Developer docs
API playgroundTry for free, no card

Search company profiles

Esperanto Technologies

Full company profile

uuid0000gtf

Namestring
Esperanto Technologies
Legal namestring
Esperanto Technologies Inc.
Websiteurl
esperanto.ai
Company typeenum
Private
Founded yearint
2014
Descriptiontext

Esperanto Technologies was a privately held semiconductor company founded in 2014 and headquartered in Mountain View, California, that designed and commercialized RISC-V-based massively parallel AI and HPC accelerators. Its core product, the ET-SoC-1 chip fabricated on TSMC 7nm, integrated 1,088 energy-efficient ET-Minion 64-bit in-order cores with vector/tensor extensions alongside four ET-Maxion out-of-order cores running SMP Linux, delivering PetaOps-scale inference performance at sub-30W typical power. The company wrapped the silicon in a four-tier product line: the bare ET-SoC-1 chip, a certified PCIe Gen4 card (co-developed with Penguin Solutions), a 2U AI/HPC compute server with up to 16 cards and 16,000 RISC-V cores, and the Generative AI Appliance bundling hardware with the Primo AI/ML software platform and pre-loaded foundation models including LLaMA 2, Vicuna, StarCoder, Stable Diffusion, and OpenJourney.

Esperanto monetized primarily through one-time hardware sales of these accelerator products, complemented by a paid Esperanto Access Program (EAP) for evaluation, with software (Primo, EtGlowExecutionProvider for ONNXRuntime) sold as part of integrated solutions rather than as standalone licenses. Go-to-market combined direct enterprise field sales with a global channel network - Penguin Solutions worldwide, plus regional value-added partners E4 Computer Engineering (Italy), MEGWARE (Germany), and Elematec (Japan) - targeting data centers, HPC facilities, and enterprise edge customers with a hybrid segmentation across use cases (LLM/Generative AI inference, transformers, computer vision, recommendation systems) and verticals (healthcare/life sciences, robotics, supply chain). Strategic partnerships with NEC Corporation (next-generation HPC RISC-V chips) and Rapidus (2nm foundry) anchored a forward roadmap. The company ceased operations in November 2025, with its intellectual property acquired by Ainekko (Nekko.ai).

Short descriptiontext

Esperanto Technologies was a Mountain View-based RISC-V AI/HPC chip company that designed the ET-SoC-1, a TSMC 7nm accelerator with over 1,000 RISC-V cores under 30W, selling accelerator chips, PCIe cards, compute servers, and a Generative AI Appliance to enterprise data centers and HPC customers globally before ceasing operations in November 2025 with its IP acquired by Ainekko.

Operating statusenum
Closed
Ownership categoryenum
Headcount rangeband
11–50
akta.pro rankint
HeadquartersMountain View, United States
HQ citystring
Mountain View
HQ countrystring
United States
HQ regionstring
North America
Markets served

Serves global market

Offices5 records

Each record includes

City, Country, Type, Description, Source

Keyword5 values
RISC-V processors, AI inference accelerators, HPC compute servers, Generative AI appliances, Energy efficient chips
Industry4 codes
PRIMARYHDAHABAC
Compute & Acceleration ICs (CPU/GPU/AI/FPGA/ASIC)
SECONDARYHDAHAJAF
AI/Neural Processing Units (NPUs) & ML Accelerators
SECONDARYHDAHAAAB
Advanced SoCs & Application Processors
SECONDARYHDAAAAAL
AI Hardware IP & Chiplet/Packaging for Accelerators (EDA/IP, advanced packaging)
Hardware, Technology & Telecom/Artificial Intelligence & Machine Learning/AI Infrastructure & Compute (GPUs/ASICs, accelerators)
NAICS code2 codes
  • Computer and Peripheral Equipment Manufacturing33411
  • Computer Systems Design Services541512
SIC code2 codes
  • Electronic Computers3571
  • Services-Computer Integrated Systems Design7373
Product category
AI Accelerator Semiconductors
GTM motion1 record

Each record includes

Type, Description, Source

Revenue model3 records
1Hardware Sales
TypeHardware Sales
Description

Sale of AI/HPC accelerator chips (ET-SoC-1), PCIe cards, compute servers, and Generative AI Appliances. Products available for purchase or secure remote cloud access.

esperanto.ai
2Software Licensing
TypeLicensing Royalties
Description

Primo AI/ML software platform sold as part of integrated solutions with hardware, including model development SDK, optimization tools (AWQ quantization, LoRA fine-tuning), and ONNX Runtime integration.

esperanto.ai
3Evaluation Program
TypeSubscription Recurring
Description

Paid Esperanto Access Program (EAP) providing evaluation access to ET-SoC-1 through remote and on-premise access. Fees credited against initial product orders.

