AMD has just introduced a revolutionary product for industrial AI: the Ryzen 8000 series of embedded processors. These processors are the first to integrate a neural processing unit (NPU) built on AMD’s XDNA architecture, which is specifically designed to handle the high workloads of industrial AI applications.
The Ryzen 8000 series boasts impressive performance. Built on an advanced 4nm process technology, it combines the Zen 4 processor architecture, a GPU based on the RDNA 3 architecture, and an all-new neural processor based on the XDNA architecture. This powerful system delivers a staggering 39 TFlops of AI computing power, with 16 TFlops coming from the NPU.
But the power doesn’t stop with artificial intelligence. Ryzen 8000 series processors have additional features: 20 lanes of high-speed PCIe Gen 4 interface provide efficient data transfer, and support for dual-channel DDR5 5600 memory with error correction (ECC) ensures smooth operation. The machines can deliver stunning images with the ability to power four independent 4K displays. Hardware decoding and encoding of popular video codecs such as AV1, H.264, and H.265 simplifies video processing.
To meet a variety of needs, the Ryzen 8000 series comes in four different models. To build a strong ecosystem around these processors, AMD has collaborated with partners such as Advantech, ASRock, and iBASE, who are all Original Design Manufacturer (ODM) embedded system manufacturers. As artificial intelligence becomes more commonplace, research in this area is gaining momentum. That’s why companies like Nvidia and AMD are producing newer AI chips for larger-scale research.
