Hengqin, Guangdong, July 31, 2026- Lanqi Technology announced today that it will be the first to pilot produce CXL in the industry ® 3.2 Memory Expansion Controller (MXC) chip. This product is designed to meet the memory expansion and resource pooling needs of scenarios such as artificial intelligence, cloud computing, and data centers. It aims to assist customers in building larger capacity, higher utilization, and more flexible memory infrastructure through CXL technology.
Lanqi Technology's CXL 3.2 memory expansion controller supports CXL.mem and CXL.io protocols of CXL Type 3 specification, based on PCIe ® 6. x and CXL 3.2 protocol design, supporting a maximum data transfer rate of 64 GT/s. The product integrates dual DDR5 memory controllers, which can support up to 8000 MT/s of DDR5 memory. By converting CXL memory requests on the host side into DDR access commands in real time, efficient data exchange between the host and backend memory is achieved, thus breaking through the memory capacity limitations of traditional servers.
In addition, the chip integrates PCIe 6. x PHY, DDR5 PHY, dual RISC-V processor subsystems, and rich management interfaces, which can support the development of various forms of CXL memory expansion products such as PCIe standard card (AIC) and EDSFF, providing key support for new memory applications such as memory expansion, memory pooling, and layered memory architecture.
广东横琴，2026年7月31日 -- 澜起科技今日宣布，率先在业界试产CXL® 3.2内存扩展控制器 (MXC) 芯片。该产品面向人工智能、云计算、数据中心等场景的内存扩展与资源池化需求，旨在通过CXL技术助力客户构建更大容量、更高利用率、更灵活的内存基础设施。
澜起科技的CXL 3.2内存扩展控制器支持CXL Type 3规范的CXL.mem和CXL.io协议，基于PCIe® 6.x和CXL 3.2协议设计，最高支持64 GT/s数据传输速率。产品集成双DDR5内存控制器，可支持最高8000 MT/s的DDR5内存，通过将主机侧CXL内存请求实时转换为DDR访问命令，实现主机与后端内存间的高效数据交互，从而突破传统服务器的内存容量限制。
此外，该芯片集成了PCIe 6.x PHY、DDR5 PHY、双 RISC-V处理器子系统及丰富的管理接口，可支持开发PCIe标准插卡 (AIC)、EDSFF等多种形态的CXL内存扩展产品，为内存扩展、内存池化和分层内存架构等新型内存应用提供关键支撑。
Mr. Stephen Tai, President of Lanqi Technology, said, "Being the first in the industry to pilot produce CXL 3.2 MXC chips is an important milestone for Lanqi Technology to promote CXL technology innovation and industrial development. In the face of the continuously growing memory demand in the AI era, we provide our customers with memory expansion solutions that combine high bandwidth, large capacity, and high reliability, accelerating the large-scale implementation of CXL technology in data centers and AI infrastructure. ”
In addition to chips, Lanqi Technology also provides complete software design kits (SDKs) and analysis testing tools to help customers quickly complete product development, compatibility verification, and mass production import. At present, the Lanqi Technology CXL 3.2 MXC has been successfully imported into the next generation CXL products of major memory module manufacturers in the industry, such as Samsung and SK Hynix, and has completed preliminary verification. This solution supports Intel ® Xeon ®、 Mainstream server platforms such as AMD EPYCTM achieve extensive interoperability and outstanding performance.
