I am trying to add 64 float values in a thread. I have created 32400 threads for processing of 32400 8X8 blocks of an image. The range of float values varies from 100.000 to 50000.00. When I add all the 64 values for a 8x8 block and store them in a float array, I am not getting proper results. I hav…

Problem with addition of float values on GTX8800 Float addition problem

seibert November 16, 2007, 11:32am 3

I don’t quite understand your problem, but I can tell you that the 8800 GTX supports integer operations.

Topic		Replies	Views
CUDA integer ops in hardware the skinny on ints in CUDA and hardware CUDA Programming and Performance	3	20135	March 26, 2007
CUDA kernel converging to CPU double result thought G80 is single precision? CUDA Programming and Performance	6	8615	March 7, 2008
Working on Floats as Integers Tips needed CUDA Programming and Performance	10	9096	January 15, 2008
float/int question CUDA Programming and Performance	3	1527	February 22, 2009
Floating Point Precision of GPU CUDA Programming and Performance	6	2224	September 9, 2010
Floating points CUDA Programming and Performance	3	2065	October 28, 2010
floating point precision CUDA Programming and Performance	3	1472	April 10, 2009
I have two question. CUDA Programming and Performance	11	6926	December 2, 2007
Floats and floats... difference between CPU and GPU? CUDA Programming and Performance	12	14249	February 2, 2010
Integer size per coprocessor CUDA Programming and Performance	14	8499	March 12, 2007