NVIDIA Developer Forums
cudaMemcpy() and L2 cache.
Accelerated Computing
CUDA
CUDA Programming and Performance
Robert_Crovella
May 17, 2016, 9:01pm
2
ALL
accesses to DRAM take place through the L2 cache. And the L2 cache cannot be disabled.
1 Like
What the difference using L2 by cudaMemcpy() and from kernel?
show post in topic
Related topics
Topic
Replies
Views
Activity
FAO: Nvidia Engineers:- Memory Leak in cudaMemcpyAsync Only occurs on Host To Device memory transfer
CUDA Programming and Performance
4
5870
August 18, 2010
Understanding Caching/Flushing Behavior/Performance in computeprof for Kepler
CUDA Programming and Performance
6
3321
September 19, 2014
cudaMemcpy2D() and a few gray hairs It's very slow
CUDA Programming and Performance
8
4533
February 13, 2009
What the difference using L2 by cudaMemcpy() and from kernel?
General Topics and Other SDKs
1
457
May 4, 2023
Got out of memory from cudaMemcpy
CUDA Programming and Performance
13
3911
January 28, 2022
Difference between L2 read/write transactions and L2_L1 read/write transactions ?
CUDA Programming and Performance
3
1346
August 28, 2019
Understanding the functioning of nvprof and .cv data load option
CUDA Programming and Performance
8
3047
December 11, 2014
Tesla K40 L2 bandwidth
CUDA Programming and Performance
12
4027
December 23, 2015
2MB size limit cudaMemcpy
CUDA Programming and Performance
18
7290
April 18, 2011
Is cudaMemcpy() real-time safe?
CUDA Programming and Performance
cuda
11
497
March 30, 2024