At Computex 2025 in Taipei, Intel is making a bold play for the AI and workstation market with fresh GPU offerings and enhanced AI accelerators. From powerful silicon for prosumers to open-source AI toolkits, Intel is aiming squarely at developers, creators, and businesses seeking scalable AI compute.
Here’s what the chipmaker rolled out at this year’s tech showcase in Taipei, alongside a major milestone: its 40-year partnership with Taiwan’s vibrant tech ecosystem.
Arc Pro GPUs Built for Workstations and AI
Intel has unveiled two new entries in its Arc Pro graphics card lineup — the Arc Pro B60 and B50 — engineered for intensive AI inference workloads and creative professional use. Based on the Xe2 architecture, both cards come equipped with Intel Xe Matrix Extensions (XMX) AI cores and hardware-accelerated ray tracing.
The Arc Pro B60 boasts 24GB of memory, while the B50 features 16GB. The cards are tailored to support multi-GPU setups, ensuring high scalability for engineers, AI developers, and 3D creators. Designed with architecture, engineering, and construction (AEC) professionals in mind, these GPUs are certified by major software vendors and offer stable performance on both Windows and Linux platforms. Linux users, in particular, benefit from a containerised software stack that simplifies AI deployment workflows.
Vivian Lien, vice president and general manager of Client Graphics at Intel, said, “The Intel Arc Pro B-Series showcases Intel’s commitment in GPU technology and ecosystem partnerships. With Xe2 architecture’s advanced capabilities and a growing software ecosystem, the new Arc Pro GPUs deliver accessibility and scalability to small and medium-sized businesses that have been looking for targeted solutions.”
The B60 will begin sampling in June 2025 via vendors such as ASRock, Gunnir, Maxsun, and Sparkle. The B50 will hit the market in July through Intel-authorized resellers.
Gaudi 3 Gets Rack-Scale and PCIe Boost
Alongside the GPU launch, Intel expanded its Gaudi 3 AI accelerator strategy, offering options for both PCIe and rack-scale environments.
For businesses that want flexibility within existing data centre infrastructure, Gaudi 3 PCIe cards allow scalable AI inference using models like Llama 3.1 8B all the way to Llama 4 Scout or Maverick. These cards will be available in the second half of 2025.
Meanwhile, rack-scale Gaudi 3 systems are built for big AI workloads, supporting up to 64 accelerators per rack and delivering 8.2 terabytes of high-bandwidth memory. With open modular architecture and liquid cooling, Intel promises lower TCO and performance gains without vendor lock-in.
Intel AI Assistant Builder Goes Public
Developers can now get their hands on the Intel AI Assistant Builder, a lightweight software toolkit designed to create local, purpose-driven AI agents on Intel-based platforms. After its debut at CES 2025, the toolkit is now available in public beta on GitHub.
With companies like Acer and ASUS already demonstrating its potential at Computex, the toolkit is expected to accelerate local AI agent development for a range of enterprise and consumer applications.
A Milestone Moment in Taiwan
This wave of AI-centric announcements also marks four decades of Intel’s presence in Taiwan — a partnership the company says has been key to advancing the x86 ecosystem.
“For the past 40 years, the power of our partnership with the Taiwan ecosystem has fueled innovation that has changed our world for the better,” said Intel CEO Lip-Bu Tan. “This week, we are renewing our commitment to our partners as we work to build a new Intel for the future. Together, we will create great products that delight our customers and capitalise on the exciting opportunities ahead.”
With AI demands skyrocketing across industries, Intel’s Computex 2025 announcements signal a clear intent: to be at the heart of the AI hardware race — and to do it with open systems, developer-first tools, and long-term partner alliances.
