Kartikey Rawat·Aug 15Accelerating JAX on Qualcomm Snapdragon: Building a Native QNN Backend with OpenXLA PJRTIntroduction
Kartikey Rawat·Jun 15Inside-Out: Building a High-Performance On-Device LLM Client in Flutter with Qualcomm’s QNNRunning Large Language Models (LLMs) directly on smartphones used to feel like a demo you’d show once and never use again. The models were…
Kartikey Rawat·May 12Nova IDE: Redefining Local Coding with Snapdragon X Elite and WebGPUIn the era of massive cloud-based LLMs, a new frontier is emerging: On-Device AI. Today, we’re introducing Nova IDE, a browser-based coding…
InGoogle Developer ExpertsbyKartikey Rawat·Apr 18Bringing Multimodal Gemma 4 E2B to the Edge: A Deep Dive into LiteRT-LM and Qualcomm QNNRunning Large Language Models (LLMs) on mobile devices used to be a futuristic dream. Today, with Google’s release of the Gemma 4 family…A response icon1A response icon1
Kartikey Rawat·Mar 28On-Device Generative AI: Running Stable Diffusion v2.1 on Snapdragon X Elite NPUFor the longest time, Generative AI has been synonymous with massive cloud GPUs — expensive, remote, and often slow to respond. But that’s…
Kartikey Rawat·Mar 13Running Qwen3–4B On-Device: Deploying a 4B LLM on Snapdragon NPUsFrom Model Architecture to Community Contribution on Qualcomm AI Hub Org at HuggingFace
InGoogle Developer ExpertsbyKartikey Rawat·Feb 24Building a Nano Mixture-of-Experts (MoE) Language Model in JAX from ScratchA beginner-friendly deep dive into how Mixture-of-Experts works, why it matters, and how to build one in pure JAX/Flax.
InGoogle Developer ExpertsbyKartikey Rawat·Jan 29Architectural Evolution and Implementation Strategy of the LiteRT CompiledModel APIExecutive SummaryA response icon1A response icon1
InGoogle Developer ExpertsbyKartikey Rawat·Jan 14Benchmarking the Edge: Gemma 3n vs. Qwen 3 on Android with LiteRT-LMIn the rapidly evolving world of Large Language Models (LLMs), the focus is shifting from massive cloud-based monsters to lean, efficient…A response icon1A response icon1
Kartikey Rawat·Dec 19, 2025Unleashing the Beast: Running YOLOv8 on Snapdragon X Elite NPU at 65 FPSThe era of AI PCs is no longer a marketing slogan — it’s real, and it’s sitting on our desks.