Samsung is developing a dedicated AI accelerator for PCs codenamed GAIA, according to multiple Korean outlets, including Chosun. HP in the US and Lenovo in China are already testing prototypes to verify performance.
Mass production could start as early as 2027, with devices potentially arriving in late 2027 or early 2028. GAIA is being developed by Samsung’s LSI division, which also works on the Exynos mobile chips.
Samsung has not officially confirmed the project. The chip has no publicly disclosed performance figures, power specifications, or architectural details.
What Makes Samsung’s GAIA AI Chip Different
GAIA is described as a companion processor rather than a general-purpose CPU. Key features reported by Korean outlets:
- Built on a 4nm-class process node
- Described as a “memory-centric” AI accelerator that places compute close to memory
- Not intended to replace Intel, AMD, or Qualcomm processors, but to work alongside them
- Positioned specifically for PC-side generative AI workloads
- Explicitly separated from GPU-based AI accelerators used for large-scale training and inference
The chip is aimed at on-device language models, real-time translation, image generation, and similar tasks that would otherwise run on the CPU or GPU. The intent is to offload these workloads to a dedicated NPU rather than sharing resources with the primary processors.
Samsung is reportedly pushing further integration with processing-in-memory (PIM) technology, which runs computations inside the memory itself instead of moving data back and forth to a processor.
PIM has been a Samsung research project for years without a major commercial breakthrough. GPUs became fast enough, and their software ecosystems matured quickly enough, that the bottlenecks PIM was designed to address became less pressing.
A dedicated NPU with real OEM validation may provide a better fit for PIM than GPUs. Samsung is one of the few companies that can pair custom AI logic with its own DRAM manufacturing, giving it a vertical integration advantage that competitors would find difficult to match.
Samsung’s Return to PC Silicon and Business Stakes
Samsung last supplied PC silicon in 2012, when Exynos chips briefly powered early Samsung Chromebooks. The business was shelved two years later. Since then, Samsung’s Galaxy Book laptops have used Intel or Qualcomm chips, including the recent Snapdragon X2 Elite in the latest Galaxy Book.
If GAIA reaches mass production, Samsung would have its own logo on the silicon inside its own laptops for the first time in more than a decade. Third-party OEM adoption through HP and Lenovo could expand Samsung’s reach beyond its own product line.
Samsung’s LSI division has been incurring structural losses for years. A successful AI venture, combined with revenue from Exynos and automotive silicon, could provide the division with a much-needed growth opportunity.
The strategy also creates potential conflicts with existing customers. Nvidia and Qualcomm both rely on Samsung’s foundry for parts of their chip production. Samsung competing with its own clients in the AI PC space while still manufacturing for them might complicate relationships.
How Samsung manages this balance will influence both its foundry business and its ability to sell GAIA to PC manufacturers who also use Qualcomm or Nvidia components.
What PC Buyers Should Watch For and When GAIA Could Arrive
For PC buyers considering AI-capable devices in 2027 and 2028, GAIA could offer an additional option alongside existing solutions such as Intel Core Ultra with an integrated NPU, AMD Ryzen AI with XDNA NPU, Qualcomm Snapdragon X2 with Hexagon NPU, and Nvidia RTX Spark for AI workloads.
When evaluating AI PC hardware, users should consider the types of workloads they run and whether an NPU can provide meaningful performance improvements over a CPU or GPU alone.
It’s also important to assess software support for the specific NPU architecture and whether local AI processing justifies a dedicated accelerator, or if cloud-based AI services might meet their needs.
Despite industry efforts over the past two years to promote NPUs as essential components of AI PCs, many users struggle to identify tasks that their current NPU handles more effectively than traditional hardware. Introducing additional NPU vendors does not fundamentally alter this situation.
There are still many details about GAIA that remain unclear, including how its performance compares to competing NPUs, its power consumption and thermal characteristics, and the maturity of its software support and drivers at launch.
Other unknowns include which PC form factors Samsung plans to target first, whether GAIA will need dedicated memory or will work with standard DRAM, and the pricing and licensing terms for OEMs beyond HP and Lenovo.
Since Samsung has not officially announced the project, all current information is based on reports from Korean industry sources and signals from OEM testing.
Availability
GAIA is currently in the prototype testing phase with HP and Lenovo. Mass production is expected to begin in 2027, with consumer devices possibly launching late that year or early 2028.
The timeline may change depending on the validation results, the commitment from OEMs, and Samsung’s ability to provide software support alongside the hardware.
Those interested in GAIA-based PCs should keep an eye on Samsung’s official announcements and OEM product roadmaps as CES 2027 and other industry events approach.
Samsung has not yet announced a date for a public unveiling of GAIA. Additional details are likely to emerge as the chip moves from prototype testing toward readiness for mass production.
Thank you for being a Ghacks reader. The post Samsung Develops Dedicated GAIA AI Chip for PCs With HP and Lenovo Testing Prototypes appeared first on gHacks.