esperanto.ai
Marketing channels6 records

Each record includes

Title, Type, Stage, Description, Source

Distribution channels5 records

Each record includes

Title, Type, Scope, Target buyer, Description, Source

Cost components6 values
Technology or R&D, Personnel, Supply Chain, Operations, Marketing or Sales, Infrastructure
Pricing details5 tiers
1ET-SoC-1 AI/HPC Accelerator Chip - single chip for system integrators
ModelOne time/ perpetual licenseBilling cadenceMulti-year contract
Notes

Quote-based pricing. Available directly from Esperanto and through partner Penguin Solutions.

esperanto.ai
2ET-SoC-1 PCIe Card - compact low-profile PCIe Gen 4 form factor with 32GB LPDDR4x DRAM
ModelOne time/ perpetual licenseBilling cadenceMulti-year contract
Notes

Jointly developed with Penguin Solutions for production quality and international certification.

esperanto.ai
3AI/HPC Compute Server - 2U rack-mounted with 8 or 16 ET-SoC-1 cards, up to 16,000 RISC-V cores
ModelOne time/ perpetual licenseBilling cadenceMulti-year contract
Notes

Standard 2U 19 inch rack-mounted chassis with Intel Xeon host processors and up to 1TB DDR4-3200 memory.

esperanto.ai
4Generative AI Appliance - full solution including hardware, LLM software stack and foundation models
ModelOne time/ perpetual licenseBilling cadenceMulti-year contract
Notes

Currently running LLaMA 2, Vicuna, StarCoder, OpenJourney and Stable Diffusion. Includes 4 ET-SoC-1 cards.

esperanto.ai
5Esperanto Access Program (EAP) - evaluation program for ET-SoC-1
ModelSubscriptionBilling cadenceAnnual
Notes

Paid evaluation program with remote and on-premise access options. Fees credited against initial orders of Esperanto products.

esperanto.ai
GTM typeB2B
B2B
Offering typeHardware or Manufacturing
Hardware or Manufacturing
Brand1 of 5 records shown
1ET-SoC-1
Description

The world's most efficient RISC-V commercial chip with over 1,000 RISC-V processors on a single TSMC 7nm chip.

esperanto.ai
+4 more records
Core offering1 text field

Esperanto Technologies developed and sold RISC-V based massively parallel AI/HPC accelerator hardware and software. The flagship ET-SoC-1 chip integrates 1,088 64-bit RISC-V in-order cores (ET-Minion) with custom vector/tensor units plus 4 out-of-order RISC-V cores (ET-Maxion) on TSMC 7nm, delivered as chips, PCIe Gen4 cards, 2U compute servers, and a Generative AI Appliance. The Primo AI/ML software platform and EtGlowExecutionProvider (ONNXRuntime backend) accompany the hardware for deploying generative AI, transformer, computer vision, and recommendation models.

Differentiator
Functional benefit
Problem solved
Quantifiable outcome1 of 3 values shown
  • 8.26x speedup compared to CPU execution (Intel Xeon Platinum 8358) for small language model inference
+2 more records
Product overview1 text field

Esperanto Technologies delivered a portfolio of RISC-V-based AI/HPC products centered on the ET-SoC-1 chip — a TSMC 7nm accelerator featuring over 1 thousand RISC-V cores on a single die — and complementary software. The product line comprised four tiers: (1) the ET-SoC-1 chip itself as the foundational AI/HPC accelerator; (2) the ET-SoC-1 PCIe Card form factor (developed with Penguin Solutions) for system integration; (3) the Esperanto AI/HPC Compute Server and Generative AI Appliance, which packaged 4–16 ET-SoC-1 cards into 2U rack-mounted systems with host Xeon processors; and (4) the Primo AI/ML software platform and EtGlowExecutionProvider for ONNXRuntime-based model deployment. The company ceased operations in 2025; its IP was subsequently acquired by Ainekko (also referenced as Nekko.ai), which launched the open-source AI Foundry platform. As of the data cutoff, product availability was listed on esperanto.ai but the company announced it had ceased operations.

Product and service6 records
1ET-SoC-1 AI/HPC Accelerator Chip
CategoryAI/HPC accelerator silicon
Description

TSMC 7nm inference chip with 1,088 ET-Minion 64-bit RISC-V in-order cores (each with custom vector/tensor units) and 4 ET-Maxion out-of-order cores, over 160 million bytes of on-chip SRAM, PCIe Gen4 x8, and sub-30W typical power consumption. Designed for system integrators building AI inference and HPC solutions.

2ET-SoC-1 PCIe Card
CategoryAI/HPC accelerator hardware
Description

Low-profile PCIe Gen 4 card with a single ET-SoC-1 chip, 32 GB LPDDR4x DRAM, full monitoring sensors, and management ports. Up to 16 cards per system. Jointly developed and certified with Penguin Solutions for deployment in North America, EU, UK, South Korea, and Japan.

3Esperanto AI/HPC Compute Server
CategoryAI/HPC compute server
Description

2U 19-inch rack-mounted datacenter server with up to 16 ET-SoC-1 PCIe cards (up to 16,000 RISC-V cores), dual Intel Xeon host processors (16-core 6326 or 32-core 8358P), up to 1 TB DDR4-3200 memory, and Esperanto ML/GPU SDKs. Supports generative AI, transformer, computer vision, and recommendation model inference.

