing optimization
  • Shelf inventory monitoring to reduce out-of-stock incidents
  • Behavior analysis to understand customer preferences and improve layout

These tasks rely on computer vision models like ResNet for classification and RetinaNet for object detection—both included in the MLPerf benchmark suite.

MLPerf v5.1: A Trusted Industry Standard

MLPerf is the gold standard for evaluating machine learning performance across hardware platforms. The Inference Edge suite measures latency and throughput under real-world conditions, ensuring results are:

  • Transparent and reproducible
  • Representative of diverse workloads
  • Relevant to edge deployments

SE100 vs. Jetson AGX Orin: Performance That Matters

In MLPerf v5.1, the SE100 with RTX 2000E Ada GPU outperformed the Jetson AGX Orin 64GB across key models, as shown in the table below.

Table 1. Key benchmark results
Model Task SE100 Latency ↓ SE100 Throughput ↑
3D-UNet Medical Imaging 1.2× faster 1.2× higher
ResNet Image Classification 2.22× faster 1.45× higher
RetinaNet Object Detection 1.93× faster 1.2× higher
Stable Diffusion XL Generative AI 1.53× faster 1.48× higher

Despite having a lower theoretical AI TOPS (142 INT8 Sparse TOPS vs. 275), the SE100’s real-world performance proves that MLPerf results matter more than peak specs.

Why SE100 Is Ideal for Edge AI:

  • Compact and energy-efficient for deployment in retail environments
  • Low latency ensures real-time responsiveness
  • High throughput supports multiple concurrent AI tasks
  • Powered by Intel Ultra 7 255H CPU and RTX 2000E Ada GPU

Conclusion

The SE100 is a powerful edge AI solution that delivers measurable performance gains in real-world scenarios. Whether you're optimizing retail operations, deploying smart surveillance, or driving automation projects, Lenovo’s SE100 stands out as a reliable, high-performance platform validated by MLPerf v5.1.

Resources

For more information, see these resources:

Author

Serban Militaru is a Senior Product Marketing Manager for the Tower, Edge, and Multi-Node Servers business at Lenovo. He is dedicated to understanding customer needs and leveraging technology to address them effectively. Serban is based in Romania and graduated from the Bucharest Academy of Economic Studies.

Related product families

Product families related to this document are the following:

Trademarks

Lenovo and the Lenovo logo are trademarks or registered trademarks of Lenovo in the United States, other countries, or both. A current list of Lenovo trademarks is available on the Web at https://www.lenovo.com/us/en/legal/copytrade/.

The following terms are trademarks of Lenovo in the United States, other countries, or both:
Lenovo®
ThinkEdge®
ThinkSystem®

The following terms are trademarks of other companies:

Intel®, Intel Arc®, and Xeon® are trademarks of Intel Corporation or its subsidiaries.

Arc® is a trademark of Microsoft Corporation in the United States, other countries, or both.

Other company, product, or service names may be trademarks or service marks of others.