[ { "id": 12737, "name": "Tensil", "slug": "tensil", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/3aefe6d1c4f47d9ef26a3296e1724c3efe9bd5ac.png", "website": "https://www.tensil.ai", "all_locations": "San Francisco, CA, USA; Remote", "long_description": "Tensil automates machine learning hardware accelerator design. Our goal is to enable ML in every device. We've developed an algorithm that can design the best hardware accelerator for your machine learning model, unlocking 10x better performance per watt-dollar than GPUs and other hardware accelerators. Our technology enables our customers to create unbeatable products with Tensil custom hardware at the core.", "one_liner": "We turn machine learning models into custom hardware.", "team_size": 2, "industry": "Industrials", "subindustry": "Industrials -> Manufacturing and Robotics", "launched_at": 1557445614, "tags": [ "Edge Computing Semiconductors", "Hardware", "Open Source", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2019", "status": "Inactive", "industries": [ "Industrials", "Manufacturing and Robotics" ], "regions": [ "United States of America", "America / Canada", "Remote", "Fully Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/tensil", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2019/tensil.json", "year_founded": 2019, "city": "San Francisco", "country": "US", "linkedin_url": "", "x_url": "", "fb_url": null, "cb_url": null, "github_url": "https://github.com/tensil-ai", "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/3aefe6d1c4f47d9ef26a3296e1724c3efe9bd5ac.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": null, "founders": [ { "user_id": 214251, "full_name": "Tom Alcorn", "title": "Founder", "founder_bio": "Machine learning enthusiast, self-taught hardware hacker", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/tdbalcorn/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/7df74fe3221be7d16ef42dc3b88d0cac270454ac.jpg" } ], "jobs": [], "news": [], "launches": [] }, { "id": 29747, "name": "deepsilicon", "slug": "deepsilicon", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/86bf1a250ebd174788e945f8cb80bbfa5fad5a1c.png", "website": "https://www.deepsilicon.net/", "all_locations": "San Francisco, CA, USA", "long_description": "", "one_liner": "Software and hardware to run neural networks faster and cheaper", "team_size": 2, "industry": "Industrials", "subindustry": "Industrials", "launched_at": 1724647478, "tags": [ "Edge Computing Semiconductors", "Hard Tech", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2024", "status": "Inactive", "industries": [ "Industrials" ], "regions": [ "United States of America", "America / Canada", "Remote", "Partly Remote" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/deepsilicon", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2024/deepsilicon.json", "year_founded": 2024, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/deepsilicon/", "x_url": "", "fb_url": null, "cb_url": null, "github_url": "https://github.com/deepsilicon", "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/86bf1a250ebd174788e945f8cb80bbfa5fad5a1c.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Diana Hu", "url": "https://www.ycombinator.com/people/diana-hu" }, "founders": [ { "user_id": 1269816, "full_name": "Abhinav Reddy", "title": "Founder", "founder_bio": "Computer Science and Electrical Engineering.\nFocused on creating simple software and fast hardware for running neural networks.", "is_active": false, "linkedin_url": "https://www.linkedin.com/in/abhinav-reddy-387386262/", "x_url": "https://x.com/abhireddy2004", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/70cc4f7cd720b9b145edb1547c6bc4ce0b5a88d1.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 83396, "title": "deepsilicon: Running neural nets with 5x less RAM and up to 20x as fast", "tagline": "We’re making hardware to run neural nets with orders of magnitude less latency, energy, and 5x less RAM. Right now, we’re proving it with software.", "body": "![uploaded image](/media/?type=post&id=83396&key=user_uploads/1269816/61292aaf-ce8a-4b34-ac1a-688fe1bc6517)\n\n**👥 The Team**\\\nHi everyone, we’re running deepsilicon!\n\n**📉 The Problem**\\\nTransformer-based models have become increasingly crucial in various industries, from natural language processing to Vision Language Action models for robotics. However, the deployment and operation of these models, particularly those exceeding a few billion parameters, present significant challenges regarding hardware capabilities, energy consumption, and operational costs.\\\nTraditional approaches to this problem typically fall into two categories:\n\n1. Utilize massive, power-hungry GPU clusters to distribute the computational load.\n2. Compromise on model size and capabilities to fit within existing hardware constraints on the edge.