ASUS Launches Intel Arc Graphics Card with 32GB Memory for Workstations
ASUS's new Arc Pro B70 workstation GPU arrives in Europe with premium cooling and higher power for AI workloads.
ASUS has begun selling its UGen Intel Arc Pro B70 graphics card across Europe, marking an important expansion of Intel’s professional GPU lineup. The card is now available in Germany, Austria, and Switzerland at a recommended retail price of €1,799, or CHF 1,799 in Switzerland. This launch represents the first partner-built version of Intel’s workstation focused graphics card to reach consumers, several months after Intel introduced its own branded Arc Pro B70 model in March 2026.
The arrival of ASUS’s variant adds competition and choice to the professional graphics card segment, though pricing tells an important part of the story for potential buyers. Intel’s own Arc Pro B70 launched at $949, but ASUS’s European pricing sits considerably higher. It’s important to note that the European figure includes value-added tax, making direct price comparison complex across regions. For context, some US retailers currently list Intel’s own Arc Pro B70 models at nearly $2,000, suggesting that real-world pricing has climbed well above original launch figures across markets.
What Makes This Card Different
The ASUS UGen Arc Pro B70 shares Intel’s core Xe2 architecture but arrives with distinct engineering choices. The card packs 32GB of ECC GDDR6 memory, 32 Xe2 cores, and 256 XMX engines designed specifically for AI acceleration. Its dual-slot blower-style cooling design and 310W power rating exceed Intel’s reference model, which draws 230W. These upgrades suggest ASUS tailored the card for demanding professional workloads that benefit from improved cooling and power delivery.
The 32GB memory capacity is the defining feature here. This substantial VRAM allows developers and AI practitioners to run large language models and inference workloads directly on local machines rather than relying on cloud services. For organizations handling sensitive data, local processing offers significant advantages: greater control over proprietary information, reduced dependence on external platforms, and potentially lower long-term operational costs. ASUS actively promotes multi-GPU configurations, where users can install up to four cards in a single system, combining for 128GB of total video memory. At the announced European price, however, a four-card setup would require roughly €7,200 in GPU investment alone before accounting for the host system itself.
Who This Card Targets

Unlike consumer graphics cards marketed to gamers, the ASUS UGen Arc Pro B70 aims at professional developers, AI researchers, and businesses running computationally intensive workloads. The card serves workloads that exceed the VRAM limitations of mainstream graphics cards, making it suitable for machine learning model development, 3D rendering, video encoding, and other enterprise applications.
The price premium over Intel’s original $949 recommendation raises valid questions about value proposition. Buyers will need to weigh the ASUS card’s engineering improvements and cooling advantages against competing workstation GPUs and cloud-based AI services. Organizations considering this investment should evaluate whether local processing benefits justify the higher upfront cost compared to alternatives like more affordable graphics card options or subscription-based cloud GPU access.
What Buyers Should Consider

For professionals evaluating the ASUS UGen Arc Pro B70, several factors deserve careful attention. System architecture matters significantly. The card’s 310W power rating requires a robust power supply and proper cooling infrastructure within the host system. A workstation designed for this card needs adequate space, quality power delivery, and sufficient case ventilation to handle the heat output during extended AI workload processing.
Software ecosystem and driver support remain critical considerations. Workstation GPUs depend on mature driver stacks and framework optimization from vendors like PyTorch and TensorFlow. Buyers should research current software support status and compatibility with their specific applications before committing to this investment.
The pricing landscape also influences purchasing decisions. A €1,799 price point in Europe raises important questions about US availability and pricing. ASUS has not yet announced whether this card will reach American markets or what pricing it might carry if it does. For US buyers, this uncertainty makes timing and purchasing strategy complicated. The fact that some retailers already list Intel’s own Arc Pro B70 well above its original launch price suggests that ASUS’s pricing strategy may reflect broader market dynamics rather than excessive partner markup alone.
Data privacy and compliance considerations should influence organizational purchasing as well. For companies handling regulated information in healthcare, finance, or government sectors, the ability to process sensitive data locally rather than transmitting it to cloud services can align with compliance requirements and corporate policy. The 32GB memory capacity supports this use case effectively.
Looking Ahead
With the ASUS UGen Arc Pro B70 now reaching customers across three European nations, the question becomes whether real-world performance and software ecosystem support justify the premium pricing. Professional users will ultimately decide based on actual application performance, driver quality, and total cost of ownership. As more partner models from Intel’s Arc Pro lineup reach market, competition and pricing pressure may shift the landscape for workstation GPU buyers seeking powerful, memory-rich solutions for local AI processing and demanding professional workloads.