4Esperanto Generative AI Appliance
CategoryGenerative AI appliance
Description

Industry's first RISC-V-based Generative AI Appliance integrating 4 ET-SoC-1 cards with the Primo software stack and foundation models (LLaMA 2, Vicuna, StarCoder, OpenJourney, Stable Diffusion). Full solution for fine-tuning vertical business applications with high data privacy and lower TCO vs. GPU-based systems.

5Primo AI/ML Platform
CategoryAI/ML software platform
Description

Advanced AI software platform for optimizing generative AI model inference at scale. Includes model optimization, fine-tuning, RAG-based inference, and orchestration across complex model pipelines, plus an AI/ML Model Development SDK with AWQ quantization, LoRA fine-tuning, Flash Attention, and ML compilation tools.

6Esperanto Access Program (EAP)
CategoryEvaluation service
Description

Paid evaluation program for ET-SoC-1 with remote and on-premise access options, enabling evaluation of inference on recommendation, visual analysis, and transformer-based models. Annual subscription with fees credited against initial product orders and higher supply priority for participants.

Scale indicator5 records

Each record includes

Type, Value, Description, Source

Partnership7 partners
Strategic tierCoreTypeStrategic or Co-development PartnerAnnounced on2024-11-18
Description

Cooperation with NEC Corporation on roadmap for next-generation chips and software for HPC computing. Leverages NEC's deep experience in HPC supercomputer design and software stacks combined with Esperanto's energy-efficient RISC-V multiprocessor technology to enable next generation supercomputers.

Strategic tierStrategicTypeStrategic or Co-development PartnerAnnounced on2024-05-15
Description

Partnership with Rapidus to enable more energy-efficient designs for AI inference and HPC workloads for the 'post-GPU era'. Rapidus is developing 2nm semiconductor manufacturing capability in Japan with pilot production line planned for April 2025 and mass production in 2027.

Strategic tierCoreTypeStrategic or Co-development PartnerAnnounced on2023-05-22
Description

Strategic partnership with Penguin Solutions (SGH brand, Nasdaq: SGH) to collaboratively develop multiple systems in various form factors for AI and HPC acceleration in datacenter and edge environments. Jointly developed, tested and certified ET-SoC-1 PCIe cards with safety and emissions certifications for North America, EU, UK, South Korea and Japan.

Strategic tierRegionalTypeChannel Partner/ Reseller/ DistributorAnnounced on2023-05-22
Description

Value-added partnership in Italy for developing AI and HPC acceleration systems for sophisticated enterprises, universities and centers for advanced research and computing.

Strategic tierRegionalTypeChannel Partner/ Reseller/ DistributorAnnounced on2023-05-22
Description

Value-added partnership in Germany for bringing RISC-V AI and HPC acceleration solutions to research institutes, universities, and industrial and business customers throughout Europe.

6Elematec Corporation
Strategic tierRegionalTypeChannel Partner/ Reseller/ DistributorAnnounced on2023-05-22
Description

Value-added partnership in Japan for RISC-V AI acceleration opportunities across digital electronics, automotive industry and industrial and home segments.

esperanto.ai
Strategic tierMinorTypeStrategic or Co-development Partner
Description

Esperanto's massively-parallel AI accelerators added to IFS ecosystem. Combination of Esperanto's low-power design techniques and IFS advanced silicon and packaging technologies enables thousands of RISC-V based processors to operate in a single package.

Recent move8 records

Each record includes

Date, Type, Title, Description, Source

Expansion highlight7 records

Each record includes

Type, Description

Peers10 records
TypeDirect peer
Description

Graphcore develops IPU accelerators for AI training and inference, targeting data center and enterprise customers. It is comparable to Esperanto in targeting energy-efficient AI acceleration as an alternative to GPU-based systems.

TypeDirect peer
Description

SiFive is a fabless RISC-V IP and core designer that licenses RISC-V-based CPU IP to chip makers. It overlaps with Esperanto on the RISC-V instruction set, design tooling, and target customer base of custom silicon developers, though it is more focused on IP licensing rather than AI accelerator SoCs.

3Tenstorrent
TypeDirect peer
Description

Tenstorrent designs AI accelerator chips and RISC-V-based processors for data center and edge inference. It directly competes with Esperanto's ET-SoC-1 in the AI accelerator market, with comparable RISC-V DNA and a focus on energy-efficient inference hardware.

4Rivos
TypeEmerging player
Description

Rivos is a fabless startup building RISC-V-based server-class processors targeting data center and AI workloads. Like Esperanto, it pursues a RISC-V-based architecture aimed at energy-efficient compute, though it is at an earlier stage of commercialization.

5Ventana Micro Systems
TypeEmerging player
Description

Ventana designs RISC-V-based data center CPUs and accelerator chiplets. It overlaps with Esperanto on the RISC-V compute fabric and enterprise data center target market, focusing on high-performance RISC-V cores for cloud and AI.