\n\nBoth of these approaches have significant drawbacks. GPU clusters are expensive to acquire and operate, with substantial energy costs and complex cooling requirements. They also introduce latency issues due to inter-device communication and can’t be deployed on the edge. On the other hand, compromising on model size can limit the AI's capabilities and potential applications, putting organizations at a competitive disadvantage.\n\n\\\n📈 **The Solution**\\\nWe help eliminate the need for inefficient distributed computing and compromised model capabilities by providing a full-stack system where we run transformer-based models on a single chip, including existing hardware. Our solution can run on a custom ASIC, dramatically reducing power consumption and operational costs.\n\n\n\n\\\n**_Here's why this is a game-changer:_**\n\n* **Immediate Deployment:** Your large-scale AI model can be operational on a single chip, eliminating the need for complex distributed setups. This means you can leverage the full power of multi-billion parameter models right from the start.\n* **Customization:** The chiplet architecture allows for near-infinite customization, enabling hardware/software co-design tailored to specific client needs.\n* **Efficiency:** Even on existing Nvidia hardware, our software provides a 5x reduction in memory usage\n* **Ease of Use:** Developers can switch their linear layer with our optimized version, dramatically simplifying the integration process and democratizing access to large-scale AI capabilities.\n\nIf you’re a YC company, we offer a 50% discount to help you train your model and deploy it on the device or the cloud!\n\n**🙏 How You Can Help**\\\nWe are looking for connections and collaborations to drive our mission forward. If you or anyone in your network is interested in learning more or helping out, we want to hear from you! We’re specifically looking to connect with:\n\n* Robotics, Drone, and Autonomous Vehicle Companies\n* People looking to train their own foundation model\n* Experts in edge AI deployment\n\nIf that’s you or someone you know, please email us at [founders@deepsilicon.net](mailto:founders@deepsilicon.net)!", "url": "https://www.ycombinator.com/launches/Lh6-deepsilicon-running-neural-nets-with-5x-less-ram-and-up-to-20x-as-fast", "votes": 38, "created_at": "2024-08-26T17:39:52.197Z" } ] }, { "id": 30676, "name": "Stellon Labs", "slug": "stellon-labs", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/e0f795844bb7b220e603390ba0b9ce967fb6db26.png", "website": "https://stellonlabs.com/", "all_locations": "San Francisco, CA, USA", "long_description": "Stellon Labs is an AI research lab that is building tiny frontier AI models that can run on edge devices like smartphones, wearables and embedded systems. Our first open source model, KittenTTS, is a super-tiny text-to-speech model, that got 8K Github stars and 45K model downloads within 2 weeks of launching. ", "one_liner": "Building tiny frontier AI models that run on edge devices", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Engineering, Product and Design", "launched_at": 1755664851, "tags": [ "Edge Computing Semiconductors", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Summer 2025", "status": "Active", "industries": [ "B2B", "Engineering, Product and Design" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/stellon-labs", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2025/stellon-labs.json", "year_founded": 2025, "city": "San Francisco", "country": "US", "linkedin_url": "https://www.linkedin.com/company/stellon-labs-inc", "x_url": "https://x.com/stellon_labs", "fb_url": null, "cb_url": null, "github_url": "https://github.com/KittenML", "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/e0f795844bb7b220e603390ba0b9ce967fb6db26.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Diana Hu", "url": "https://www.ycombinator.com/people/diana-hu" }, "founders": [ { "user_id": 1695666, "full_name": "Rohan Joshi", "title": "Founder/CEO", "founder_bio": "Co-founder and CEO at Stellon Labs where we train open-source frontier speech models! \nWe're better known as KittenML (kittenml.com)\nCheckout KittenTTS at https://github.com/KittenML/KittenTTS \n\nEx-Meta Research, CMU SCS", "is_active": true, "linkedin_url": "https://linkedin.com/in/rohan-m-joshi", "x_url": "https://x.com/ron_joshi", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/3681055ad6cebe46ed6a70a96243efe262959745.jpg" }, { "user_id": 741714, "full_name": "Divam Gupta", "title": "Founder/CTO", "founder_bio": "Packing intelligence in super-tiny AI models. \n", "is_active": true, "linkedin_url": "https://linkedin.com/in/divamgupta", "x_url": "https://x.com/divamgupta", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/f72f7c4e597a0e56f8b85e4aaed075f04dd2589e.jpg" } ], "jobs": [], "news": [], "launches": [ { "id": 93181, "title": "Stellon Labs: Building Tiny AI Models for Edge