Channel Avatar

Tech learn by Shalu @[email protected]

1.4K subscribers

topmate.io/shalu_chaudhary10 Here you will find videos rela


Welcoem to posts!!

in the future - u will be able to do some more stuff here,,,!! like pat catgirl- i mean um yeah... for now u can only see others's posts :c

Tech learn by Shalu
Posted 1 month ago

β€οΈβ€οΈπŸ§ΏπŸ™

0 - 0

Tech learn by Shalu
Posted 1 month ago

Will miss this place

0 - 0

Tech learn by Shalu
Posted 1 month ago

Yes AI bubble is real #aibubble

0 - 0

Tech learn by Shalu
Posted 1 month ago

NVIDIA DGX GB300 next video

0 - 0

Tech learn by Shalu
Posted 1 month ago

NVIDIA GPU & AI Quick Reference Guide
1. NVIDIA GPU Architecture Basics
CUDA Cores: Parallel processors for general-purpose GPU computing.
Tensor Cores: Specialized for deep learning matrix operations.
Memory Types: Global, shared, and texture memory optimize performance.
Streaming Multiprocessors (SMs): Core units containing CUDA cores.
2. CUDA Programming Intro
CUDA (Compute Unified Device Architecture) is NVIDIA's parallel computing platform.
Kernel functions run on the GPU, launched with grid and block dimensions.
Memory hierarchy is crucial for performance optimization.
Common languages: CUDA C/C++, Python with Numba/CuPy, PyCUDA.
3. AI Frameworks Optimized by NVIDIA
TensorFlow with TensorRT integration for inference acceleration.
PyTorch with cuDNN for deep learning training performance.
ONNX Runtime with CUDA execution provider for model deployment.
RAPIDS for GPU-accelerated data science workflows.
4. Common Interview Questions
Q: What is the difference between CUDA cores and Tensor cores?
A: CUDA cores handle general parallel tasks; Tensor cores accelerate matrix math used in deep
learning.
Q: How does GPU memory hierarchy affect performance?
A: Proper use of shared and global memory reduces latency and improves throughput.
Q: What is CUDA kernel launch configuration?
A: It defines how threads are organized into blocks and grids for parallel execution.
5. Key Tools & SDKs
CUDA Toolkit – Core libraries and compiler for GPU programming.
cuDNN – GPU-accelerated primitives for deep learning.
TensorRT – Inference optimizer and runtime engine.
Nsight Systems – Profiling tool for GPU applications.

0 - 0

Tech learn by Shalu
Posted 1 month ago

How is nityo infotech???

0 - 0

Tech learn by Shalu
Posted 1 month ago

Is it worth to learn databricks please let me know #databricks

0 - 0

Tech learn by Shalu
Posted 7 months ago

Will upload renewal video for azure data engineer #dp203 today

0 - 0

Tech learn by Shalu
Posted 7 months ago

If you want to learn new update about azure connect to azure site it’s really helpful #azure#newupdate

0 - 0

Tech learn by Shalu
Posted 7 months ago

Preparing for aws data engineer #aws

0 - 0