TypeDirect peer
Description

Cerebras builds wafer-scale AI accelerator systems for training and inference in data centers. It competes head-to-head with Esperanto's compute server in the AI accelerator market with a similar focus on enterprise and HPC customers.

TypeDirect peer
Description

SambaNova develops AI accelerator systems and RDU chips targeting data center inference and training. It is a direct competitor to Esperanto in the energy-efficient AI accelerator category, with a comparable enterprise go-to-market motion.

TypeDirect peer
Description

Groq builds LPU inference accelerators optimized for low-latency generative AI serving. It competes with Esperanto in the AI inference accelerator market with a similar value proposition of fast, energy-efficient inference for LLMs.

TypeBroad incumbent
Description

NVIDIA is the dominant GPU and AI accelerator vendor (H100, Blackwell) and the de facto incumbent that Esperanto's energy-efficiency story is positioned against. It defines the performance and software ecosystem benchmark for AI inference.

TypeOthers
Description

Penguin Solutions (SGH brand) is Esperanto's strategic systems integration partner and a system integrator for HPC and AI workloads. It is comparable as a system-level channel and reference design partner in the same data center AI hardware ecosystem.

Market position
Strengths5 records

Each record includes

Headline, Details, Source

Weaknesses5 records

Each record includes

Headline, Details, Source

Competitive moat4 records

Each record includes

Type, Details

Key risks6 records

Each record includes

Headline, Details, Source

Key highlights7 records

Each record includes

Headline, Details, Source

Customer concentration

Classification, Details

Named customers1 record

Each record includes

Name, Industry, Type, Use case, Source, UUID

Segment3 records

Each record includes

Title, Type, Primary, Description, Pain point addressed, Use case, Source

Ideal customer profile3 records

Each record includes

Profile, Firmographic size, Sales motion, Sales cycle length, Buying structure, Purchase trigger, Buyer persona, Geography, Industry vertical, Primary use case, Description, Pain points, Evidence proof points, Target buyer

Technology focused
Yes
API detail
Has APIbool
No

Docs URL, Description

Integration18 records

Each record includes

Title, Type, Description, Source

AI capability15 records

Each record includes

Type, Description, Source

AI maturity
App detail

Has app

Feature5 records

Each record includes

Title, Differentiator, Description, Source

Core technology
Revenue estimate
Valuation estimate
Number of profiles
Profiles6 records

Each record includes

Name, Designation, Designation category, Overview, Profile commentary, Source

No data
Compliance6 records

Each record includes

Name, Class, Description

Funding overview

Funding stage, Last funding date, Total funding USD

Funding rounds3 records

Each record includes

Round, Amount USD, Date, Pre money valuation, Total investors, Investors, News

Investors7 records

Each record includes

Name, Type, Date of entry, Rounds participated, Website

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Each record includes

Name, Acquisition type, Announced date, Completed date, Status, Website, News

Investment

Each record includes

Name, Round, Announced date, Lead investor, Website, News

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Esperanto Technologies

AI Accelerator Semiconductorsesperanto.ai

Esperanto Technologies was a Mountain View-based RISC-V AI/HPC chip company that designed the ET-SoC-1, a TSMC 7nm accelerator with over 1,000 RISC-V cores under 30W, selling accelerator chips, PCIe cards, compute servers, and a Generative AI Appliance to enterprise data centers and HPC customers globally before ceasing operations in November 2025 with its IP acquired by Ainekko.

What Esperanto Technologies does

Esperanto Technologies was a privately held semiconductor company founded in 2014 and headquartered in Mountain View, California, that designed and commercialized RISC-V-based massively parallel AI and HPC accelerators. Its core product, the ET-SoC-1 chip fabricated on TSMC 7nm, integrated 1,088 energy-efficient ET-Minion 64-bit in-order cores with vector/tensor extensions alongside four ET-Maxion out-of-order cores running SMP Linux, delivering PetaOps-scale inference performance at sub-30W typical power. The company wrapped the silicon in a four-tier product line: the bare ET-SoC-1 chip, a certified PCIe Gen4 card (co-developed with Penguin Solutions), a 2U AI/HPC compute server with up to 16 cards and 16,000 RISC-V cores, and the Generative AI Appliance bundling hardware with the Primo AI/ML software platform and pre-loaded foundation models including LLaMA 2, Vicuna, StarCoder, Stable Diffusion, and OpenJourney.

Esperanto monetized primarily through one-time hardware sales of these accelerator products, complemented by a paid Esperanto Access Program (EAP) for evaluation, with software (Primo, EtGlowExecutionProvider for ONNXRuntime) sold as part of integrated solutions rather than as standalone licenses. Go-to-market combined direct enterprise field sales with a global channel network - Penguin Solutions worldwide, plus regional value-added partners E4 Computer Engineering (Italy), MEGWARE (Germany), and Elematec (Japan) - targeting data centers, HPC facilities, and enterprise edge customers with a hybrid segmentation across use cases (LLM/Generative AI inference, transformers, computer vision, recommendation systems) and verticals (healthcare/life sciences, robotics, supply chain). Strategic partnerships with NEC Corporation (next-generation HPC RISC-V chips) and Rapidus (2nm foundry) anchored a forward roadmap. The company ceased operations in November 2025, with its intellectual property acquired by Ainekko (Nekko.ai).

