Author:
Nikita Shiledarbaxi
Software Tools Technical Marketing Engineer, Intel
Learn how next gen developers utilized Intel® Tiber™ AI Cloud and AI PCs
Team Intel supported the HackTX 2024 hackathon held at The University of Texas at Austin (UT Austin) on November 2-3. The attendees got exposure to Intel’s latest AI hardware and software technologies available on the Intel™ Tiber™ AI Cloud, our unified cloud platform provided as a compute resource at the hackathon. We also enabled the attendees to learn about and unlock the potential of Intel® AI PCs for developing hackathon projects and other AI workloads in the future.
Key highlights of the hackathon:
- 800+ developers including university students, professionals and other tech enthusiasts attended the event.
- 500+ new users signed up to Intel Tiber AI Cloud.
- 300+ attendees got guidance on Intel Tiber AI Cloud, Intel® Gaudi® AI accelerator and AI PCs at Intel-hosted workshop (‘Kickstart Your Projects with Intel Technology’).
- 18 real-world project solutions were submitted in the ‘Best Use of Intel AI’ hackathon track.
- Our experts introduced the attendees to the Intel® Liftoff Program for AI startups and our educational initiatives namely, the Intel® Student Ambassador Program and the Educator Program.
Winning Projects at HackTX
Following were the top-notch projects that won the ‘Best Use of Intel AI’ hackathon track:
- Angel’s Protection, the 1st prize winner, is a project that assists security officers in enhancing safety. It stores data from live video feeds based on individuals’ clothing rather than facial recognition, thus prioritizing privacy. The team trained the YOLOv11 model used in the project on Intel Tiber AI Cloud for real-time human detection. A segmentation model extracts clothing attributes and securely stores the data using the high-speed capabilities of Intel’s AI PC. From model training on our cloud platform to model inferencing on our AI PC, all the components of the solution are powered by Intel’s AI technologies.
- PrimeTime, the 2nd place winner, is an advanced productivity application that helps optimize time spent on your computer, thereby maximizing your focus and boosting efficiency. It captures usage data and then categorizes it into four distinct categories based on productivity. It shows under which of the categories each application you use daily falls. It thus helps you recognize which applications’ usage can make you productive. The team learned how to utilize an NPU (Neural Processing Unit) on an AI PC for efficiently executing a multimodal LLM for the first time at the hackathon and learned how to leverage OpenVINO™ Toolkit.
- SuperMAN, the 3rd prize winner, is a project that improves the experience of command-line users, especially those using Linux* commands. It automatically translates natural human language into terminal commands. The team leveraged Intel® Extension for PyTorch* for accelerated LLM inferencing on Intel’s GPUs available on the Intel Tiber AI Cloud platform. The team found powerful Intel AI PC to be capable of handling complex AI workloads with low latency and delivering responsive user experience. By running the project on Intel AI PC locally, the team was able to ensure both high speed and data privacy.
Exciting Prizes for the Winning Teams!
The 1st prize winning team was awarded a Samsung* AI PC, the 2nd prize winning team received Intel® Arc A770 Graphics cards while the 3rd place team was given received Intel® Arc A750 Graphics cards.
Additionally, all three teams were awarded free credits of Intel Tiber AI Cloud for further enhancement of their AI workloads.
What’s Next?
We highly encourage you to get started with developing, testing and deploying high-performance AI or Machine Learning workloads on Intel Tiber AI Cloud – sign up today! Accelerate AI development with our AI PC. Also explore our AI Tools powered by the oneAPI programming model for multiarchitecture, accelerated parallel programming on the latest hardware from diverse vendors.
Additional Resources
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.