Reduce development cycles and costs by automating the most intricate component compatibility checks. Try us out: https://app.niyam.xyz/
External link for Niyam AI
11 irving place, Singapore
Singapore, 369551, SG
Niyam AI reposted this
Kudos to Agrim Singh for this idea: AI Analyzes AI – Our Demo Day's Unexpected Sequel. We just wrapped up our demo day with @IterativeVC with over 700 VCs registered to attend. It was an awesome opportunity to showcase what we’ve been building these past months with Niyam AI. But the wild part? We ran our pitch through Notebook LLM, and it generated an 11-minute deep dive on our startup—complete with a podcast-like audio file. Should have just let the AI pitch for us lol. Hsu Ken Ooi Brian Ma Cathy Guo Nathaniel Yim Thank you for guiding us throughout this process. Don't think i'm out of line in suggesting that Iterative is probably the best VC/accelerator experience startup founders can have.
Niyam AI reposted this
An MVP should be just that—a minimum viable product. It doesn’t need to look fancy, and it’s okay if it crashes sometimes. The goal is to tell your story: This solves X pain point by doing Y, and it’s Z times better than the industry standard. It’s not about building everything at once, but about proving your concept. For example, we used Infineon Technologies SiC MOSFETs and built tooling that optimizes power and thermal management for EV inverters. It’s not pretty, but it drastically cuts iteration times and we used it as a strong wedge into the market, opening doors for broader opportunities. Check it out below. Agrim Singh and I have been cookin
Niyam AI reposted this
Hot take (kinda) of the day: Mass AI adoption is gated by hardware challenges! I just wrapped up an insightful session at Penang with BC Ooi from Broadcom and Vivek Raghuraman from Mixx Technologies Inc. The biggest takeaway? Mass AI adoption is gated by hardware challenges—especially in optical integration. Sharing some of my thoughts/insights from the session: 1) Optical integration is a cornerstone for advancing high-performance computing, yet it's hindered by significant challenges in manufacturability, interconnects, and heat dissipation. These obstacles are particularly evident in the development of co-packaged optics (CPO) and ASICs, where the complexity of combining multiple chiplets onto a single substrate creates further difficulties. 2) Fragmentation in EDA tools is a significant problem. While we knew this was an issue in the broader semiconductor space, it was eye-opening to realize its impact on the design of systems like CPOs and ASICs. This insight validated, again, why we're developing AI-based EDA solutions at Niyam AI. We need more integrated tools that streamline the design process and make these sophisticated technologies more accessible and efficient. It won't come from giants like Cadence and Synopsys, but through the new age mom-and-pop EDA solution providers. 3) One possible answer to the issues Vivek highlighted with optical integration: Substrate CPOs. Broadcom deserves credit here because they were among the first to successfully implement substrate CPOs at scale. While it's a promising solution, they don’t fully resolve broader issues such as optical integration complexity, heat dissipation in high-density packages, and manufacturability at scale. Additionally, design flexibility and the lack of standardization across the industry remain significant hurdles. Promising, but lots of development and refinement yet to come. 4) Tencent has been piloting advanced packaging with optical integration, achieving significant milestones in reducing costs and improving energy efficiency. I think the industry buzz is that the early results show that they’ve managed to cut optics costs by nearly 50% and achieve power savings of around 70%. The potential of integrating optics into semiconductor packaging is massive. Looking ahead, it's clear that innovation in both hardware and software is vital to meet the demands of mass adoption in AI. The road is challenging, but companies like Mixx Technologies Inc are driving the necessary changes. Snapshot below of the session and a quick snippet of the MVP we've been building for optimizing SIC MOSFETS for EV Inverters.
Niyam AI reposted this
Niyam AI is your AI co-pilot for electronics manufacturing. Each year, the electronics industry loses billions due to component compatibility issues, delays, defects, and component obsolescence. This is where Niyam AI comes in. Co-founders Samarth Shyam and 🥃 Agrim Singh interviewed over 100 hardware engineers who shared similar frustrations: manual data aggregation, poor collaboration tools, and fragmented processes. To tackle these problems, they created Niyam, a platform that helps engineers build better and faster. Niyam AI reduces development time and costs by: 1) Using predictive analytics to identify and mitigate risks 2) Solving supply chain issues with AI-powered component analysis 3) Streamlining design and validation processes Congrats Samarth Shyam, 🥃 Agrim Singh and the team on the launch! Find the 📌 in the comments for the website link.
Niyam AI reposted this
We're a few weeks away from the Summer 2024 batch, and we're excited to welcome these four amazing companies joining us 🎊 1. Beed (Vietnam) - Modern software for construction tender teams. Founded by Victor Lee and Hung Nguyen. 2. Opilot (Singapore) - Secure, private, and compliant AI copilot solutions for industries that demand the highest standards of data protection and confidentiality. Founded by Julien Lauret, Theodore Garson and Angelica Handover. 3. Yuuki (Singapore) - AI Coach for Managers. Founded by Nihit Nirmal and Nitin Dhawan. 4. Niyam AI (Singapore) - AI co-pilot for electronics manufacturers. Founded by Samarth Shyam and 🥃 Agrim Singh. The next batch starts in July 2024, and the deadline to apply for the Summer 2024 batch is June 24. There's only less than one week left! Find the 📌 emoji in the comments for the link to apply.