TL;DR
AMD’s upcoming Ryzen AI MAX+ 495 “Strix Halo” flagship has leaked with a 10% performance uplift over the previous Strix Halo generation, a new Radeon 8065S integrated GPU, and support for up to 192GB of unified memory. This positions AMD to challenge both high-end mobile CPUs and discrete GPUs in a single package, making it a critical product for the company’s AI PC and gaming ambitions in the second half of 2026.
What Happened
A comprehensive leak from Wccftech has revealed AMD’s next-generation Halo flagship, the Ryzen AI MAX+ 495, codenamed “Gorgon Halo,” which delivers a 10% generational performance increase over its predecessor. The chip packs a new Radeon 8065S integrated GPU and support for up to 192GB of unified memory, setting a new bar for integrated graphics in laptop-class processors.
Key Facts
- The Ryzen AI MAX+ 495 is AMD’s flagship “Gorgon Halo” part, succeeding the Strix Halo series launched in early 2025.
- Performance is 10% higher than the previous Strix Halo flagship, according to leaked benchmarks from Wccftech.
- The chip features a new Radeon 8065S integrated GPU, a step up from the Radeon 8050S found in Strix Halo.
- Memory support expands to 192GB of unified LPDDR5X memory, up from 128GB in the prior generation.
- The leak originates from Wccftech on May 3, 2026, citing unnamed industry sources with access to pre-production silicon.
- AMD is expected to target AI workloads, high-end gaming laptops, and workstation-class mobile devices with this chip.
- The Ryzen AI MAX+ 495 leverages AMD’s Zen 5 CPU cores paired with the RDNA 3.5 GPU architecture, both refined for the Halo segment.
Breaking It Down
The 10% performance bump in the Ryzen AI MAX+ 495 is notable not because it is massive — it is modest by historical generational standards — but because it comes from a platform that already pushed the boundaries of integrated graphics. The previous Strix Halo chips, such as the Ryzen AI 395+, delivered integrated GPU performance that rivaled NVIDIA’s GeForce RTX 4060 laptop GPU in many titles. A 10% improvement means the Radeon 8065S could match or exceed the RTX 4070 mobile in rasterization, while consuming less power and occupying less motherboard space.
The 192GB unified memory ceiling is the single most striking figure in this leak — it is 4x the memory capacity of most current high-end gaming laptops and matches what many workstation desktops shipped with just two years ago.
This memory capacity is not for gamers. It targets AI developers and content creators who need to load large language models (LLMs) or massive 3D scenes entirely in unified memory. Apple’s M4 Ultra chips support up to 256GB, but AMD is now closing that gap at a lower price point. The 192GB limit suggests AMD is using 32GB LPDDR5X memory modules, likely from SK Hynix or Samsung, in a quad-channel configuration. This also means the memory bandwidth could exceed 500 GB/s, critical for both GPU compute and AI inference.
The Radeon 8065S naming is a departure from AMD’s previous integrated GPU naming scheme. The “S” suffix likely denotes “Strix” or “Super,” indicating this is a higher-tier part than the Radeon 8050S in Strix Halo. Given that AMD’s discrete Radeon RX 8000 series is also expected in 2026, the 8065S may share architectural features with those discrete GPUs, such as improved ray tracing accelerators and AI upscaling hardware. This would make the Ryzen AI MAX+ 495 a legitimate competitor to NVIDIA’s GeForce RTX 5070 laptop GPU in ray-traced titles, a segment where integrated graphics have historically struggled.
What Comes Next
The leak points to a product that is still in pre-production, meaning the final specifications and performance could shift before launch. Here are the key developments to watch:
-
Official AMD announcement timeline: Expect AMD to formally unveil the Ryzen AI MAX+ 495 at CES 2027 in January, with laptops shipping in Q1 2027. However, AMD may preview the architecture at Computex 2026 in June to build hype.
-
Laptop OEM adoption: Key partners like ASUS, Lenovo, and HP will need to design new chassis to accommodate the high-power Halo platform. The first wave of Gorgon Halo laptops is expected to include 16-inch and 18-inch models with advanced cooling solutions.
-
Competitive response from NVIDIA and Intel: NVIDIA is expected to launch its GeForce RTX 5060 and 5070 laptop GPUs in late 2026, while Intel will counter with Arrow Lake-HX and Panther Lake platforms. The Ryzen AI MAX+ 495 will need to prove its performance edge in real-world benchmarks, not just leaked numbers.
-
Pricing and availability: The 192GB memory configuration will command a premium. Expect the top-end Ryzen AI MAX+ 495 laptops to start at $2,500–$3,000, with the 192GB models exceeding $4,000. Volume availability will depend on LPDDR5X supply from memory manufacturers.
The Bigger Picture
This leak fits into two larger trends reshaping the PC industry. First, the convergence of CPU and GPU into a single high-performance package is accelerating. AMD’s Halo line, Apple’s M-series, and Intel’s upcoming Arrow Lake with integrated Arc graphics all point toward a future where discrete GPUs become optional for many users. The Ryzen AI MAX+ 495, with its 192GB memory ceiling, directly challenges the notion that serious AI work requires a desktop with separate GPU memory.
Second, the AI PC revolution is driving memory capacity higher across all segments. Microsoft’s Copilot+ PC initiative requires at least 16GB of RAM, but AI workloads like running local LLMs demand 64GB or more. AMD’s decision to support 192GB in a mobile platform signals that the company expects users to run 70B-parameter models locally, a use case that was unimaginable in a laptop just two years ago. This positions AMD to capture the growing market of developers and researchers who need portable AI workstations.
Key Takeaways
- [Performance Uplift]: The Ryzen AI MAX+ 495 delivers a 10% generational performance increase over Strix Halo, pushing integrated graphics into RTX 4070 territory.
- [Memory Ceiling]: Support for 192GB of unified memory quadruples the previous generation’s limit, targeting AI and workstation workloads.
- [New GPU Architecture]: The Radeon 8065S iGPU introduces a new naming scheme and likely shares architectural features with discrete Radeon RX 8000 series GPUs.
- [Market Timing]: Expect official launch at CES 2027, with the chip competing against NVIDIA’s RTX 5000-series laptop GPUs and Intel’s Arrow Lake-HX.
