MAXSUN Intel Arc Pro at Lenovo 2026 Partner Conference

Shenzhen, April 17, 2026 — Under the theme “United in Purpose, Shared Success, A New Chapter Ahead,” Lenovo’s 2026 Government & Enterprise Partner Conference was held in Shenzhen, bringing together industry leaders to explore the next stage of intelligent transformation. The event focused on the implementation of the “AI+” initiative and the core needs of government and enterprise customers, while unveiling a broad portfolio of AI products and solutions spanning device, edge, and cloud scenarios.

At this high-profile industry gathering, AI computing products powered by the MAXSUN Intel Arc Pro Series GPUs were showcased, injecting powerful AI computing capabilities into Lenovo’s government and enterprise ecosystem. Their presence highlighted MAXSUN’s strong technological foundation and continued innovation in AI deployment.

As enterprise digital transformation accelerates, customers are moving beyond the simple “+AI” stage of adding AI features and into a new phase of “AI+” value delivery, where AI becomes a core driver of productivity and business innovation. With a forward-looking product strategy, MAXSUN has built a full-spectrum AI portfolio covering single-GPU to multi-GPU configurations, and workstations to servers, enabling flexible deployment for organizations of different sizes and industries.

As a close strategic partner of Intel, MAXSUN has introduced the MAXSUN Intel Arc Pro Series with a design philosophy built for the future of AI. Backed by strong hardware specifications and a robust software ecosystem, the series delivers breakthrough solutions for AI developers, multi-GPU collaborative workloads, and productivity users handling intensive professional applications.

（）

One of its standout advantages lies in AI-powered video encoding, decoding, and super resolution technologies. For example, taking a native 1080p at 30 FPS video stream at 8 Mbps as a reference, the sending side can use a MAXSUN Intel Arc Pro GPU to downsample the stream to 540p while maintaining 30 FPS, reducing the bitrate from 8 Mbps to 4 Mbps. This effectively cuts transmission bandwidth requirements — and corresponding costs — by half. On the receiving side, the GPU leverages AI Super Resolution to restore the stream to 1080p in real time, while AI Frame Rate Conversion (AIFRC) boosts the frame rate from 30 FPS to 60 FPS. The final result is a near-lossless 1080p at 60 FPS output, achieving the remarkable benefit of half the bitrate, half the bandwidth, and double the frame rate, significantly lowering transmission costs while preserving a smooth and clear viewing experience.

Among the products highlighted at the event, the MAXSUN Intel Arc Pro B70 Series stood out by redefining AI computing and professional visual workloads with 32GB of onboard memory on a single card.

The MAXSUN Intel Arc Pro B70 32G Turbo features 32GB of memory, 32 Xe cores, 608GB/s memory bandwidth, and up to 367 TOPS (INT8) of peak AI performance. It supports up to 256K context length, enabling fast and efficient AI inference, video encoding and decoding, and graphics rendering for creative professionals and intelligent applications.

The MAXSUN Intel Arc Pro B70 Series also offers native support for the PyTorch framework and has obtained ISV certifications, ensuring stable and efficient performance with mainstream professional software. In addition, it supports vLLM-based Multi-Arc, delivering an out-of-the-box optimized experience for large language model deployment. With Docker containerized deployment, applications, runtime environments, configurations, and dependencies can be packaged into lightweight, portable containers, making one-click deployment and easy operation a reality.

Also attracting strong attention was the MAXSUN Intel Arc Pro B60 Series. In particular, the MAXSUN Intel Arc Pro B60 Dual 48G Turbo introduces a breakthrough solution for AI inference and high-performance computing with its innovative dual-GPU design and 48GB of GDDR6 memory. Supporting a PCIe 5.0 x8+x8 interface, the card can run efficiently on consumer platforms as long as the motherboard supports bifurcation of a PCIe x16 slot into x8+x8, significantly lowering the barrier to local deployment of large AI models.

AI inference, quantization, and multimodal processing all place heavy demands on GPU memory. As system memory prices continue to rise, achieving higher performance often means higher build costs. Dynamic quantization is a key technology for running large models within limited VRAM, but traditional methods typically require model weights to be loaded into system memory before conversion, resulting in higher memory usage and slower performance. By optimizing environment variable settings, the MAXSUN Intel Arc Pro B60 Series can significantly reduce memory consumption during dynamic quantization, helping users maintain strong runtime performance while lowering overall system memory costs.

Looking ahead, MAXSUN will continue to deepen its strategic partnership with Intel, stay aligned with global AI technology trends, and further explore the deployment needs of the AI era. With more innovative hardware products and more complete solution offerings, MAXSUN is committed to empowering digital transformation across industries worldwide.