ADLINK Technology DLAP Supreme AI platform series

ADLINK Technology has launched its DLAP Supreme edge generative AI platform series. By integrating Phison’s innovative aiDAPTIV+ AI solution, this series overcomes memory limitations in edge generative AI applications, thereby enhancing AI computing capabilities on edge devices.

As generative AI continues to penetrate various industries, many edge devices encounter performance bottlenecks due to insufficient DRAM capacity when executing large language models, affecting model operations and even causing issues such as inadequate token length. The DLAP Supreme series overcomes these limitations and enhances computing performance. It also supports edge devices in conducting generative language model training, enabling them with AI model training capabilities and improving their autonomous learning and adaptability.

For instance, when using the DLAP-411 Orin Supreme equipped with an NVIDIA Jetson AGX Orin 32GB configuration alongside the Gemma 27B model, inference TTFT improved by 8 times, and token length increased by 4 times, according to the company. It can also execute training tasks for models that would typically require H100/A100 80G Tensor Core GPUs, thereby enhancing the performance of generative AI model operations.

The launch of the DLAP Supreme series opens up new opportunities for edge generative AI applications while helping various industries overcome memory limitations. In addition to optimisation for AGX Orin, the aiDAPTIV+ technology can also extend the memory of entry-level Orin Nano, NX and Super modules, enabling more devices to perform efficient AI computing. ADLINK plans to continue launching Orin Nano Super and NX Super versions to further expand the application scope of the DLAP Supreme series, meeting diverse AI computing needs.