Customer and partner evaluations and feedback:
澜起科技总裁Stephen Tai先生表示：“率先在业界试产CXL 3.2 MXC芯片，是澜起科技推动CXL技术创新和产业化发展的重要里程碑。面对AI时代持续增长的内存需求，我们为客户提供兼具高带宽、大容量和高可靠性的内存扩展解决方案，加速CXL技术在数据中心和AI基础设施中的规模化落地。”
除芯片外，澜起科技还同步提供完整的软件设计套件 (SDK) 及分析测试工具，助力客户快速完成产品开发、兼容性验证及量产导入。目前，澜起科技CXL 3.2 MXC已成功导入三星、SK海力士等业界主要内存模组厂商的下一代CXL产品中，并已完成初步验证。该解决方案支持Intel® Xeon®、AMD EPYCTM等主流服务器平台，实现广泛的互操作性和卓越性能。
客户与合作伙伴的评价与反馈：
Jangseok Choi, Vice President of Memory Product Planning at Samsung Electronics, stated that the combination of CXL 3.2 MXC and DDR5 memory technology provides data center customers with a solution that balances performance and capacity. The cooperation results of both parties have verified the application value of CXL technology in improving system memory capacity and resource utilization efficiency, and also provided more choices for the construction of next-generation AI infrastructure. ”
Uksong Kang, the head of planning and standards for SK Hynix's next-generation products, believes that "with the rapid popularity of large models and generative AI applications, memory has become a key factor in system performance. Lanqi Technology's MXC products have shown promising application prospects in supporting high-capacity memory expansion and resource sharing. We look forward to further cooperation in ecological verification and product innovation in the future. ”
Dr. Debendra Das Sharma, Senior Researcher and Chief I/O Architect at Intel, said, "Artificial intelligence, next-generation cloud systems, and today's industry economic landscape are reshaping server memory architecture and driving demand for CXL. The Lanqi Technology CXL 3.2 MXC showcases how the increasingly mature CXL ecosystem provides protocol support, compatibility, and performance, helping to unleash the memory expansion capabilities of the next generation Intel Xeon platform. ”
三星电子内存产品规划副总裁Jangseok Choi表示：“CXL 3.2 MXC与DDR5内存技术的结合，为数据中心客户提供了兼顾性能与容量的解决方案。双方合作成果验证了CXL技术在提升系统内存容量和资源利用效率方面的应用价值，也为下一代AI基础设施建设提供了更多选择。”
SK海力士下一代产品规划与标准负责人Uksong Kang认为：“随着大模型和生成式AI应用的快速普及，内存已成为系统性能的关键因素。澜起科技MXC产品在支持高容量内存扩展和资源共享方面展现出良好的应用前景，我们期待未来在生态验证和产品创新方面展开更深入的合作。”
英特尔高级研究员兼首席I/O架构师Debendra Das Sharma博士表示：“人工智能、下一代云系统以及当今的行业经济格局正在重塑服务器内存架构，并推动对CXL的需求。澜起科技CXL 3.2 MXC展示了日益成熟的CXL生态系统如何提供协议支持、兼容性和性能，助力释放下一代英特尔至强平台的内存扩展能力。”
Amit Goel, Vice President of Computing and Enterprise AI Platform Solutions Engineering at AMD, pointed out that "artificial intelligence and data center workloads are driving the continuous evolution of system architecture, and memory has become a key factor affecting performance and scalability. We look forward to continuing to collaborate with Lanqi Technology in the CXL technology field to lay the foundation for building a more flexible and efficient memory architecture. ”
Market Activity Preview
Lanqi Technology will showcase its CXL 3.2 MXC products at the CXL DevCon and FMS exhibitions in the United States. We sincerely invite customers, friends, industry colleagues, and partners to visit and exchange ideas, and explore industry development trends and cooperation opportunities together!
AMD计算与企业AI平台解决方案工程副总裁Amit Goel指出：“人工智能和数据中心工作负载正推动系统架构持续演进，内存已成为影响性能与扩展能力的关键因素。我们期待与澜起科技在CXL技术领域持续合作，为构建更加灵活高效的内存架构奠定基础。”
市场活动预告
澜起科技将携CXL 3.2 MXC产品亮相美国CXL DevCon和FMS展会，诚邀各位客户朋友、业界同仁及合作伙伴莅临参观交流，共同探讨行业发展趋势与合作机遇！
CXL Mini DevCon
Date: August 3, 2026
Location: Santa Clara Marriott, USA
CXL Mini DevCon
时间：2026年8月3日
地点：美国圣克拉拉万豪酒店 (Santa Clara Marriott)
FMS
Date: August 4-6, 2026
Location: Santa Clara Convention Center Booth # 845
FMS
时间：2026年8月4-6日
地点：美国圣克拉拉会展中心 (Santa Clara Convention Center) 展位号#845
(Using AI translation)
（使用AI翻译）
