How to specify architecture
while compiling CUDA program in Visual profiler
How to change command
line flag in CUDA
Hello, today I’ll let you know how to change compilation
option such as flag in visual profiler. Sometimes it happens we need to specify
the architecture of our GPU to run any sample code. For example, if your code
contain “atomicAdd()” function then you SM architecture
should be > SM10, although you SM architecture is greater than 1.0 but you can’t Build your project in Visual profiler because of compiler flag is SM10. When you try to do this visual profiler through an error like this;
should be > SM10, although you SM architecture is greater than 1.0 but you can’t Build your project in Visual profiler because of compiler flag is SM10. When you try to do this visual profiler through an error like this;
atomicAdd undefined identifier
|
In this case you specifically, specify architecture greater
than SM10 in compiler flag.
Let us compile our code using command line;
nvcc –arch=sm__30
|
Same you want to do in your visual profiler. So you need to
change this line
nvcc -gencode=arch=compute_10,code=\"sm_30,compute_30\
|
To
nvcc -gencode=arch=compute_30,code=\"sm_30,compute_30\
|
Procedure to achieve this;
Step 1: Open your project
Step 2: Go to your GPU option
Step 3: Change GPU architecture
value to required SM value, in this case I have change it from SM10 to SM30
Step 4 : You are done
Got Questions?
Feel free to ask me any question by
commenting below because I'd be happy to walk you through step by step!
Want to Contact us? Click here
What a great approach to the subject, I was looking for. Thanks for these tips
ReplyDeleteC++ > Games Code Examples
ReplyDeleteAs I can see, there are even more different programming languages, and I hear about it for the first time. Anyway, I am a user of applications running in the cloud because I know that such services always offer help in case of problems. I like to use solutions from https://grapeup.com/cloudboostr/support-and-services/ and more I've never had any problems.
ReplyDeleteGreat Post Keep sharing such informative posts.
ReplyDeleteApply Now For android application development course
Call Now on 7070905090
What’s up, I wish for to subscribe for this website to get latest updates, therefore where can I do it please help out.
ReplyDeleteJaipur Raj University BA Final Year Result
This comment has been removed by the author.
ReplyDeleteAsk him any Questions about herpes
ReplyDelete[…] It’s a crazy miracle product. Read my review here. […]
I was cured of herpes simplex virus his contact ___________Robinsonbuckler11@gmail. com.......................
this is the best herbal remedy…
Helpful data accessible here, Thanks This is very interesting article, Very skilled blogger.
ReplyDeleteExcellent post. I was always check this blog, and I’m impressed. Greatjob!
ReplyDeleteWow! Such an amazing and helpful post this is. you're great buddy
ReplyDeleteSatisfied with all the information I found in this article. Very helpful.
ReplyDeleteI enjoy every little bit of it. I bookmarked to check out new stuff of this blog. Thanks... MM
ReplyDeleteYour blog have nice information, I got good ideas from this amazing blog... MM
ReplyDeleteI admire this blog here. Awesome article, keep doing it! thanks!... MM
ReplyDeleteI learned a lot in this blog. Thank you very much for sharing good content... MM
ReplyDeleteI will support you. Such a very useful article. A must read post, Thanks a lot... MM
ReplyDeleteAw, this was an extremely good post. thanks. Pretty great post. Thankyou... MM
ReplyDeleteI have bookmarked this site to check out the new stuff post. Keep it up, you're great... MM
ReplyDeleteThis is a clear and practical explanation of an issue that trips up a lot of people when they first start working with CUDA tooling. The disconnect between the actual GPU capability and the compiler flags in Visual Profiler can be frustrating, especially when errors like atomicAdd being undefined do not immediately point to an architecture mismatch.
ReplyDeleteWhat works well here is how you tie the command line approach back to the Visual Profiler configuration. Showing the exact change from compute_10 to compute_30 makes it obvious what the tool is really doing under the hood and helps readers build a mental model instead of just following clicks. That understanding pays off when they move between IDEs, command line builds, or CI environments.
From a quality perspective, this kind of documentation is invaluable. Configuration issues are often harder to diagnose than code bugs, and having clear steps reduces trial and error. Capturing these setup nuances alongside test scenarios and build requirements in a shared place, such as Tuskr test management software, helps teams avoid rediscovering the same problems as hardware and toolchains evolve.