15K+ @LinkedIn | Staff SoC Engineer | Full-Chip Signoff | RTL2GDS | STA | ECO | LEC | EMIR | Package MCM | 9 Tape Outs | Semiconductor Professional with 6+ Years' Experience | Career Guide for Aspiring Engineers
LVS Explorer Debugging 🌏Connection to ICV VU & ICV ShortFinder 🌏Interactive Debugging through GUI environmen 🌏Highlights only shortest Short paths
Learn how to manage and store footage for large volumes of video cameras with NVIDIA Video Storage Toolkit (VST), an application service offered by #NVIDIAMetropolis microservices for Jetson. This video provides a comprehensive guide from installation, user interface, code samples, and debugging.
Dive into the intricate mechanics governing the operation of car door windows! 🚗💨 Unlock the secrets behind this everyday marvel and gain insights into the engineering finesse that keeps us rolling smoothly. . . #CarWindows #MechanicalEngineering #AutomotiveTech #EngineeringMarvels #VehicleInnovation #STEM #EngineeringInsights
⚡️ Boost your render times! With GPU acceleration built into today's OS, Voke Video Productions turns hours of rendering into minutes. Time to create, not wait! #GPUPower #RenderFast
To anyone who has had a project go up in smoke—literally. It’s time for a safer approach with the MPLAB® SiC Power Simulator. https://lnkd.in/gDEfCMXq #SiCPower #SiliconCarbide #PowerManagement #MPLABSimulator
Build AM32 ESC project from source code with Vimdrones AM32 ESC DEV Board
At first, we thought software decoding AV1 might be too slow for our purposes. But after we pushed the planar YUV (output of the decoder) --> RGB (input of our renderer) conversion to the GPU it's pretty darn fast 🎥🕺🏎💨 Playback at 20x speed for this video, and not a single stutter! Another nice side effect: All the new formats we needed for that internally are now exposed, so you can log pictures using those directly to Rerun. You can find the PR, including all new formats supported, from the link in the comments 👇
Semiconductor Test Professional | Edge AI & Computer Vision Specialist | Optimizing Test Strategies with Machine Learning
Got my Nvidia Jetson Orin NX 8 gb a fresh install of Jetpack SDK 6.0. I kept a backup NVME with Jetpack SDK 5.1.2 as backup just in case. After the install I check for basic functionalities. Wifi is ok but bluetooth is unexpectedly buggy. I connected my logitech MX3 mouse and MX keyboard via bluetooth but are not tunctional. Tested my UVC cameras and are working fine. BTW the included opencv 4.8 is not cuda enabled. I needed to install another version 4.10 and is cuda enabled. Would do further testing and try loading my models using the upgraded tensorrt.
Generative AI Leader at Lyric.ai | Founding Partner at Multiversal Ventures | Founder, Orfeo Health
So here is a crazy idea that came up while I was spitballing with Michael Yang of Ollama crew... 1. Make a unikernel with unikraft + rust + candle (or ollama go) 2. use vAccel + unikraft to get access to GPU on qemu-vaccel 3. Use vulkan to map and "snapshot" GPU RAM Now superfast load unikernel, load GPU memory snapshot of the model so we get rid of loading time. I don't know how to do this yet but apparently can be done on A100s. Have a service that "seeds" the tensors for the inputs and boom serverless embeddings at massive scale .... #listofthingstohack #unikernels #serverless
Working on a small #rustlang CLI tool "affected" to aid myself with monorepos like #Nx, especially on the CI/CD side. Detecting modified files and projects, running tasks against those, etc. Still in the early stages but hopefully evolves into a valuable utility. https://lnkd.in/ePfi_mD4
