From July 8th to 10th, the Electronica China 2024 was held grandly at the Shanghai New International Expo Center. The exhibition brought together thousands of exhibitors from all over the world, covering cutting-edge fields such as semiconductors, sensors, connectors and wire harnesses, intelligent manufacturing, and automotive electronics. Luxshare-Tech, a subsidiary of Luxshare, as a core infrastructure solution provider in the ICT field, made a stunning appearance with the most cutting-edge "AI data center solution".

While AI intelligent computing blue ocean surges, Luxshare-Tech makes comprehensive deployment

With the explosive growth of generative artificial intelligence (AI) technology, the demand for AI computing power is surging like a tidal wave, driving the global AI server market into an unprecedented expansion phase. According to authoritative data from Trendforce, global AI server shipments in 2023 have exceeded 1.205 million units, and are expected to surge to 2.369 million units by 2026, with a compound annual growth rate of up to 25%. With its deep accumulation in communication and data center technologies, Luxshare-Tech is riding the AI wave, making comprehensive move in the AI server market, and leading the industry forward.

Breakthroughs in High-speed interconnection technology define new standards for future data transmission

At the exhibition, Luxshare-Tech showcased a series of innovative high-speed interconnection products and technical solutions carefully built around the stringent needs of AI intelligent computing centers. From cable backplanes with custom interfaces and Chip2IO efficient connection solutions, to NPC connector modules, and the trend-setting AI cluster optical interconnection solutions, Luxshare-Tech impressed the audience with an application rate of 224Gbps, and proactively launched the pre-research work of 448Gbps technology. What is particularly striking is that Luxshare Technology has achieved large-scale commercial use of 112Gbps high-speed interconnection products, and 224Gbps products have also entered the small-scale mass production stage. Luxshare-Tech has also partnered with international leading chip manufacturers to outline a grand blueprint for the next generation of high-speed interconnection standards such as 1.6T and 3.2T.

Luxshare-Tech has insisted on continuous investment for a long time in silicon photonic solutions for data centers, developed 400G single-mode and 800G single-mode high-speed silicon photonic modules, and established a complete silicon photonic solution design and silicon photonic engine process manufacturing system. Based on rich experience in silicon photonics design and driven by the needs of leading customers, Luxshare-Tech is set to launch the 1.6T OSFP DR8 for 200G SerDes, providing strong support for the rapidly growing demand for AI optical interconnection.

"Light active" technology innovation, a new leap in intelligent computing energy efficiency

Facing the energy efficiency challenges of high density and high efficiency in intelligent computing centers, Luxshare-Tech has launched an energy efficiency revolution centered around its "light active" technology. Relying on the innovative integration of high-speed bare wire database and optoelectronic conversion technology, this technology has achieved a significant improvement in the signal quality of high-speed transmission links through cutting-edge means such as beam shaping, time reconstruction, and direct chip driving. At the same time, it has greatly reduced power consumption, extended transmission distance, and reduced cost expenditure, providing solid technical support for the green and sustainable development of intelligent computing centers.

Extreme heat dissipation, green energy saving

In the key field of thermal management, Luxshare-Tech has also demonstrated its strong innovation strength and forward-looking vision. At the exhibition, the company grandly launched the most cutting-edge stacked CDU cabinets and immersion liquid cooling tanks, providing revolutionary cooling solutions for high-density, high-power AI data centers.

The immersion liquid cooling tank from Luxshare-Tech stands out with its unique features. It effectively separates high-heat components from environmental heat dissipation, utilizing dedicated cooling channels specifically for CPU, GPU, and other core heat-generating elements. Compared to traditional immersion cooling solutions, this design achieves precise heat control and rapid heat transfer, effectively preventing localized heat from affecting the entire system. At the same time, independent cooling channels ensure that the coolant can continuously and stably provide heat dissipation support for high-heat components, and can maintain the low temperature and stable operation of the system even under long-term high-load operation, thereby extending the service life of the hardware equipment and improving the reliability of the overall system.

Customized services enable personalized development of AI intelligent computing centers

Luxshare-Tech deeply understands that in the era of rapid development of AI intelligent computing centers, customers' demand for customized and integrated solutions is becoming increasingly urgent. Therefore, the company has built a comprehensive product lines and R&D system closely around the diversified needs of next-generation servers, switches, entire cabinets, and AI intelligent computing clusters. Through in-depth insights into customer needs, precise customization of product design, and continuous optimization of performance and reliability, Luxshare-Tech provides customers with end-to-end high-speed interconnection, thermal management, and power management solutions covering multiple nodes at the component level, device level, cabinet level, and cluster level. What is more worth mentioning is that Luxshare-Tech also actively establishes close collaborative R&D relationships with customers, jointly exploring new technologies and applications to meet rapidly changing market demands. This collaborative approach injects strong momentum into the burgeoning development of AI large models.