Skip to main content

NeuroBlade

Category: AI Infrastructure

NeuroBlade builds purpose-built hardware (SQL Processing Unit) to massively accelerate data analytics query processing for large-scale enterprise data. NeuroBlade was founded in 2018. The company is led by Elad Sity. Based in Tel Aviv, Israel. Team size: 101-500. Total funding raised: $110M. Latest round: Series B. Key investors include Corner Ventures, Intel Capital, StageOne Ventures, Grove Ventures, Marius Nacht, UMC Capital, Marubeni, MediaTek.

Founded
2018
Headquarters
Tel Aviv, Israel
Team size
101-500
Total funding
$110M

Value proposition

Delivers 100x or greater speedup for data analytics query processing at lower total cost of ownership by offloading SQL query execution to a custom-engineered SPU processor.

Products and solutions

SQL Processing Unit (SPU) architecture, deployable via cloud EC2 instances or on-premises PCIe cards, DAXL developer platform, Hardware Enhanced Query System (HEQS).

Unique value

Purpose-built silicon (SPU) for SQL/analytics acceleration vs. general-purpose CPUs, claiming order-of-magnitude throughput and TCO improvements while preserving existing infrastructure investments.

Target customer

Enterprises and data platform providers running large-scale analytics workloads (cloud-native and on-premises), integrating with frameworks like Spark, Presto, ClickHouse.

Industries served

Data analytics, cloud infrastructure, big data/database acceleration, semiconductor/hardware

Technology advantage

Custom semiconductor architecture (SPU) specifically engineered for query execution rather than general compute; deep integration with open-source big data engines (Spark, Presto, ClickHouse, Velox).

How they differentiate

Focus on purpose-built hardware acceleration (SPU) specifically for SQL/analytics workloads rather than general-purpose AI accelerators or software-only optimization.

Main competitors

MangoBoost, Speedata, Fungible, also referenced: Hazelcast, H2O.ai, GigaSpaces (broader analytics acceleration space)

Key partnerships

AWS (SPU on EC2 F2 instances; core engineering team joined Annapurna Labs Oct 2025), Dell Technologies (SPU on select PowerEdge servers, May 2023), Meta Velox open-source community (HW acceleration for Presto/Velox, 2023), KIOXIA (NVMe SSD integration for HEQS), Intel Capital (strategic investor)

Major milestones

2018: Founded (Tel Aviv). 2019-06: $23M Series A (Marius Nacht lead; total ~$27.5M). 2021-10: $83M Series B led by Corner Ventures (total $110M). 2023-05: SPU G200 + Dell PowerEdge partnership. 2023: Meta Velox HW acceleration collaboration. 2024-12: SPU available on Amazon EC2 F2 instances. 2025-10: Core engineering team joins AWS Annapurna Labs (quiet exit/acquihire; end of independent ops per CTech); website remains up for product/support info.

Growth metrics

Raised $110M total (through Series B Oct 2021). Peak ~100–120+ employees (~2021–2023); Oct 2025 AWS Annapurna Labs deal moved core engineering team (dozens) off independent headcount — independent commercial scale effectively ended.

Market positioning

Niche hardware-accelerated data analytics infrastructure provider targeting enterprise and cloud data platform performance/cost optimization.

Geographic focus

Global, with primary operations in Israel (Tel Aviv) and US (Palo Alto/Mountain View).

Patents and IP

Patents assigned to NeuroBlade, Ltd. (e.g. US11126511 — distributed processors / compile methods); Crunchbase/IPqwery reports ~30 registered patents primarily in computing/calculating category.

About Elad Sity

VP of Embedded SW and System Integration at SolarEdge (Nasdaq: SEDG); co-founded NeuroBlade in 2018.

Latest news about NeuroBlade

More AI Infrastructure companies

Official website: