Habana is a technology company.
Habana has raised $75.0M across 1 funding round.
Habana has raised $75.0M in total across 1 funding round.
Habana has raised $75.0M in total across 1 funding round.
Habana's investors include 5Y Capital, DAG Ventures, Digital Horizon VC, Koch Fund, Sequoia Capital, Two Bear Capital, Walden International, John Scull, Steve Goldberg.
Habana (Habana Labs) is an Israeli semiconductor company that builds AI accelerators — purpose‑built processors and software for training and inference of deep neural networks — and was acquired by Intel in 2020 and continues to operate as part of Intel’s data-products organization[2][3].
High‑Level Overview
Habana designs AI processors and a software stack aimed at delivering higher throughput and lower cost for data‑center AI training and inference workloads compared with general‑purpose CPUs and some GPUs[1][3]. Its product lines include Gaudi (training accelerators) and Goya/Dali (inference accelerators); Habana sells hardware plus software integrations for datacenter and cloud customers as well as systems and OEM partners[1][2]. The company’s work addresses compute bottlenecks in large‑scale deep learning deployments and has influenced the datacenter AI hardware market by providing a competitive, specialized alternative to incumbent GPU vendors[3][2].
Origin Story
Habana was founded in 2016 by Avigdor Willenz, David Dahan and Ran Halutz; the founders had prior experience in fabless semiconductor ventures and systems companies, which informed their decision to build processors optimized specifically for modern deep‑learning workloads[3][2]. The idea emerged from recognizing a performance and efficiency gap between general‑purpose datacenter chips and the needs of neural‑network training and inference; Habana launched inference and training products in 2018–2019 and attracted venture and strategic investment before being acquired by Intel in 2020 for roughly $2 billion[2][3].
Core Differentiators
Role in the Broader Tech Landscape
Habana is riding two major trends: the specialization of datacenter hardware for AI workloads, and the rapid growth of large‑scale model training and production inference needs across cloud and enterprise. Timing mattered because surging demand for transformer-scale training and cost pressures in cloud deployments created an opening for accelerator vendors with lower TCO claims[3][1]. Market forces in Habana’s favor include expanding AI workloads, operator demand for efficient inference at scale, and enterprise adoption of heterogeneous accelerator stacks. Habana’s presence pushed competitive dynamics in the AI accelerator market, giving hyperscalers and enterprises more choice beyond dominant GPU suppliers and encouraging richer software‑hardware integration practices[3][2].
Quick Take & Future Outlook
Quick reminder: Habana is now an Intel business unit that continues to develop AI accelerators (Gaudi, Goya/Dali) and supporting software for datacenter training and inference workloads[2][3][4].
Habana has raised $75.0M across 1 funding round. Most recently, it raised $75.0M Series B in November 2018.
| Date | Round | Lead Investors | Other Investors |
|---|---|---|---|
| Nov 1, 2018 | $75.0M Series B | 5Y Capital, DAG Ventures, Digital Horizon VC, Koch Fund, Sequoia Capital, Two Bear Capital, Walden International, John Scull, Steve Goldberg |