Esperanto Technologies firmographics

Firmographics
Name
Esperanto Technologies
Legal name
Esperanto Technologies Inc.
Website
https://esperanto.ai
Company type
Private
Founded year
2014
Operating status
Closed
Headcount range
11–50 employees
Short description
Esperanto Technologies was a Mountain View-based RISC-V AI/HPC chip company that designed the ET-SoC-1, a TSMC 7nm accelerator with over 1,000 RISC-V cores under 30W, selling accelerator chips, PCIe cards, compute servers, and a Generative AI Appliance to enterprise data centers and HPC customers globally before ceasing operations in November 2025 with its IP acquired by Ainekko.
Ownership category
akta.pro rank

Esperanto Technologies industry classification

Industry
Product category
AI Accelerator Semiconductors
NAICS
Computer and Peripheral Equipment Manufacturing (33411), Computer Systems Design Services (541512)
SIC
Electronic Computers (3571), Services-Computer Integrated Systems Design (7373)
akta.pro industry
PRIMARYHDAHABAC
Compute & Acceleration ICs (CPU/GPU/AI/FPGA/ASIC)
SECONDARYHDAHAJAF
AI/Neural Processing Units (NPUs) & ML Accelerators
SECONDARYHDAHAAAB
Advanced SoCs & Application Processors
SECONDARYHDAAAAAL
AI Hardware IP & Chiplet/Packaging for Accelerators (EDA/IP, advanced packaging)
Hardware, Technology & Telecom/Artificial Intelligence & Machine Learning/AI Infrastructure & Compute (GPUs/ASICs, accelerators)

Keywords

  • RISC-V processors
  • AI inference accelerators
  • HPC compute servers
  • Generative AI appliances
  • Energy efficient chips

Where Esperanto Technologies is headquartered

Location

Headquarters

HQ city
Mountain View
HQ country
United States
HQ region
North America

Offices5 records

Markets served

Esperanto Technologies business model

Business model
GTM type
B2B
Offering type
Hardware or Manufacturing
Cost components
Technology or R&D, Personnel, Supply Chain, Operations, Marketing or Sales, Infrastructure

Revenue model

  1. Hardware Sales: Sale of AI/HPC accelerator chips (ET-SoC-1), PCIe cards, compute servers, and Generative AI Appliances. Products available for purchase or secure remote cloud access.
  2. Software Licensing: Primo AI/ML software platform sold as part of integrated solutions with hardware, including model development SDK, optimization tools (AWQ quantization, LoRA fine-tuning), and ONNX Runtime integration.
  3. Evaluation Program: Paid Esperanto Access Program (EAP) providing evaluation access to ET-SoC-1 through remote and on-premise access. Fees credited against initial product orders.

Pricing tiers

ModelBillingPrice
One time/ perpetual licenseMulti-year contractET-SoC-1 AI/HPC Accelerator Chip - single chip for system integrators
One time/ perpetual licenseMulti-year contractET-SoC-1 PCIe Card - compact low-profile PCIe Gen 4 form factor with 32GB LPDDR4x DRAM
One time/ perpetual licenseMulti-year contractAI/HPC Compute Server - 2U rack-mounted with 8 or 16 ET-SoC-1 cards, up to 16,000 RISC-V cores
One time/ perpetual licenseMulti-year contractGenerative AI Appliance - full solution including hardware, LLM software stack and foundation models
SubscriptionAnnualEsperanto Access Program (EAP) - evaluation program for ET-SoC-1

Go-to-market motion1 record

Distribution channels5 records

Marketing channels6 records

Esperanto Technologies product offering

Product offering

Core offering

Esperanto Technologies developed and sold RISC-V based massively parallel AI/HPC accelerator hardware and software. The flagship ET-SoC-1 chip integrates 1,088 64-bit RISC-V in-order cores (ET-Minion) with custom vector/tensor units plus 4 out-of-order RISC-V cores (ET-Maxion) on TSMC 7nm, delivered as chips, PCIe Gen4 cards, 2U compute servers, and a Generative AI Appliance. The Primo AI/ML software platform and EtGlowExecutionProvider (ONNXRuntime backend) accompany the hardware for deploying generative AI, transformer, computer vision, and recommendation models.

