Sapphire launches rack-optimized Radeon AI PRO R9700 graphics card

Sapphire targets edge computing and visualization with its Radeon AI PRO R9700, a rack-friendly professional graphics card designed to handle intensive Artificial Intelligence workloads.

Sapphire has announced its new Radeon AI PRO R9700 graphics card, aiming squarely at professionals in visualization and edge Artificial Intelligence acceleration. The design closely mirrors a recent ASUS variant, showcasing a form factor tailored for rack-friendly deployment: at just 26.6 cm long, 2 slots wide, and 11.1 cm tall, the card maintains a full-height design that fits seamlessly into 3U rackmount enclosures, a frequent choice in dense workstation environments.

The Radeon AI PRO R9700 stands out thanks to its focus on airflow management and secure installation. A lateral airflow blower cools the card, while precision cutouts in both the PCB and cooler shroud facilitate air intake for adjacent cards—an essential feature when up to four units operate side by side. Additional enterprise features include a rear-mounted 12V-2×6 power connector, simplifying cable routing and reducing clutter in tightly packed racks, and rear bracket mounting holes for reliable chassis attachment using expanders.

Under the hood, the card leverages a maxed-out 4 nm ´Navi 48´ GPU, the same core found in the Radeon RX 9070 XT desktop model. However, Sapphire´s professional version doubles the available memory to a beefy 32 GB, intended to support larger large language models (LLMs) and more demanding Artificial Intelligence tasks. The R9700 promises substantial computing performance, achieving up to 1,531 trillion operations per second (TOPS) at INT4 precision and reaching 95.7 teraflops per second (TFLOP/s) for FP16 workloads. By combining robust hardware specifications with rackmount optimization, Sapphire´s latest card addresses the growing demand for scalable Artificial Intelligence solutions in data centers and professional visualization markets.

56

Impact Score

Seedance video models launch on Atlas Cloud with tiered pricing

Atlas Cloud has added ByteDance’s Seedance video generation family to its model platform, offering Lite and Pro text-to-video and image-to-video options across 480p, 720p, and 1080p resolutions with transparent per‑second pricing. The integration targets production use cases with low latency, unified APIs, and enterprise‑grade reliability and compliance.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.