SAN FRANCISCO--(BUSINESS WIRE)--HEAVY.AI, the leader in GPU-accelerated analytics, today announced the general availability of the HEAVY.AI analytics platform with the NVIDIA GH200 Grace Hopper Superchip, an advanced hardware architecture that features an NVIDIA-designed Arm-based CPU with an ultra-fast NVIDIA Hopper GPU. The launch is part of the broader 8.2 release of the HEAVY.AI accelerated analytics platform.
Support for running HEAVY.AI on the Grace Hopper Superchip means that users of the platform will be able to process faster and cheaper than previously possible. By leveraging the ultra-fast NVIDIA NVLink-C2C interconnect between the CPU and GPU, which features 900GB/sec of bidirectional bandwidth, data can be transferred between CPU and GPU at speeds of up to 7X faster than traditional PCIe-based systems. This means that users of HEAVY.AI can now query and visualize massive datasets that exceed GPU memory capacity at interactive speeds. Furthermore, users of HEAVY.AI will be able to also take advantage of the NVIDIA GB200 Grace Blackwell Superchip and GB200 NVL72, a liquid-cooled, rack-scale solution that boasts a 72-GPU NVLink domain that acts as a single massive GPU and delivers 30X faster real-time trillion-parameter LLM inference.
The release of the HEAVY.AI platform for the Grace Hopper and Blackwell architectures also offers significant cost savings for customers, who now can now scale to larger datasets with less GPU resources. As an illustration of these cost savings, with this release, HEAVY.AI has launched a publicly available demo featuring over 20 billion records of ship locations (AIS data) in US coastal waters spanning the last 7 years, running on a single NVIDIA GH200 Superchip on Vultr Cloud. Previously, this demo would have required at least 4 NVIDIA A100 GPUs with 320GB of combined GPU VRAM to cache the relevant data and ensure interactive performance, representing nearly 70% hardware cost savings.
“The NVIDIA Grace Hopper Superchip changes the game in terms of being able to provide best-in-class performance over our customers’ largest datasets,” said Todd Mostak, Co-Founder and CEO of HEAVY.AI. “Now customers no longer have to choose between faster performance and lower cost; with HEAVY.AI running on NVIDIA Grace systems, they can have both.”
“Customers worldwide are looking to boost performance and reduce cost when analyzing large datasets,” said Ivan Goldwasser, director of data center CPUs at NVIDIA. “By accelerating HEAVY.AI’s analytics platform with NVIDIA Grace Hopper and Grace Blackwell Superchips, customers can speed up high-performance data processing and visualization for big data analytics.”
Underscoring the performance and cost benefits of the Grace Hopper architecture, HEAVY.AI recently released benchmarks comparing the performance of the HeavyDB GPU-accelerated database on the industry standard TPC-H SQL Data Warehouse benchmark against three of the most popular CPU-based data warehouses. Running on a NVIDIA GH200 Grace Hopper system, HeavyDB was up to 21X faster on average than its CPU-based competitors while being up to 9X cheaper per hour to operate. More details from the benchmark can be found here.
The release of the HEAVY.AI analytics platform on NVIDIA Grace architecture is part of a larger effort by the company to further increase the speed and cost advantages of the platform compared to CPU-based data warehouses and analytics systems. “Our overarching goal is to make the performance and cost advantages of GPU-accelerated analytics so compelling that it becomes the default means of analyzing large datasets,” said Mostak.
About HEAVY.AI
HEAVY.AI provides a groundbreaking GPU-accelerated analytics platform that empowers organizations to instantly query and visualize multi-billion-record datasets, including geospatial and time-series data, delivering a complete view of what is happening, when, and where. By integrating massive data volumes from multiple sources, HEAVY.AI provides an immersive, real-time interactive visual analytics experience. Industry leaders in telecommunications, energy, government, utilities, and higher education rely on HEAVY.AI to drive high-impact decisions at unprecedented speeds. Born from research at Harvard and MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), HEAVY.AI is backed by GV, In-Q-Tel, New Enterprise Associates (NEA), NVIDIA, Tiger Global Management, Vanedge Capital, and Verizon Ventures. Headquartered in San Francisco, HEAVY.AI is reshaping data analytics. Learn more about HEAVY.AI.