Product overview

Esperanto Technologies delivered a portfolio of RISC-V-based AI/HPC products centered on the ET-SoC-1 chip — a TSMC 7nm accelerator featuring over 1 thousand RISC-V cores on a single die — and complementary software. The product line comprised four tiers: (1) the ET-SoC-1 chip itself as the foundational AI/HPC accelerator; (2) the ET-SoC-1 PCIe Card form factor (developed with Penguin Solutions) for system integration; (3) the Esperanto AI/HPC Compute Server and Generative AI Appliance, which packaged 4–16 ET-SoC-1 cards into 2U rack-mounted systems with host Xeon processors; and (4) the Primo AI/ML software platform and EtGlowExecutionProvider for ONNXRuntime-based model deployment. The company ceased operations in 2025; its IP was subsequently acquired by Ainekko (also referenced as Nekko.ai), which launched the open-source AI Foundry platform. As of the data cutoff, product availability was listed on esperanto.ai but the company announced it had ceased operations.

Differentiator

Problem solved

Functional benefit

Brands

  • ET-SoC-1: The world's most efficient RISC-V commercial chip with over 1,000 RISC-V processors on a single TSMC 7nm chip.
  • ET-Minion
  • ET-Maxion
  • Primo
  • Esperanto Generative AI Appliance

Products and services

  • ET-SoC-1 AI/HPC Accelerator Chip TSMC 7nm inference chip with 1,088 ET-Minion 64-bit RISC-V in-order cores (each with custom vector/tensor units) and 4 ET-Maxion out-of-order cores, over 160 million bytes of on-chip SRAM, PCIe Gen4 x8, and sub-30W typical power consumption. Designed for system integrators building AI inference and HPC solutions.
  • ET-SoC-1 PCIe Card Low-profile PCIe Gen 4 card with a single ET-SoC-1 chip, 32 GB LPDDR4x DRAM, full monitoring sensors, and management ports. Up to 16 cards per system. Jointly developed and certified with Penguin Solutions for deployment in North America, EU, UK, South Korea, and Japan.
  • Esperanto AI/HPC Compute Server 2U 19-inch rack-mounted datacenter server with up to 16 ET-SoC-1 PCIe cards (up to 16,000 RISC-V cores), dual Intel Xeon host processors (16-core 6326 or 32-core 8358P), up to 1 TB DDR4-3200 memory, and Esperanto ML/GPU SDKs. Supports generative AI, transformer, computer vision, and recommendation model inference.
  • Esperanto Generative AI Appliance Industry's first RISC-V-based Generative AI Appliance integrating 4 ET-SoC-1 cards with the Primo software stack and foundation models (LLaMA 2, Vicuna, StarCoder, OpenJourney, Stable Diffusion). Full solution for fine-tuning vertical business applications with high data privacy and lower TCO vs. GPU-based systems.
  • Primo AI/ML Platform Advanced AI software platform for optimizing generative AI model inference at scale. Includes model optimization, fine-tuning, RAG-based inference, and orchestration across complex model pipelines, plus an AI/ML Model Development SDK with AWQ quantization, LoRA fine-tuning, Flash Attention, and ML compilation tools.
  • Esperanto Access Program (EAP) Paid evaluation program for ET-SoC-1 with remote and on-premise access options, enabling evaluation of inference on recommendation, visual analysis, and transformer-based models. Annual subscription with fees credited against initial product orders and higher supply priority for participants.

Quantifiable outcome

  • 8.26x speedup compared to CPU execution (Intel Xeon Platinum 8358) for small language model inference
  • +2 more outcomes

Companies that use Esperanto Technologies

Customer profile

Named customers1 record

Segments3 records

Ideal customer profiles3 records

Esperanto Technologies technology and API

Technology

Technology focussed Yes

API detail

Has API
No
API docs
API detail

Core technology

AI maturity

App detail

Integration18 records

AI capability15 records

Feature5 records

Esperanto Technologies partnerships and signals

Strategic signal

Partnerships

Seven partnerships are on record, tiered core, strategic, regional and minor.

  • NEC CorporationcoreStrategic or Co-development Partner · 18 November 2024Cooperation with NEC Corporation on roadmap for next-generation chips and software for HPC computing. Leverages NEC's deep experience in HPC supercomputer design and software stacks combined with Esperanto's energy-efficient RISC-V multiprocessor technology to enable next generation supercomputers.
  • Rapidus CorporationstrategicStrategic or Co-development Partner · 15 May 2024Partnership with Rapidus to enable more energy-efficient designs for AI inference and HPC workloads for the 'post-GPU era'. Rapidus is developing 2nm semiconductor manufacturing capability in Japan with pilot production line planned for April 2025 and mass production in 2027.
  • Penguin SolutionscoreStrategic or Co-development Partner · 22 May 2023Strategic partnership with Penguin Solutions (SGH brand, Nasdaq: SGH) to collaboratively develop multiple systems in various form factors for AI and HPC acceleration in datacenter and edge environments. Jointly developed, tested and certified ET-SoC-1 PCIe cards with safety and emissions certifications for North America, EU, UK, South Korea and Japan.
  • E4 Computer EngineeringregionalChannel Partner/ Reseller/ Distributor · 22 May 2023Value-added partnership in Italy for developing AI and HPC acceleration systems for sophisticated enterprises, universities and centers for advanced research and computing.
  • MEGWARE GmbHregionalChannel Partner/ Reseller/ Distributor · 22 May 2023Value-added partnership in Germany for bringing RISC-V AI and HPC acceleration solutions to research institutes, universities, and industrial and business customers throughout Europe.
  • Elematec CorporationregionalChannel Partner/ Reseller/ Distributor · 22 May 2023Value-added partnership in Japan for RISC-V AI acceleration opportunities across digital electronics, automotive industry and industrial and home segments.
  • Intel Foundry ServicesminorStrategic or Co-development PartnerEsperanto's massively-parallel AI accelerators added to IFS ecosystem. Combination of Esperanto's low-power design techniques and IFS advanced silicon and packaging technologies enables thousands of RISC-V based processors to operate in a single package.

