A fresh round of leaks has sketched what looks like a modest but meaningful refresh of AMD’s mobile AI-focused APUs: the Ryzen AI Max 400 family, codenamed “Gorgon Halo.” If the snippets circulating through HKEPC, Wccftech, TechPowerUp and others are accurate, this isn’t a ground-up redesign — it’s a focused tuning exercise: higher CPU and iGPU clocks, faster memory support, and the same core counts buyers already know.

What leaked (and where caution matters)

The leak list centers on five Ryzen AI Max 400 SKUs: the Max 485, Max+ 488, Max 490, Max+ 492 and the flagship Max+ 495. Key points reported across multiple outlets:

  • The flagship Ryzen AI Max+ 495 is said to carry 16 Zen 5 CPU cores (32 threads) with clocks of 3.1 GHz base and a 5.2 GHz boost, paired with a Radeon 8060S iGPU running up to 3.0 GHz.
  • Other SKUs mirror existing Max 300-series core counts but come with raised CPU/iGPU clocks (many listed at 5.0 GHz boost for non-495 parts) and the same or slightly higher CU counts depending on the SKU.
  • Memory support reportedly jumps to LPDDR5X-8533 — an upgrade from the Max 300 family’s LPDDR5X-8000 limit.
  • TDPs are reported as a 55 W base with configurable ranges (45–120 W), and NPU performance is expected to sit somewhere around the mid-50s TOPS, though exact figures weren’t published.

    • Wccftech, Notebookcheck and TechPowerUp summarized HKEPC’s leaked OPN numbers and clock figures; VideoCardz also reported that “Gorgon Halo” engineering samples may already be in distribution. A note of caution: Notebookcheck flags that an image on HKEPC might have been doctored, and several original pages show Cloudflare verification pages or blocks — typical signs that this is still unofficial, early-stage information. Treat these numbers as plausible but unconfirmed until AMD speaks.

    Why this refresh makes sense

    This is the sort of refresh you expect when the competitive environment tightens. Intel’s Core Ultra “Panther Lake” family is pushing into the same thin-and-light and AI-capable laptop space, and AMD’s Ryzen AI line has become a visible differentiator for devices that want substantive on-device inferencing and capable integrated graphics.

    A few practical implications if the leaks hold up:

  • Gaming and graphics: the faster 8060S clocks in several SKUs will improve iGPU performance for thin-and-light gaming and GPU-accelerated workloads (think creative apps and some local model inference).
  • AI workloads: slightly higher NPU throughput plus the move to LPDDR5X-8533 give vendors more headroom for on-device models and responsive AI features. That ties into the wider trend of heavier on-device AI, as new models and tools emerge and push silicon demands in unexpected directions (see recent moves like Microsoft’s MAI-Image-1 and evolving integrations in productivity software).
  • OEM flexibility: the reported 45–120 W configurable envelope lets laptop makers tune thermals and performance for ultra-thin designs or thicker gaming/creator machines.

Those shifts matter outside just benchmarks. As cloud and local AI use cases blur — for example with heavier search and assistant features being integrated into apps — there’s a stronger argument for hardware that can accelerate inference on the endpoint. Google’s experiments with deeper AI integration in Docs and Gmail are another indicator of the kind of client-side demand that chips like these aim to satisfy (Gemini’s Deep Research).

Timing and what to watch for

Sources aren’t unanimous on timing. Some reporting suggests Gorgon Halo could show up later this year or early next, while Notebookcheck hedges toward an October–December 2026 window. That staggered timeline would let AMD phase inventory and give OEMs time to slot refreshed designs into holiday or next-year product cycles.

There’s also the question of platform reach. The rumor mill hints that the Gorgon Halo refinements could appear on desktop AM5 as well as in mobile designs, but that would depend on AMD’s roadmap priorities and demand from PC-makers.

Reading the leaks with a practical eye

If you’re shopping for a laptop today, these leaks don’t change much immediately: the current Ryzen AI Max/Strix Halo machines are still solid choices for users needing integrated AI and strong iGPU performance. For buyers planning a later purchase, the Gorgon Halo rumors suggest worthwhile gains in clocks and memory speed — enough to justify waiting if you need the latest in on-device AI acceleration.

Either way, expect AMD to be cautious: this looks like a performance-tuning refresh rather than a headline-grabbing new architecture. Keep an eye on official AMD channels for confirmation, and on OEM announcements that will reveal which SKUs land in which machines.

If you want to track how on-device AI demand is shaping hardware, those software-service stories are worth following in parallel — developments like Microsoft’s MAI-Image-1 and Google’s wider AI integrations (Gemini’s Deep Research) are part of the same picture that pushes vendors toward faster NPUs and beefier memory support.

Until AMD confirms, consider these leaks an intriguing glimpse rather than a shopping list. But if the clock bumps and LPDDR5X-8533 support are real, Gorgon Halo could be one of 2026’s quieter yet consequential hardware updates.

AMDRyzenAPUGorgon HaloLeak

Related Articles