Devices", "tagline": "We build super-tiny frontier models that can run literally anywhere - No GPU Required! ", "body": "[Launch Video](https://www.youtube.com/watch?v=aUoLxgi5hQA)\\\n\\\n**The problem:**\\\nExisting foundation models are so large that they need an insane amount of compute and memory, making them inaccessible for edge devices like smartphones, laptops, robots and other embedded systems. \\\n\\\n**What we do:**\\\nWe are training state-of-the-art tiny models designed specially to run on these devices without compromising on quality. Our first model is KittenTTS 😻 , which is a tiny open-source text-to-speech model that is under 25MB in size. It got 8K Github stars and 45K model downloads within 2 weeks of launch. At Stellon Labs, we are building tiny frontier models for Speech, Language and Video intelligence to power every single device in the world.\\\n\\\n**Our ask:**\\\nWe would like to connect with developers or companies( both software and hardware) that are interested in running AI models locally. We are happy to provide support for custom models, training and deployments across different applications.\\\n\\\n**Our links:**\\\nGithub - https://github.com/KittenML/KittenTTS (pls star ^^)\\\nWebsite - [https://stellonlabs.com](https://stellonlabs.com/)\\\nEmail - [info@stellonlabs.com](mailto:info@stellonlabs.com)", "url": "https://www.ycombinator.com/launches/OEv-stellon-labs-building-tiny-ai-models-for-edge-devices", "votes": 20, "created_at": "2025-08-20T06:10:11.895Z" } ] }, { "id": 32168, "name": "Synapse Semiconductor", "slug": "synapse-semiconductor", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/37650fbbc73ea1ea0b895fd819e93836825d1520.png", "website": "https://synapsesemi.org", "all_locations": "Durham, NC, USA", "long_description": "We want to push machines to where humans cant go. \r\nPresidential early career award winning Duke Semiconductor ECE Faculty Tania Roy and Sanjeev Chauhan are building a frontier semiconductor research lab. Their focus is on solving physical AI challenges from the semiconductor device level and building products around the moat of impossible to copy innovation. Their first work is on a new vision sensor that can run neural network filters right in the camera pixels and only collect data that machines care about. As physical AI develops, machines needs to understand the world around them if they aim to interact with them.\r\n", "one_liner": "New semiconductors that merge compute and vision", "team_size": 2, "industry": "Industrials", "subindustry": "Industrials", "launched_at": 1785109099, "tags": [ "Edge Computing Semiconductors", "Hard Tech", "Robotics", "Defense", "AI" ], "tags_highlighted": [], "top_company": false, "isHiring": true, "nonprofit": false, "batch": "Summer 2026", "status": "Active", "industries": [ "Industrials" ], "regions": [ "United States of America", "America / Canada" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/synapse-semiconductor", "api": "https://devasheeshg.github.io/yc-api/batches/summer-2026/synapse-semiconductor.json", "year_founded": null, "city": "Durham", "country": "US", "linkedin_url": "https://www.linkedin.com/company/synapse-semiconductor/", "x_url": "https://x.com/SynapseSemi", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/37650fbbc73ea1ea0b895fd819e93836825d1520.png", "app_video_url": null, "dday_video_url": null, "company_photos": [], "primary_partner": { "name": "Tyler Bosmeny", "url": "https://www.ycombinator.com/people/tyler-bosmeny" }, "founders": [ { "user_id": 1383306, "full_name": "Sanjeev Chauhan", "title": "Founder/CEO", "founder_bio": "I like to break things and build things\n\n2 IEDM publications, 3 years Seed Stage VC, scaled deaplearning.com to 300k users, Stanford Linear Accelerator National Lab ML Researcher, IPAC 1st Author", "is_active": true, "linkedin_url": "https://linkedin.com/in/sanjeev-one", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/ed6ba01db61837566ea4234d9d7ec34da6e0e4c1.jpg" }, { "user_id": 3682345, "full_name": "Tania Roy", "title": "Founder", "founder_bio": "Academic by profession, entrepreneur by nature. Love to control the user interface to reality.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/tania-roy-ba272b24/", "x_url": "", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/22e5f2656b361e2b88271fcd468b9cc08d0ba91f.jpg" } ], "jobs": [ { "id": 106484, "title": "Founding Engineer - RETINA", "url": "https://www.ycombinator.com/companies/synapse-semiconductor/jobs/T7HUC4J-founding-engineer-retina", "location": "San Francisco, CA, US / Durham, NC, US", "type": "Full-time", "role": "Engineering", "role_type": "Electrical", "salary_range": "$1M", "equity_range": "10.00%", "experience": "1+ years", "visa": "US citizen/visa only", "skills": [ "Verilog", "VHDL", "Computer Vision" ] } ], "news": [ { "title": "Duke ECE Hardware Startup Selected for Y Combinator | Duke Pratt School of Engineering", "url": "https://pratt.duke.edu/news/synapse-semiconductor-y-combinator/", "date": "Jun 08, 2026" } ], "launches": [ { "id": 107691, "title": "Synapse Semiconductor: We put a neural network inside the camera pixel.", "tagline": "Semiconductors to push machines where humans cant go.", "body": "\n\n**TLDR:** Synapse Semiconductor builds image sensors that process **AI on the sensor** itself. Each pixel captures light and runs neural-network computation in the same device, replacing the need for a GPU while cutting the power and latency of edge vision.