LLR GPU application

carpetpool · 2020-10-02, 21:41

Hi,

I'm wondering if the GPU LLR software found at Jean Penne's website is in progress or of any use.

I've installed CUDA Toolkit 11.1 (deb local) for Linux Ubuntu 20.04. After that I tried to compile the CUDA version of LLR:

(llrcuda Version 3.8.1 complete source, presently only for Linux build)

Code:

~/llrcuda381src$ make
/usr/local/cuda/bin/nvcc -g -O -I/usr/local/cuda/include -I. -Wno-deprecated-gpu-targets lprime.cu -c
nvcc fatal   : '-I/usr/local/cuda/include': expected a number
make: *** [Makefile:11: lprime.o] Error 1

There's also no readme which came along with the download so now I'm stuck on what I should do.

Thanks for help.

BenR · 2020-10-20, 19:01

The optimization option, -O, is causing the error (it's no longer legal after CUDA 10). Try using -O3.

2020-10-02, 21:41	#1
carpetpool "Sam" Nov 2016 2²×83 Posts	LLR GPU application Hi, I'm wondering if the GPU LLR software found at Jean Penne's website is in progress or of any use. I've installed CUDA Toolkit 11.1 (deb local) for Linux Ubuntu 20.04. After that I tried to compile the CUDA version of LLR: (llrcuda Version 3.8.1 complete source, presently only for Linux build) Code: ~/llrcuda381src$ make /usr/local/cuda/bin/nvcc -g -O -I/usr/local/cuda/include -I. -Wno-deprecated-gpu-targets lprime.cu -c nvcc fatal : '-I/usr/local/cuda/include': expected a number make: *** [Makefile:11: lprime.o] Error 1 There's also no readme which came along with the download so now I'm stuck on what I should do. Thanks for help. Last fiddled with by carpetpool on 2020-10-02 at 21:41

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Funny job application replies	pinhodecarlos	Lounge	18	2022-05-06 12:08
Srsieve Application failed	PawnProver44	Information & Answers	12	2016-03-11 00:43
lasievee application	pinhodecarlos	NFS@Home	2	2013-05-19 20:23
Ports of the application	nohero	Twin Prime Search	9	2007-06-17 10:12
application of euler's phi function	TalX	Math	3	2007-04-27 11:50

2020-10-20, 19:01	#2
BenR Nov 2014 3² Posts	The optimization option, -O, is causing the error (it's no longer legal after CUDA 10). Try using -O3.