Posts avec #GPUProgramming
Les posts publics sur X qui utilisent #GPUProgramming.
Thakur @iamthakur404Solved my first LeetGPU CUDA challenge: Vector Addition. Turning CUDA concepts into actual GPU code, one kernel at a time. #CUDA #GPUProgramming #NVIDIA #CPlusPlus
Thakur @iamthakur404
Thakur @iamthakur404Day 23/90 | GPU Programming Using cudaStreamAddCallback(); Applications should use or switch to using cudaLaunchHostFunc(). #GPUProgramming #CUDAProgramming
Thakur @iamthakur404Day 17/90 GPU Programming! CUDA Tile memory operations: tile-space loads/stores, partition_view, boundary handling, gather/scatter, and masked access. #CUDA #CUDACpp #GPUProgramming #NVIDIA #GPULearning #ParallelComputing #CPlusPlus #AIEngineering
Thakur @iamthakur404Day 16/90 GPU Programming! A lighter session today, learning the basics of CUDA Tile: block positioning, tile creation, and compile-time constants.
Thakur @iamthakur404Day 1/90 of learning GPU Programming! Started with GPU architecture and Apple's Metal documentation to understand how GPUs execute work in parallel. #GPUProgramming #CUDA #CPP #BuildInPublic #90DaysChallenge
Thakur @iamthakur404Day 28/90 | GPU Programming How to influence and inspect the optimization/resource stage with nvcc. As a C++ developer, I’m obsessed with control, optimization, and speed. Now learning how those decisions happen closer to the GPU hardware. #NVIDIA
Thakur @iamthakur404Day 34/90 | GPU Programming Understanding the 4 Unified Memory Paradigms in CUDA. #CUDA #GPUProgramming #NVIDIA #CPlusPlus
Thakur @iamthakur404Somewhere between the code, the errors, and another late night.
Thakur @iamthakur404Day 36/90 | GPU Programming Back to the GPU grind after a short break. Today, diving deeper into CUDA Unified Memory Hardware vs. Software Coherency and why fine-grained CPU–GPU sharing matters. #CUDA #GPUProgramming #NVIDIA #CPlusPlus #UnifiedMemory
Thakur @iamthakur404Day 3/90 of learning GPU Programming. Today I learned how GPU memory is organized: Global memory (GPU DRAM),Registers & Shared Memory (on-chip) L1/L2 caches • Unified Memory for CPUtoGPU access
Thakur @iamthakur404Day 2/90 of learning GPU Programming Diving into the CUDA Programming Guide and learning the GPU programming model, SIMT, and tiles. One step closer to writing efficient GPU kernels. #GPUProgramming #CUDA #CPP #BuildInPublic #90DaysChallenge
Thakur @iamthakur404Day 19/90 GPU Programming! Today I covered CUDA Tile Atomics, Optimization Hints, and C++ Performance Tips.
Thakur @iamthakur404Day 18/90 GPU Programming! Covered control flow, broadcasting, element-wise arithmetic, and key tile primitives including matrix multiply/MMA, reductions, scans, transpose, permutation, and element-wise selection. #GPUProgramming #CUDAProgramming
Thakur @iamthakur404Day 10/90 GPU Programming writing CUDA kernels, making friends with the GPU #CUDA #90DaysChallenge
Thakur @iamthakur404Plan: Write GPU code today. Reality: Let me just understand this one concept first… opens CUDA documentation 2 hours later: still reading. But hey, that’s also part of learning GPU programming.
Thakur @iamthakur404Day 37/90 | GPU Programming — Mixing Hardware & Software Coherency in CUDA Unified Memory. #CUDA #GPUProgramming #NVIDIA #CPlusPlus #UnifiedMemory
Thakur @iamthakur404Day 25/90 | GPU Programming Unified and System Memory. Understanding where data physically resides, how it moves, and how the CPU-GPU interconnect affects latency and bandwidth is becoming much clearer. #GPUProgramming #CUDA #NVIDIA
Thakur @iamthakur404Day 24/90 | GPU Programming Diving deeper into CUDA Asynchronous Execution today. #GPUProgramming #CUDA #90DaysOfCode
Thakur @iamthakur404Day 25/90 | GPU Programming Unified and System Memory. Understanding where data physically resides, how it moves, and how the CPU-GPU interconnect affects latency and bandwidth is becoming much clearer. #GPUProgramming #CUDA #NVIDIA
Aman Swar @Compile_Conquerwas getting bored So I made my GPU go Brrrrr Introducing KernelLab – A collection of high performance CUDA kernels optimized for deep learning, matrix operations, and image processing. Check the chaos : #CUDA #GPU #AI #MachineLearning [1/X] 🧵
