The latest AMD news has everybody excited as they officially announce the existence of the Ryzen AI MAX+ 395 mobile processor.
This APU comes with “Zen 5” cores that offer 50+ AI TOPS with its XDNA 2 NPU plus the Radeon 8060S iGPU with 40 RNDA 3.5-based Compute Units for an absolute banger laptop performance. With support up to 128GB of unified memory, up to 96GB of those can be assigned for VRAM via the Variable Graphics Memory tech.
All of this resulted in exceptional performance in AI applications like LM Studio, outperforming competitors (Yes, the Intel Core Ultra 7 258V, not any Core Ultra 9 chip by any means) by up to 2.2x in token throughput and up to 12.2x in time to first token for large language models. Specifically, the IBM Granite Vision and Google Gemma 3 is around 7x faster, with Team Red stating that a top-specced system with the processor can comfortable handle the Gemma 3 27B Vision model.
AMD has also extended its AI capabilities to space applications with the Versal AI Edge XQRVE2302 adaptive SoC, which recently achieved Class B qualification for spaceflight. This radiation-tolerant device brings accelerated AI inferencing to space with integrated AI Engines optimized for machine learning.
The SoC features a dual-core Arm Cortex-A72 application processor, a dual-core Arm Cortex-R5F real-time processor, along with AI engines, DSP blocks, and FPGA programmable logic in a small 23mm x 23mm package. This allows for powerful edge processing in space for applications like image detection and classification, autonomous navigation, and sensor data processing. The device supports unlimited reprogramming even after deployment in the harsh radiation environment of space, making it flexible for evolving mission requirements.
