Intel Unveils Powerful Arc Pro B-Series GPUs, Ushering in a New Era for AI and Professional Workstations

Intel has once again stepped into the spotlight with the launch of its latest graphics processing units, the Intel Arc Pro B60 and B50, announced at Computex 2025 in Taipei, Taiwan. Designed specifically for AI inference workloads and demanding professional applications, these GPUs leverage Intel’s cutting-edge Xe2 architecture. Featuring Intel Xe Matrix Extensions AI cores and enhanced ray tracing capabilities, the B60 and B50 GPUs boast impressive memory capacities of 24GB and 16GB, respectively—specs aimed squarely at creators, AI developers, and professionals in architecture, engineering, and construction.

The unveiling highlights Intel’s ongoing commitment to expanding its GPU ecosystem and delivering scalable, accessible solutions tailored for small and medium-sized businesses. Vivian Lien, Intel’s Vice President and General Manager of Client Graphics, emphasized how the B-Series reflects a broader vision to empower diverse industries with advanced graphics and AI technologies. Alongside the new GPUs, Intel introduced Project Battlematrix—a workstation platform capable of supporting up to eight Arc Pro B60 GPUs, enabling AI models with a staggering 150 billion parameters, a massive leap for enterprise AI capabilities.

This launch also commemorates Intel’s 40 years of collaboration with Taiwan’s tech ecosystem, with CEO Lip-Bu Tan underscoring the importance of these partnerships for future innovation. Intel also took the opportunity to announce the availability of Gaudi 3 AI accelerators in PCIe and rack-scale formats, aimed at enterprise and cloud AI inference, plus the public release of the Intel AI Assistant Builder on GitHub—tools that further extend Intel’s AI and software ecosystem reach.

Intel Arc Pro B60 GPUs will begin sampling through partners such as ASRock, Gunnir, and Sparkle in June 2025, with the B50 becoming available through authorized resellers the following month.

Intel’s Latest GPU Launch

At Computex 2025, Intel showcased its newest GPU offerings in the professional and AI sectors: the Intel Arc Pro B60 and B50. Built on the Xe2 architecture, these GPUs feature specialized AI cores designed to accelerate matrix computations, combined with advanced ray tracing units for superior visual rendering. With 24GB of memory on the B60 and 16GB on the B50, Intel is targeting professional markets including content creators, AI developers, and sectors such as architecture, engineering, and construction that require robust computational power.

Complementing these GPUs is Project Battlematrix, a scalable workstation platform that can house up to eight Arc Pro B60 units, enabling the training and deployment of AI models with an unprecedented 150 billion parameters. This represents a major capability increase for AI applications in enterprises.

Intel’s announcements also included broader AI ecosystem advancements: the Gaudi 3 AI accelerators now available for enterprise inference workloads and a new AI Assistant Builder platform, publicly accessible on GitHub to foster developer engagement. This marks a strategic push by Intel to integrate hardware and software innovation for AI and professional workloads.

The rollout will begin with partner sampling in June 2025, followed by general availability in July, indicating Intel’s intent to capture significant market share in the competitive GPU landscape for AI and professional computing.

What Undercode Say:

Intel’s launch of the Arc Pro B-Series GPUs is a clear strategic move to reclaim and expand its footprint in the professional graphics and AI computing markets, long dominated by competitors like NVIDIA and AMD. By leveraging the new Xe2 architecture with specialized AI cores, Intel is not only addressing the growing demand for AI inference acceleration but also targeting industries that require a fusion of high-performance graphics and computational accuracy—such as AEC (Architecture, Engineering, and Construction).

The memory configurations of 24GB and 16GB in these GPUs position them well against existing workstation GPUs, enabling large datasets and complex models to be handled more efficiently. Project Battlematrix, with its support for up to eight GPUs running massive AI models, signals Intel’s intention to compete seriously in the enterprise AI training and inference arena, a space critical for next-gen AI applications from large language models to complex simulation environments.

Intel’s parallel announcement of Gaudi 3 AI accelerators and open-source AI development tools underscores a holistic ecosystem approach. By providing both hardware for high-performance computing and accessible developer tools, Intel is fostering a community-driven growth model that can accelerate adoption across industries.

However, Intel faces challenges. The GPU market is fiercely competitive, with NVIDIA still leading in AI-centric GPU performance and software ecosystem maturity. Intel’s success will depend on how well its partners integrate and optimize these GPUs, and how swiftly the software ecosystem matures to fully utilize the Xe Matrix Extensions AI cores.

In terms of timing, aligning the launch with Intel’s 40th anniversary in Taiwan reflects both a nod to legacy and a commitment to innovation in a key regional market. The focus on SMBs also highlights Intel’s effort to democratize access to advanced AI and graphics technologies, a segment often underserved by top-tier GPU vendors.

Overall, Intel’s Arc Pro B-Series and associated platforms represent a significant leap forward in their GPU roadmap, potentially reshaping competitive dynamics in professional and AI computing markets. The next year will be critical to see how adoption scales and performance benchmarks compare in real-world applications.

🔍 Fact Checker Results:

✅ Intel officially announced the Arc Pro B60 and B50 GPUs at Computex 2025 in Taipei.
✅ The GPUs are based on the Xe2 architecture and include specialized AI cores and ray tracing units.
✅ Project Battlematrix supports up to eight B60 GPUs and targets AI models with up to 150 billion parameters.

📊 Prediction:

Intel’s Arc Pro B-Series GPUs are poised to disrupt the mid-tier professional GPU market, particularly within AI inference and specialized workstation environments. Over the next 12 to 18 months, expect growing adoption among SMBs and enterprise customers seeking cost-effective AI acceleration solutions. If Intel can rapidly expand software support and partner integrations, this launch could significantly challenge NVIDIA’s dominance, especially in sectors like architecture and engineering that demand both graphical fidelity and AI computation. Moreover, the ecosystem tools like the AI Assistant Builder will likely accelerate developer interest and innovation, further strengthening Intel’s position in the AI hardware landscape.