AMD’s Ryzen processor (Fig. 1) incorporates a wide range of improvements, including artificial intelligence for its branch prediction. The 8-core chip implements simultaneous multithreading supporting 16 threads (8C/16T). It has a TDP under 100 w and is delivering performance similar or better than Intel’s Core i7-6900K that has a 140 W TDP. Some of the chip’s details are sketchy, but it clock starts at 3.4 GHz.
AMD employs a neural network in its branch prediction subsystem. It builds a model of the code being executed so that its Smart Prefetch can pre-load instructions and optimize the path through the processor pipeline. The neural network is designed to learn from the currently running applications rather than the predefined static analysis often used in other deep learning applications.
AMD’s Pure Power technology (Fig. 2) is designed to minimize power requirements while providing optimum performance. The “infinity fabric” is a power management system that works throughout the chip to monitor and control the temperature, clock frequency, and system voltages. The Infinity System Management Unit provides the adjustments.
The Ryzen SenseMI (pronounced “Sense Em-Eye”) technology implements automatic overclocking. The system also improves cooling when combined with the Pure Power support and the Precision Boost uses a 25 MHz granularity. The Extended Frequency Range (XFR) is designed for overclockers that have high-end cooling systems.
The chip will run on AMD’s AM4 motherboard with the Bristol Ridge chipset. It has a dual memory channel DDR4 memory subsystem. It supports USB 3.1 Gen 2, NVMe, and SATA Express.
Ryzen represents a significant performance boost that puts it on par with the fastest Intel desktop chips while utilizing less power. The anticipated price for Ryzen is less than Intel’s Intel Core i7 6900K that has a street price around $1,000, and it should bring some price competition in the high-end gaming space in 2017.
