Author: Nikita Shiledarbaxi
The latest generation of personal computers called the AI PCs have revolutionized the way one can unlock AI acceleration to more efficiently execute AI workloads locally, combining the computational potential of CPU, GPU and NPU (Neural Processing Unit). Intel recently sponsored a ‘GenAI Goes Local’ hackathon at the AGI House on September 7, 2024. The participant developers harnessed the power of GenAI to develop high-quality, optimized solutions using LLMs (Large Language Models) and Vision Language Models (VLMs) on Intel-provided powerful AI PCs at the event.
About the AGI House
The AGI House is a community of AI founders and researchers that serves as a "hacker house" for startups, engineers, and researchers for marathon coding sessions.
Highlights of the hackathon:
- Intel, partnering with Lenovo*, provided 23 AI PC machines at the event to help developers build innovative GenAI projects.
- Participants included premier local startups and AI thought leaders (such as Jeffrey Morgan, the co-founder of Ollama* framework for LLMs).
- There were 200 total attendees and 16 project submissions out of which 12 projects utilized AI PCs.
- 2 startups enthusiastically came forward to join the Intel® Liftoff program that helps AI startups grow into large-scale industry-standard companies.
Insightful Workshop Sessions at the Hackathon
A diverse lineup of speakers from leading tech companies like Nexa AI*, Ollama, Google*, Hugging Face*, and more showcased their cutting-edge technologies at the event. Intel hosted a workshop introducing the attendees to AI PCs and outlining their purpose, functionalities, key components, and the importance of privacy, low latency, and essential libraries and tools.
Winning GenAI Projects at the Hackathon
The following projects were identified as winners for their efforts and ability to enhance productivity and user experiences through GenAI.
- DejaView, the 1st -place winner, is an AI-powered local semantic image search engine that handles complex human language queries. The team leveraged AI PC, used local images to ensure privacy and local Llama LLMs and VLMs to annotate images and search over images.
- ThirdEye , the 2nd-place winner, is an AI application that stores users’ entire UI history as context, enabling easy and quick access to required information in the future. The team used AI PC and employed VLM to cache users’ information followed by LLM for recalling the stored information when required. The app allows one to search for not only words, but also calendars, screenshots, or anything that went across the UI.
- Open Motion is a fast, AI-powered API that automates calendaring and resolves calendar collisions. It locally automates users’ work plan for free.
Each of the 1st-prize winning team members was awarded an AI PC by Intel. While the other winning teams received free credits of Intel® Tiber™ Developer Cloud to build, test and deploy innovative solutions using Intel’s oneAPI-powered extensive software portfolio on the latest hardware including CPUs, GPUs, and other accelerators such as the Intel® Gaudi® AI processors.
| → Details of the other notable GenAI projects developed during the hackathon are available here. |
Accelerate AI Development with Intel
We encourage you to jumpstart AI development with Intel’s high-performance developer resources. Leverage AI PC development kits, AI frameworks and tools from Intel, allowing your real-world application projects to attain greater heights of efficiency and acceleration.
Learn More
Technical Software Product Marketing Engineer, Intel
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.