In the case of Topaz, this tool is able to harness Intel XMX (matrix extensions) on board the Arc GPU, and also other GPU processing resources on board the CPU’s Iris Xe integrated graphics engine, to accelerate an AI denoise and upscaling algorithm with a short, low quality video clip I captured on my phone. Let’s look at a couple examples of Hyper Encode and Hyper Compute in action, as well as just standard media encoding acceleration on the Arc GPU alone.Ībobe Premiere Pro Video Rendering On Intel Arc A370M GPU Alone Dave AltavillaĪll of the above benchmarks represent some form of video encoding, image quality enhancement, or post-production and special effects processing. Deep Link consists of three primary features: Dynamic Power Share, which dynamically balances power delivery between the CPU and GPU, depending on workload requirements in real-time, Hyper Compute, which utilizes both discrete Arc GPU resources as well as the integrated Iris Xe Graphics engine on board Intel 12th Gen CPUs, for accelerating compute workloads like AI-driven algorithms in various applications, and Hyper Encode, which also leverages the CPU and GPUs for multimedia transcoding. Intel calls its conjoining and balancing of CPU and GPU resources Deep Link technology.
In terms of performance, a unique feature that’s interesting to look at with Intel’s new Arc mobile graphics offering, takes advantage of the tight coupling between CPU and GPU resources that Intel can bring to bear, since it now engineers the majority of the laptop platform (versus historically having to rely on a 3rd party discrete GPU). Intel Deep Link With Hyper Encode For Content Creation Workloads Intel
0 kommentar(er)
