How Speedata Is Powering Next-Gen AI and Big Data With APU Chips

In the heart of Tel Aviv, a semiconductor startup is quietly redefining how enterprises crunch data. Speedata is forging ahead with a bold aim: transforming analytics infrastructure through its pioneering Analytics Processing Unit or APU.

A Breakthrough in Data Processing

Established in 2019 by a team of six researchers including early innovators in Multi-Threaded Coarse-Grained Reconfigurable Architecture (CGRA)—Speedata saw a critical market gap. Data analytics were lingering on general-purpose processors, often requiring hundreds of servers to handle complex workloads. From this realization emerged an ambitious vision: a single, purpose-built chip capable of doing much more with far less energy.

The APU Advantage: Custom Silicon for Analytics

At the core of Speedata’s innovation lies the C200 Analytics Accelerator, which houses their purpose-built APU, powered by the custom Callisto chip. Distinct from repurposed GPUs or CPUs, this ASIC is tailored from the ground up for data-intensive tasks particularly Apache Spark workloads.

Clients embedding this advanced hardware into their data centers have reported striking performance gains:

• Up to 100× faster analytics performance
• ~91% reduction in capital infrastructure, 94% footprint reduction, and 86% lower power consumption compared to equivalent GPU setups
• A standout test: a pharmaceutical analytics workload processed in just 19 minutes versus 90 hours previously, a staggering 280× speed-up

Fueling Growth: Series B Funding and Strategic Backers

In June 2025, Speedata announced a milestone $44 million Series B round, bringing its total funding to $114 million.

The round was spearheaded by returning investors like Walden Catalyst Ventures, 83North, Koch Disruptive Technologies, Pitango First, and Viola Ventures. Notably, strategic industry heavyweights joined in Lip-Bu Tan, CEO of Intel and Managing Partner at Walden Catalyst, and Eyal Waldman, co-founder and former CEO of Mellanox Technologies.

These endorsements speak volumes, reflecting confidence in the company’s unique technology and its potential to shape enterprise-grade data analytics.

Leadership Driving Innovation

Speedata’s leadership, under the guidance of newly appointed CEO Adi Gelvan, brings deep experience from high-performance data infrastructure. Gelvan previously led Speedb (acquired by Redis in 2024) and held executive roles at SQream, Infinidat, and EMC. Under his stewardship, Speedata is poised for its next growth phase.

Seamless Integration for Real-World Impact

Speedata places special emphasis on ease of adoption. Its Dash software stack integrates with Apache Spark's Catalyst optimizer, offloading heavy compute loads to the APU seamlessly without requiring customers to overhaul existing pipelines or applications.

Additionally, the company provides a Workload Analyzer tool, offering data engineers clear insight into which Spark jobs will benefit most from hardware acceleration.

Commercial Momentum & Future Directions

Speedata is swiftly gaining traction across sectors from finance and real estate to healthcare, AdTech, and pharmaceuticals—where accelerated insights are mission-critical.

With the new funding and its first-chips already taped out, expectations are high for a product launch at major industry events, such as the anticipated Databricks Data & AI Summit in mid-2025. From there, expansion into broader analytics platforms is on the roadmap.

Why Speedata Matters

Speedata signals a fundamental shift in how enterprises handle data at scale. By delivering tailored hardware specifically for analytics, the company promises not only speed but also efficiency and sustainability. In a world where AI and big data increasingly define competitive differentiation, Speedata’s innovations may well become the unseen engine powering tomorrow’s breakthroughs.