Phison Electronics announced expanded capabilities for aiDAPTIV+ to deliver more powerful AI Processing on PC Platforms. The company now supports integrated GPU architectures and accelerates inference while increasing memory capacity. Furthermore, Phison simplifies deployment so users can unlock large-model AI capabilities on notebooks, desktops, and mini-PCs. Phison highlights AI Processing on PC Platforms as a critical innovation pathway for scaling everyday computing power.
As AI workloads grow, organizations fac
Phison also talked about collaborations that help build a stronger ecosystem. In essence, working with Acer is a perfect example of how aiDAPTIV+ makes larger LLM training possible on laptops with less DRAM. As a result, users benefit from improved privacy, scalability, lower costs, and simpler deployment processes. Moreover, OEMs and system integrators have the capability to offer complete solutions that not only break the GPU VRAM constraints but also facilitate wider AI adoption.
“As AI models grow into tens and hundreds of billions of parameters, the industry keeps hitting the same wall with GPU memory limitations,” said Michael Wu, President and GM, Phison US. “By expanding GPU memory with high-capacity, flash-based architecture in aiDAPTIV+, we offer everyone, from consumers and SMBs to large enterprises, the ability to train and run large-scale models on affordable hardware. In effect, we are turning everyday devices into supercomputers.”
“Our engineering collaboration enables Phison’s aiDAPTIV+ technology to accommodate and accelerate large models such as gpt-oss-120b on an Acer laptop with just 32GB of memory,” said Mark Yang, AVP, Compute Software Technology at Acer. “This can significantly enhance the user experience interacting with on-device Agentic AI, for actions ranging from simple search to intelligent inquiries that support productivity and creativity.”
Reduced TCO and Memory Consumption
aiDAPTIV+ lowers memory pressure by shifting MoE inference loads from DRAM to flash-based cache memory. Phison’s internal testing confirms that a 120B-parameter model now runs with 32GB of DRAM instead of 96GB. As a result, more devices can execute complex MoE workloads without expensive upgrades.
Faster Inference Performance
aiDAPTIV+ stores tokens that exceed KV cache limits and reuses them during new prompts. This increases response speed by up to ten times, based on Phison testing. Additionally, power consumption drops while notebo
Bigger Models on Smaller Devices
Phison claims that aiDAPTIV+, combined with new Intel Core Ultra Series 3 processors, supports direct LLM training on notebooks. Early testing shows a notebook can now fine-tune a 70B-parameter model, which previously required far costlier workstation-class hardware. This opens AI development to students, researchers, and smaller organizations.
Ecosystem Demonstrations at CES 2026
Phison will showcase aiDAPTIV+ with partner systems at CES 2026, including notebooks, desktops, and mini-PCs from Acer, Corsair, MSI, and NVIDIA. MSI will also demonstrate an AI-accelerated workflow that summarizes meeting notes through aiDAPTIV+. ASUS and Emdoor will present additional systems running aiDAPTIV+ for enhanced inference performance.
Explore IT Tech News for the latest advancements in Information Technology & insightful updates from industry experts!
News Source: Businesswire.com
