Habana Labs
Gaudi and Goya AI processors for data center training and inference
- Sector
- Semiconductors
- Federal grants
- None on recordNo awards found in SBIR/STTR, NIH, NSF or DoD records.
- Private capital raised
- None on recordNo Form D filings on SEC EDGAR. Some raises are not filed.
- Federal contracts
- None on recordNo federal contract awards found.
- Granted patents
- 23
- 6 pending · first 2019
- Rank in peer group#36 of 224
Capital timeline
No federal awards or Form D filings are on record for this company.
Many early companies raise without filing a Form D, or have not yet won a federal award. New records appear here as the sources update.
Checked: SEC EDGAR Form D through Oct 23, 2026; SBIR and STTR awards through Oct 21, 2026; USAspending contract awards
Founder? Add funding contextStanding among 224 peers
Line = medianSemiconductors · Device
- Granted patents#36 · top 16%
Cumulative measures favor older companies. Rankings describe position today, not outcome. Below the median, ranks show as a band rather than a number.
Founders and team
No founders are linked to this company yet.
Founders appear once a person is matched to the company through its own patents, grants or filings. We never infer people from social profiles.
Checked: USPTO inventors; federal award principal investigators; SEC Form D related persons · Oct 23, 2026
Claim this profile to add your teamWhat it has protected
USPTO · 23 granted of 29- GrantedAccelerated Sparse Neural Network Training using Transposable Pruning MasksUSPTOUS 12,688,421
- GrantedIn-network Computing With Reduction Operation By SwitchUSPTOUS 12,537,771
- GrantedDynamic Cache Allocation In Artificial Intelligence AcceleratorUSPTOUS 12,475,054
Show all 29 patent records
- Accelerated Sparse Neural Network Training using Transposable Pruning MasksJul 21, 2026 · US 12,688,421Granted
- In-network Computing With Reduction Operation By SwitchJan 27, 2026 · US 12,537,771Granted
- Dynamic Cache Allocation In Artificial Intelligence AcceleratorNov 18, 2025 · US 12,475,054Granted
- Performance Modeling and Analysis of Artificial Intelligence (AI) Accelerator ArchitecturesSep 30, 2025 · US 12,430,566Granted
- Machine learning with variable-shape tensorsSep 2, 2025 · US 12,406,204Granted
- Rendezvous Flow With Rdma (remote Direct Memory Access) Write ExchangeNov 12, 2024 · US 12,141,093Granted
- Cache Memory With Decoupled Control PathsAug 20, 2024 · US 12,066,936Granted
- Compressing Neural Networks Through Unbiased Minimum Variance PruningAug 8, 2024 · Application 18164875Pending
- Low latency execution of a machine learning modelDec 19, 2023 · US 11,847,491Granted
- Fine-grained pipelining using index space mappingAug 1, 2023 · US 11,714,653Granted
- Mediating between asynchronous clock domains while preventing false indications of FIFO occupancyDec 20, 2022 · US 11,532,338Granted
- Activation Function Approximation in Deep Neural Networks using Rectified-Linear-Unit FunctionOct 11, 2022 · US 11,468,147Granted
- Index Space Mapping Using Static Code AnalysisOct 11, 2022 · US 11,467,827Granted
- Remote Direct Memory Access (rdma) Transfer Of An Information Unit Over An Ethernet NetworkJun 23, 2022 · Application 17125984Pending
- Tensor-based Memory AccessMay 3, 2022 · US 11,321,092Granted
- Processing-Memory Architectures Performing Atomic Read-Modify-Write Operations in Deep Learning SystemsFeb 15, 2022 · US 11,249,724Granted
- Method For Remote Direct Memory Access (rdma) Congestion Control Over An Ethernet NetworkFeb 1, 2022 · US 11,240,162Granted
- Hardware accelerator for systolic matrix multiplicationFeb 9, 2021 · US 10,915,297Granted
- Approximation of Mathematical Functions in a Vector ProcessorFeb 9, 2021 · US 10,915,494Granted
- Processor Suspension Buffer and Instruction QueueDec 1, 2020 · US 10,853,070Granted
- Hiding latency of multiplier-accumulator using partial resultsDec 1, 2020 · US 10,853,448Granted
- Hardware accelerator for outer-product matrix multiplicationJul 14, 2020 · US 10,713,214Granted
- Large-scale Computations Using An Adaptive Numerical FormatNov 26, 2019 · US 10,491,239Granted
- Data Compression Scheme Utilizing A Repetitive Value Within The Data StreamNov 26, 2019 · US 10,491,241Granted
- Matrix Multiplication EngineNov 26, 2019 · US 10,489,479Granted
- Pipelining between processorsApplication 63008845Pending
- Switch In-network ComputingApplication 63386051Pending
- Rendezvous Flow With Rdma (remote Direct Memory Access) Write ExchangeApplication 63292767Pending
- Accelerated Sparse Neural Training: A Provable and Efficient Method to Find N:M Transposable MasksApplication 63155323Pending
Related companies, and why
- M/A-COM Technology Solutions Holdings, Inc.Similar federal contract value in Semiconductors · null in contracts
- AMS AGSimilar federal contract value in Semiconductors · null in contracts
- Phononic Devices, Inc.Similar federal contract value in Semiconductors · null in contracts
- Movandi CorpSimilar federal contract value in Semiconductors · null in contracts
- Cornelis Networks, Inc.Similar federal contract value in Semiconductors · null in contracts
- MatX Inc.Similar federal contract value in Semiconductors · null in contracts
Sources
Federal grants from SBIR/STTR, NIH, NSF and DoD records. Federal contracts from USAspending.gov. Private capital from SEC Form D. Patents from USPTO. Publications from OpenAlex. Regulatory activity from the Federal Register. Press matched on company name.
How to read this page
Figures show associations in public records, not causes or predictions. Nothing here is investment advice. An em-dash means no record was found under this company's name, not that none exists.
Data freshness
SEC Form D filings through . SBIR and STTR awards through .