Scale indicators5 records

Recent moves8 records

Expansion highlights7 records

Esperanto Technologies competitors and assessment

Company assessment

Direct peers

  • Graphcore: Graphcore develops IPU accelerators for AI training and inference, targeting data center and enterprise customers. It is comparable to Esperanto in targeting energy-efficient AI acceleration as an alternative to GPU-based systems.
  • SiFive: SiFive is a fabless RISC-V IP and core designer that licenses RISC-V-based CPU IP to chip makers. It overlaps with Esperanto on the RISC-V instruction set, design tooling, and target customer base of custom silicon developers, though it is more focused on IP licensing rather than AI accelerator SoCs.
  • Tenstorrent: Tenstorrent designs AI accelerator chips and RISC-V-based processors for data center and edge inference. It directly competes with Esperanto's ET-SoC-1 in the AI accelerator market, with comparable RISC-V DNA and a focus on energy-efficient inference hardware.
  • Cerebras Systems: Cerebras builds wafer-scale AI accelerator systems for training and inference in data centers. It competes head-to-head with Esperanto's compute server in the AI accelerator market with a similar focus on enterprise and HPC customers.
  • SambaNova Systems: SambaNova develops AI accelerator systems and RDU chips targeting data center inference and training. It is a direct competitor to Esperanto in the energy-efficient AI accelerator category, with a comparable enterprise go-to-market motion.
  • Groq: Groq builds LPU inference accelerators optimized for low-latency generative AI serving. It competes with Esperanto in the AI inference accelerator market with a similar value proposition of fast, energy-efficient inference for LLMs.

Emerging players

  • Rivos: Rivos is a fabless startup building RISC-V-based server-class processors targeting data center and AI workloads. Like Esperanto, it pursues a RISC-V-based architecture aimed at energy-efficient compute, though it is at an earlier stage of commercialization.
  • Ventana Micro Systems: Ventana designs RISC-V-based data center CPUs and accelerator chiplets. It overlaps with Esperanto on the RISC-V compute fabric and enterprise data center target market, focusing on high-performance RISC-V cores for cloud and AI.

Broad incumbents

  • NVIDIA: NVIDIA is the dominant GPU and AI accelerator vendor (H100, Blackwell) and the de facto incumbent that Esperanto's energy-efficiency story is positioned against. It defines the performance and software ecosystem benchmark for AI inference.

Others

  • Penguin Solutions (SGH): Penguin Solutions (SGH brand) is Esperanto's strategic systems integration partner and a system integrator for HPC and AI workloads. It is comparable as a system-level channel and reference design partner in the same data center AI hardware ecosystem.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat4 records

Key risks6 records

Key highlights7 records

Customer concentration

Esperanto Technologies social profiles

Digital presence

Esperanto Technologies compliance and trust

Trust signal

Compliance6 records

Esperanto Technologies financial estimates

Financial estimate

Revenue estimate

Valuation estimate

Esperanto Technologies leadership team

Management profile

Number of profiles

Profiles6 records

Esperanto Technologies funding detail

Funding detail

Funding overview

Funding rounds3 records

Investors7 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

Esperanto Technologies M&A and investment

M&A and investment

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Esperanto Technologies

What does Esperanto Technologies do?

Esperanto Technologies developed and sold RISC-V based massively parallel AI/HPC accelerator hardware and software. The flagship ET-SoC-1 chip integrates 1,088 64-bit RISC-V in-order cores (ET-Minion) with custom vector/tensor units plus 4 out-of-order RISC-V cores (ET-Maxion) on TSMC 7nm, delivered as chips, PCIe Gen4 cards, 2U compute servers, and a Generative AI Appliance. The Primo AI/ML software platform and EtGlowExecutionProvider (ONNXRuntime backend) accompany the hardware for deploying generative AI, transformer, computer vision, and recommendation models.

Is Esperanto Technologies a public or private company?

Esperanto Technologies is a private company. It is classified as venture growth investor backed and is currently closed.

When was Esperanto Technologies founded?

Esperanto Technologies was founded in 2014. It employs 11 to 50 people.

Where is Esperanto Technologies based?

Esperanto Technologies is headquartered in Mountain View, United States, in the North America region.

How does Esperanto Technologies make money?