\\\n\\\n\\\nA camera has no use by itself. Its value is only when its information gets evaluated. How can a camera that only has value when combined with a GPU support local intelligence? GPUs will get better, yet the fundamental problem is the separation of the camera and the GPU. Solving that separation will unlock bountiful innovation in machine perception. \n\n**RETINA: a chip designed like the human retina which runs neural network operations in the camera pixel.** \n\n\\\nWe have compressed the whole edge AI vision hardware stack into **one wafer**. CNNS, VLMS, etc can run on just one chip that senses light. Our compute transistors are also the same photosensors that see light. There is no longer separation between compute (Nvidia Jetson Nano) and a camera (RealSense Depth Camera). \\\n\\\nIf you are building robotic perception, drones, and vision systems for physical AI —> [book.synapsesemi.org](http://book.synapsesemi.org)\n\n![uploaded image](/media/?type=post&id=107691&key=user_uploads/1383306/bfa2012d-1c74-47b5-bdad-3127e94cd8bf)\n\n", "url": "https://www.ycombinator.com/launches/S0x-synapse-semiconductor-we-put-a-neural-network-inside-the-camera-pixel", "votes": 57, "created_at": "2026-07-27T03:40:58.979Z" } ] }, { "id": 34828, "name": "Kailash Labs", "slug": "kailash-labs", "former_names": [], "small_logo_thumb_url": "https://bookface-images.s3.amazonaws.com/small_logos/9a65a6096b4135a40d95b668829c8fca7b87354d.png", "website": "https://www.kailashlabs.com/", "all_locations": "Bengaluru, KA, India", "long_description": "The world generates petabytes of video every day, and almost none of it is readable to machines. Frontier VLMs that can unlock this (like Gemini) are too expensive to run at the volume real industries have, too generic for any single task, and can't be deployed on-prem where sensitive video lives.\r\n\r\nWe are building an auto vision factory: take video plus a customer's objective, distill a frontier model, and post-train a small open model that is SOTA on that one task and serve it blazingly fast. Our first product, Marlin, is a 2B-parameter open-weights video VLM that matches Gemini-2.5-flash on dense captioning and temporal grounding at a fraction of the cost. ", "one_liner": "A factory for small and specialized video reasoning models", "team_size": 2, "industry": "B2B", "subindustry": "B2B -> Analytics", "launched_at": 1789754590, "tags": [ "Deep Learning", "Edge Computing Semiconductors" ], "tags_highlighted": [], "top_company": false, "isHiring": false, "nonprofit": false, "batch": "Fall 2026", "status": "Active", "industries": [ "B2B", "Analytics" ], "regions": [ "India", "South Asia" ], "stage": "Early", "app_video_public": false, "demo_day_video_public": false, "app_answers": null, "question_answers": null, "url": "https://www.ycombinator.com/companies/kailash-labs", "api": "https://devasheeshg.github.io/yc-api/batches/fall-2026/kailash-labs.json", "year_founded": 2026, "city": "Bengaluru", "country": "IN", "linkedin_url": "https://www.linkedin.com/company/kailash-labs/", "x_url": "https://x.com/KailashLabs", "fb_url": null, "cb_url": null, "github_url": null, "logo_url": "https://bookface-images.s3.amazonaws.com/small_logos/9a65a6096b4135a40d95b668829c8fca7b87354d.png", "app_video_url": null, "dday_video_url": null, "company_photos": [ "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/7ed96e3a2a70109795a500adac0ddb05a732ae64.png", "https://bookface-images.s3.us-west-2.amazonaws.com/attachments/52cccc6bf7d1cffb8653e974a88d0f756caf60eb.png" ], "primary_partner": { "name": "Harj Taggar", "url": "https://www.ycombinator.com/people/harj-taggar" }, "founders": [ { "user_id": 2326582, "full_name": "shubham sharma", "title": "Founder/CEO", "founder_bio": "Founder at Kailash Labs. Previously visiting researcher at KAIST working on robot learning and long-term memory for decision-making agents. Published at NeurIPS (spotlight) on offline RL during research at Tsinghua. Aerospace Engineering from IIT Bombay.", "is_active": true, "linkedin_url": "https://www.linkedin.com/in/shubh-am-sharma/", "x_url": "https://x.com/happyypablo?s=11", "avatar_thumb_url": "https://bookface-images.s3.us-west-2.amazonaws.com/avatars/b9159802a0bf905460372aec0622850ce7fa0e26.jpg" } ], "jobs": [], "news": [], "launches": [] } ]