Chinese original design manufacturer Sixunited has officially unveiled the AXN88B, marking the first laptop reference design powered by the highly anticipated AMD Gorgon Halo APU. Built to handle massive local AI workloads, the system supports up to a staggering 192 GB of RAM, signaling a major shift in portable workstation capabilities.
At the heart of the AXN88B is the AMD Ryzen AI Max+ PRO 495 APU. This processor features a 16-core, 32-thread configuration built entirely on full-fat Zen 5 cores. Compared to its predecessor, the Ryzen AI Max+ 395, the new Gorgon Halo silicon delivers higher boost clocks across both the CPU and GPU. Leaked Passmark benchmarks suggest the chip operates roughly 5% faster than the existing Strix Halo platform, while supporting ultra-fast 8533 MT/s memory.
Graphics performance receives a significant architectural push. The system utilizes the Radeon 8065S iGPU, which offers a slight performance edge over the Radeon 8060S. Both integrated graphics units feature a 40 Compute Unit (CU) setup based on the RDNA 3.5 architecture. This configuration promises synthetic benchmark performance that rivals a discrete RTX 4070 Laptop GPU, all within an integrated package.
Display and Hardware Specifications
The reference chassis houses a 16-inch display with a 2560 x 1600 resolution and a 120 Hz refresh rate, complete with P3 color gamut support. Powering the hardware is an 85 Wh battery, alongside dual 2 W speakers, a full-size keyboard with RGB backlighting, fingerprint recognition, and a 2 MP IR front camera.
Sixunited has equipped the AXN88B with a comprehensive array of I/O ports for professional connectivity:
- Dual USB 4 Type-C ports
- Single USB 3.2 Gen 1 Type-A port
- Single USB 3.2 Gen 2 Type-A port
- UHS-II SD card reader
- HDMI 2.1
- 2.5G Ethernet
The Local LLM Workstation Redefined
Because Sixunited operates as an original design manufacturer (ODM), the AXN88B will not hit retail shelves under its own name. Instead, consumer-facing brands will license and modify this reference design in the coming weeks. However, the baseline specifications reveal AMD's aggressive strategy for the AI era.
By allowing up to 192 GB of unified memory to be shared with a 40 CU integrated GPU, AMD is solving the biggest bottleneck for developers running large language models (LLMs) locally. Traditional discrete laptop GPUs are strictly limited by their VRAM capacity, but the Gorgon Halo architecture effectively turns the entire system memory into a massive pool for AI processing, making it a formidable alternative to traditional mobile workstations.
