AMD Ryzen AI MAX 495 Leaks with iGPU Upgrade

Post LinkedIn

🇨🇳Read original on cnBeta (Full RSS)

#ai-pc #cpu-refresh #integrated-gpuryzen-ai-max-400

💡AMD's next AI laptop CPU leaks: higher clocks, more RAM, RDNA 3.5

⚡ 30-Second TL;DR

What Changed

Ryzen AI MAX 400 series codenamed Gorgon Halo

Why It Matters

Enhances AI PC performance for laptops, positioning AMD strongly in mobile AI compute market against Intel and Qualcomm.

What To Do Next

Benchmark upcoming Ryzen AI MAX 400 in MLPerf for laptop AI inference.

Who should care:Developers & AI Engineers

🧠 Deep Insight

AI-generated analysis for this event.

🔑 Enhanced Key Takeaways

•The Gorgon Halo series is positioned to leverage a massive 120W+ TDP envelope, targeting high-end mobile workstations and enthusiast gaming laptops that require discrete-class graphics performance without a dedicated GPU.
•The architecture utilizes a multi-chiplet design, integrating a dedicated NPU capable of exceeding 80 TOPS, aligning with Microsoft's Copilot+ PC requirements for next-generation AI workloads.
•Memory support is expected to transition to LPDDR5X-8533 or potentially LPDDR6, addressing the bandwidth bottlenecks inherent in high-core-count APUs when driving large RDNA 3.5 compute units.

📊 Competitor Analysis▸ Show

Feature	AMD Ryzen AI MAX 495 (Gorgon Halo)	Intel Core Ultra 200 Series (Lunar Lake/Arrow Lake)	Apple M4 Pro/Max
iGPU Performance	High (Targeting 40+ CUs)	Moderate (Xe2 Architecture)	High (Unified Memory Architecture)
NPU TOPS	>80 TOPS	~48 TOPS	~38 TOPS
Memory Architecture	LPDDR5X/6 (High Bandwidth)	LPDDR5X (On-Package)	Unified Memory (High Bandwidth)
Target Segment	Enthusiast/Workstation APU	Thin-and-Light/Mainstream	Premium Laptop/Workstation

🛠️ Technical Deep Dive

Architecture: Multi-chiplet design utilizing AMD's Infinity Fabric to connect CPU complex (Zen 5) and GPU complex (RDNA 3.5).
GPU Configuration: Expected to feature up to 40 Compute Units (CUs), significantly higher than standard mobile APUs, requiring high-speed memory to sustain throughput.
Platform: FP11 socket, designed specifically for high-TDP mobile applications, supporting advanced power management features for thermal throttling control.
AI Engine: Integrated XDNA 2/3 architecture NPU, optimized for low-latency inference in local LLMs and generative AI applications.

🔮 Future ImplicationsAI analysis grounded in cited sources

AMD will capture significant market share in the premium thin-and-light workstation segment.

By eliminating the need for a discrete GPU, Gorgon Halo allows for thinner chassis designs while maintaining high-performance compute and graphics capabilities.

The 80+ TOPS NPU will become the new baseline for high-end mobile processors by late 2026.

The industry is rapidly shifting toward local AI processing, and AMD's aggressive NPU scaling forces competitors to accelerate their own AI silicon roadmaps.