Three revenue lines are on record. Hardware Sales are the primary driver. The others are software Licensing and evaluation Program.

Who are Esperanto Technologies's main competitors?

Direct peers on record are Graphcore, SiFive, Tenstorrent, Cerebras Systems, SambaNova Systems and Groq. Emerging players are Rivos and Ventana Micro Systems. NVIDIA is listed as a broad incumbent. Penguin Solutions (SGH) is listed as an others.

Does Esperanto Technologies have an API?

No public API is recorded for Esperanto Technologies.

What industry is Esperanto Technologies in?

Esperanto Technologies's product category is AI Accelerator Semiconductors. Its primary akta.pro industry code is HDAHABAC, Compute & Acceleration ICs (CPU/GPU/AI/FPGA/ASIC), with a secondary code of HDAHAJAF, AI/Neural Processing Units (NPUs) & ML Accelerators. Its NAICS code is 33411 and its SIC code is 3571.

Unlock the full company data

50 free credits on sign-up, no credit card required.

Contact sales
Live signals
GoogleET-SoC-1 Chip with More Than 1,000 RISC-V Cores Aimed at Accelerating Machine LearningEsperanto Technologies announced the ET-SoC-1 chip at RISC-V Summit 2020, featuring over 1,000 RISC-V cores and 23.8 billion transistors. The chip claims up to 50X better performance on recommendation networks and 30X on image classification, with up to six chips per card.GoogleEsperanto Technologies Begins Trailing 1,093-Core ET-SoC-1 RISC-V Machine Learning AcceleratorEsperanto Technologies is sampling its 1,093-core ET-SoC-1 RISC-V AI accelerator for evaluation with unnamed lead customers, with general availability expected soon. The 7nm chip packs 1,088 64-bit cores plus vector/tensor units and four out-of-order cores. Pricing has not been announced.GoogleDo You Want To Be An AI Plumber?Ainekko, co-founded by Tanya Dadasheva and Roman Shaposhnik, acquired Esperanto Technologies' IP for a 1,000-core RISC-V inference chip. The company launched AI Foundry to open-source RTL, emulators, and tooling, and is forming an AI Plumbers community. It aims to create an open AI substrate, stack, and systems community.SalesTech StarAinekko Acquires Esperanto Technologies’ Intellectual Property to Power Open-Source Edge AI PlatformAinekko, a startup developing open, software-defined AI infrastructure, has acquired all intellectual property and select assets from Esperanto Technologies, including chip designs, software tooling, and development frameworks for a many-core RISC-V architecture. The company plans to open-source the production-grade ET-SoC-1 technology (featuring over 1,000 Minion cores) to enable community-driven hardware innovation focused on edge AI inference rather than data center deployments. The acquisition aims to unlock latent value in Esperanto's technically advanced but underutilized chip design by making it accessible to developers, researchers, and enterprises across robotics, industrial automation, IoT, and other edge computing applications.GlobeNewswireAinekko Acquires Esperanto Technologies’ Intellectual Property to Power Open-Source Edge AI PlatformAinekko acquired all intellectual property and select assets from Esperanto Technologies, including chip designs and software tooling. The company will open-source a many-core RISC-V architecture with over 1,000 Minion cores, aiming to enable community-driven edge AI hardware. Ainekko is exploring foundation-based governance for long-term transparency.GlobeNewswireAinekko Acquires Esperanto Technologies’ Intellectual Property to Power Open-Source Edge AI PlatformAinekko acquired all intellectual property and select assets from Esperanto Technologies, including chip designs and software tooling. The company will open-source the many-core RISC-V architecture, enabling community-driven edge AI hardware. Ainekko is building on this with its AI Foundry platform.YahooAinekko Acquires Esperanto Technologies’ Intellectual Property to Power Open-Source Edge AI PlatformAinekko has acquired the intellectual property and select assets from Esperanto Technologies, including chip designs and software tooling for its many-core RISC-V architecture. The company plans to open-source these assets to facilitate community-driven innovation in AI inference at the edge. This move aims to provide developers with modular, energy-efficient hardware foundations for applications ranging from robotics to industrial automation.Business Wire BlogEsperanto Technologies and NEC Cooperate on Initiative to Advance Next Generation RISC-V Chips and Software Solutions for HPCEsperanto Technologies and NEC announced cooperation to develop next-generation RISC-V chips and software for high-performance computing. Esperanto's ET-SoC-1 packs over 1,000 RISC-V cores with power under 30 watts, and next-generation products will add HPC features. The collaboration aims to improve performance per watt for supercomputers.GlobeNewswireArteris Selected by Esperanto Technologies to Integrate RISC-V Processors for High-Performance AI and Machine Learning SolutionsArteris announced Esperanto Technologies selected its CSRCompiler for integrating RISC-V processors in high-performance AI and HPC solutions. Esperanto's ET-SoC-1 integrates over 1,000 RISC-V cores with custom vector/tensor units for ML workloads. The partnership aims to develop next-generation energy-efficient solutions for data center and enterprise-edge applications.