Speedata has announced a strategic partnership with Nebul under which Nebul will be the first cloud provider to deploy Speedata’s Analytics Processing Unit (APU) in its European sovereign cloud infrastructure. The companies say the deployment targets European enterprises running advanced analytics and AI data-processing workloads at scale, while keeping data under European legal jurisdiction.
Speedata positions the APU as a processor purpose-built for big data and advanced analytics. It is a hardware-optimized application-specific integrated circuit (ASIC) that executes Apache Spark SQL natively, running complex queries, joins, aggregations, and transformations directly in silicon rather than memory. Speedata reports that, in standard benchmarks, the APU has demonstrated up to 100x performance gains over CPUs and GPUs, and that it works alongside CPUs and GPUs by addressing “the data layer bottleneck,” with claimed gains in time-to-insights, energy efficiency, and cost.
For data center and platform teams, the companies frame the initial value around Apache Spark acceleration and infrastructure reduction. “Running advanced analytics at scale has always been extremely costly. The APU changes that equation,” said Adi Gelvan, CEO of Speedata. “In one enterprise AI data preprocessing deployment, customers replaced 38 servers with just 3, achieving over 90% cost reduction. With Speedata on Nebul’s sovereign cloud, European enterprises can lower their infrastructure and operational costs while keeping their data under European governance.”
Speedata also says the APU extends beyond batch extract, transform, load (ETL) into AI-oriented pipelines, including AI data preparation and cleaning, and real-time query-oriented retrieval-augmented generation (RAG) and table-oriented generation (TAG). The company says this is intended to bring accelerated structured and table data to large language models (LLMs).
Nebul says it will host the APU within its European data center network and offer it as part of its Data Platform, positioning it as an advanced analytics component alongside its existing AI infrastructure. “Sovereignty isn’t a value statement; it’s an operational requirement,” said Arnold Juffer, CEO of Nebul. “True data sovereignty extends beyond GDPR compliance and EU-based data centers to encompass ownership structure, operational control, and jurisdictional authority. Our customers need to know who owns their infrastructure, who operates it, and who can be compelled to access it. With Speedata, we can now add high-performance data processing performance to that offer.” The companies say the partnership will initially focus on Europe, the Middle East, and Africa (EMEA) enterprise customers in sectors where analytics performance and data governance are critical.
Source: Speedata
