ASUS Launches Its First Intel Arc Pro B70 Graphics Card: 32GB ECC VRAM, AI Compute Power Reaching 367 TOPS
ASUS launches its first Intel Arc Pro B70 graphics card, targeting the AI inference market, equipped with 32GB ECC VRAM and INT8 AI compute power reaching 367 TOPS.
Key Takeaways
- Key Highlight：ASUS launches its first Intel Arc Pro B70 graphics card, targeting the AI inference market, equipped with 32GB ECC VRAM and INT8 AI compute power reaching 367 TOPS.
- Innovation & Tech：Highlights advancements in ASUS, Launches, Its, demonstrating rapid progress in model capabilities.
- Industry Impact：Reported via IT之家 (CN), offering actionable signals for developers and technology leaders.
The UGen Intel Arc Pro B70 32GB GDDR6 graphics card released by ASUS is the company's first professional-grade product based on the Intel Arc Pro series, explicitly targeting AI inference and workstation scenarios. The product features an Intel Xe2 architecture Big Battlemage GPU, equipped with 32 Xe cores and 256 XMX AI engines, delivering AI compute power of up to 367 TOPS at INT8 precision, and is equipped with 32GB GDDR6 ECC VRAM to meet the demands of professional inference tasks for VRAM capacity and data reliability.
ASUS had been absent from multiple generations of Intel Arc discrete graphics cards, and its return to the Intel discrete graphics lineup marks that Intel has gained more OEM partner support in the professional AI inference hardware space. The XMX engine is a matrix compute unit designed by Intel specifically for AI workloads, and the 367 TOPS INT8 compute power is somewhat competitive among professional inference cards, making it particularly suitable for deploying small-to-medium scale large model inference services.
In the AI inference hardware market, NVIDIA has long held a dominant position, and Intel is attempting to establish alternatives in the professional inference space through the Arc Pro series. ASUS's entry means more brands will offer AI inference solutions based on Intel GPUs, helping to reduce hardware procurement costs for enterprises on the inference side, while providing developers with more diversified hardware choices.
However, the Intel Arc Pro series still needs to be compared and validated against the NVIDIA CUDA ecosystem in terms of software ecosystem, driver maturity, and compatibility with mainstream AI frameworks. The product's actual market performance will depend on its energy efficiency ratio, stability, and level of ecosystem support in real inference scenarios.
This page provides an editorial summary based on publicly available information. It is not a republished article. Use the source link below for the original report.
Industry Insights & Analysis
As artificial intelligence rapidly evolves, breakthroughs surrounding ASUS, Launches, Its, First are shifting toward scalable, robust real-world implementations.
Driven by both open-source ecosystems and proprietary model architectures, the integration between compute optimization, data engineering, and agentic workflows is accelerating. This development provides a strategic benchmark for upcoming AI tooling and developer workflows.