Scribd
Upload
52 views27 pages

Gpu Cuda Part1

This document provides an introduction to GPUs and CUDA programming. It discusses the evolution of GPU microarchitectures from early graphics accelerators to modern GPUs with programmable hardware. GPUs have thousands of cores devoted to highly parallel computation and follow a single-program multiple-data processing model. The document outlines GPU terminology and components like streaming multiprocessors and execution cores. It also compares CPU and GPU architectures and provides examples of NVIDIA GPU microarchitectures from Fermi to Volta over several generations.

Uploaded by

Raghav Ganesh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
52 views27 pages

Gpu Cuda Part1

This document provides an introduction to GPUs and CUDA programming. It discusses the evolution of GPU microarchitectures from early graphics accelerators to modern GPUs with programmable hardware. GPUs have thousands of cores devoted to highly parallel computation and follow a single-program multiple-data processing model. The document outlines GPU terminology and components like streaming multiprocessors and execution cores. It also compares CPU and GPU architectures and provides examples of NVIDIA GPU microarchitectures from Fermi to Volta over several generations.

Uploaded by

Raghav Ganesh
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 27

IT301: INTRODUCTION TO

CUDA
By,
Ms. Thanmayee
Adhoc Faculty,
Department of IT,
NITK, Surathkal
OUTLINE
● Introduction to GPU
● Evolution of GPU microarchitectures
● General Purpose GPU
● Introduction to CUDA
● CUDA Execution Model
● CUDA Memory Model
● Steps in GPU Execution
● Hello World Program
● CUDA Device Variables
● CUDA Programming examples
Let's learn about GPU..

● A little history:
− The first GPUs were designed as graphics accelerators
● supported only specific fixed-function pipelines.
− In the late 1990s, the hardware became increasingly programmable
● Culminating in NVIDIA's first GPU in 1999.
Graphics Processing Unit

● Has thousands of cores and ALUs.

● They can handle billions of repetitive low level tasks.

● GPU is specialized for compute-intensive, highly parallel computation.

● They are devoted to data processing rather than data caching and flow
control.

● Follows SPMD processing model.


CPU versus GPU
More closer look at GPU:
GPU Terms:
● Stream processing -- Term used to denote processing of a stream of
instructions operating in a data parallel fashion.
● Stream Processors (SPs) – the execution cores that will execute the
stream. Each stream processor has compute resources such as register
file, instruction scheduler
● Streaming multiprocessors (SMs) -- groups of streaming processors
that shares control logic and cache.
GPU Microarchitectures
Fermi Architecture (2010)

https://www.nvidia.com/content/PDF/fermi_white_papers/NVIDIA_Fermi_Compute_Architecture_Whitepaper.pdf
What each SMs
have..

● Memory operations are handled by a set of 16 load-store units in


each SM.
● A set of four Special Function Units (SFUs) is also available to
handle transcendental and other special operations such as sin,
cos, exp, and rcp (reciprocal)
● Along with the group of 16 load-store units and the four SFUs,
there are four execution blocks per SM.
○ 16 + 16 Cores (2 blocks)
○ LD/ST
○ SFU
A total of 32 instructions from one or two warps can be dispatched in
each cycle to any two of the four execution blocks within a Fermi SM
Kepler (2012)
Fermi Versus
Kepler
Maxwell (2014)
Pascal (2016)
Fermi : GTX480
GTX580

Kepler : GTX 680


GTX 780
GTX 780 Ti

Maxwell: GTX 980


GTX 980 Ti

Pascal : GTX 1080

Source : https://www.techspot.com/article/1191-nvidia-geforce-six-generations-tested/
Volta micro architecture (2017)
NVIDIA
Tesla GPUs
THANK YOU

You might also like