Intel® Deep Link Technology: Transforming CPU-GPU Interaction for Enhanced Performance
Intel® Deep Link Technology is a game-changing innovation that transforms the interaction between a systems's Central Processing Unit (CPU) and Graphics Processing Units (GPUs). Available on several Intel-based modules, Deep Link combines the computing power of a discrete GPU (dGPU) with that of a powerful integrated GPU (iGPU), simplifying code development and reducing overall effort. With a high degree of code re-use, Deep Link enables the construction of complex workloads and pipelines, leading to significant gains in performance and efficiency, and boosting the functional capabilities of a given application.
The partitioning of computational elements into logical segments allows Deep Link GPUs to equitably share the workload, with each dGPU and iGPU working independently and concurrently with the other. Intel® Iris® Xe can be used together to split tasks, reducing workload completion time by up to half.
By combining the computing power of multiple GPUs with similar characteristics, Deep Link enables Intel computing components to work together at an unprecedented level of speed and efficiency, with little additional overhead when partitioning tasks between the two graphics processors.
ADLINK solutions supporting Intel® Deep Link Technology:
MXM-AXe, with discrete GPU @Intel® Arc™, supports Deep Link Technology. It offers hardware ray tracing, AI acceleration, and 4x display, delivering precision graphics for time-sensitive edge applications in healthcare, transportation, gaming, and more. Click here to obtain more information.