Tag
AMD details its new Ryzen AI Embedded X100 Series processors and Kria AI platforms, aiming to lead in the physical AI market with high-performance compute and unified AI acceleration for edge systems.
A security research tool that exploits DRAM address scrambling on AMD CPUs to unlock hidden platform features such as PSP, SMM, C6, and microcode, demonstrating a deep hardware-level attack.
User reports running Qwen 3.6 35B A3B-Q8_0 gguf on a Radeon 7600 with llama.cpp and ROCm, achieving 21 tokens per second after VRAM overclocking, with a note about a display-related performance bug.
At OCP APAC 2026, AMD, Arm, and Microsoft executives argue that agentic AI workloads will increase CPU demand significantly, potentially shifting CPU-to-GPU ratios from 1:4 to 1:2 or even 1:1, while Microsoft positions its Cobalt 200 chip as an 'agent-native CPU'.
SemiAnalysis founder Dylan Patel will speak at the Runtime event to discuss compute constraints, Kimi K3 tokenomics, and AMD supply chain issues. Applications to attend are open.
AMD has reached a definitive agreement to acquire Toronto-based AI chip startup Taalas, which specializes in hardwiring AI models onto custom silicon for efficient inference. The deal aims to strengthen AMD's AI portfolio with differentiated inference performance and efficiency.
QuixiAI comments on Taalas Inc joining AMD, expressing hope that high-performance inference hardware becomes available to consumers rather than being reserved for data centers, quoting Taalas's acquisition announcement.
AMD acquires AI chip startup Taalas, which etches model weights directly into silicon, promising order-of-magnitude inference performance boosts. The deal is seen as AMD's move to challenge Nvidia's dominance in AI hardware.
AMD's Q2 2026 earnings show data center revenue surging 107% YoY to $6.7 billion on AI demand, while gaming revenue fell 31% due to component shortages and price hikes. Overall revenue hit a record $11.5 billion, up 50% YoY.
The tweet shares evidence that NVIDIA intentionally shapes its software ecosystem to make newer workflows appear Blackwell-exclusive while leaving Ampere paths unsupported, advising users to wait before upgrading and suggesting NVIDIA's moat is weakening in favor of Intel and AMD.
AMD highlights that running cloud models for Blender and Hermes agent desktop still needs a powerful CPU, promoting Ryzen AI Agent Computers as the platform.
A new AMD Linux kernel patch improves low-end gaming performance on Steam Deck by monitoring core workloads and temporarily setting busy cores to performance mode, resulting in up to 31.8% improvement in 1% low FPS in Civilization VI.
AMD and OpenAI engineers collaborate to share insights on performance optimization, featuring a behind-the-scenes look with OpenAI's VP of Compute Strategy, Sachin Katti.
ASRock announces the Radeon RX 9050 with 4GB VRAM, the first modern GPU in years with such low memory, likely targeting budget builds due to component shortages.
Chirag Dekate discusses that the biggest bottleneck in AI spans the entire stack, not just chips or energy, during AMD's Advancing AI event.
AMD has achieved a 46% revenue share in x86 server CPUs, up from nearly zero in 2017, driven by its new Venice processors that offer 2.2x performance per core over Nvidia's comparable chip. The shift highlights a growing preference for modular CPU+GPU setups in AI agentic workloads over Nvidia's vertically integrated racks.
A report on building the world's first cluster of AMD Ryzen AI Halo processors, which turns out to be underwhelming despite its novelty.
AMD released an open-source Mixture-of-Experts model Instella-MoE-16B-A3B on HuggingFace.
TIME highlights the concept of world models as a step beyond language models for AI understanding, featuring a conversation with Oliver Cameron of Odyssey at AMD's Advancing AI event.
Matt White, global CTO of AI at The Linux Foundation, argued at AMD's Advancing AI event that open models should not be viewed through a geopolitical lens, emphasizing that once open, they are for everyone to build on.