what are the performances of my machine ?
how much usage of the machine is free ?
okey thank you !
what does 200 CPUS mean ?
aah i see
You can check the hardware specs on the website.
When you run the simulations, you can check performance of your job after connecting to your server by pressing the Visulaizer.
You can check current pricing on the pricing page
As of now, I see 200 CPUS-Hrs per month being offered for free.
You can run machines with 2 CPU, for example, for 100Hrs each month.
Or you can run machines with 4 CPUs for 50 hrs each month.
If you need any further help, please ping again. We are here.
I wanted to participate in A.I. competition
My output log file shows python 2.7 but my simulation needs python3 and it gets terminated fast, what should I change in .sh file for bash shell on cloudshell?
#!/bin/sh #!/bin/python # run from directory where this script is cd `echo $0 | sed 's/\(.*\)\/.*/\1/'` # extract pathname pip install --upgrade pip pip install rebound python myscript.py
Oh ok. Thank you though! Good day.
I dont think you should put 2 hash-bangs on the top. You dont need second one.
Give full path to python in the last line
I am not technical support staff. If you need help you can go to Collaborate tab, share with email@example.com with admin priviledgs and then go to Discuss tab and leave a message there.
Someone from our technical support team will help.
How do i determine the number of hours to allocate for specific jobs. I submitted a job for 28 hours using 8 processor cores, i received an inconclusive result. Too bad. Can i restart the job?
I want to send a job now and i don't want an inconclusive result. Can someone help me. My two messages have not been attended to.
There is really no way to predict how long a job will take. Scientifically, it is impossible.
You can start a job with wall time limit of 400hrs, say for example. If your job ends after 30hrs, we automatically refund 370hrs. This is how the system works and it is designed to exactly support the situations like yours. I suggest you purchase a larger pool of CPU credits so you are able to select longer wall time limits before you start your job. Kogence CPU Credits do not expire, you can use them over time.
Please let us know if you need further help.
I am trying to create a new model but is not working
So, regarding the credits... Is there a way to earn them by referring or something. Actually I am student only from India and can't afford to purchase. And I really need this platform.
Kogence will be so helpful for my final year project
how to increase credits
@YuvamBhateja. Please log in first. Top right corner, where your username is shown, there is a dropdown menu. Go to "My Compute Plans". There you will see a section called "My Individual Compute Plan". Towards the right you will see a a settings button (gears icon). Clicking on that button would let you purchase additional credits.
can we have WIEN2K software for material simulation
We are happy to deploy this software on kogence. Please send us the link to software licensing and usage restrictions (if any), download and install instructions and some example models that we can use to test the setup.
I have been able to login. Thanks to your support.
I have a friend who wants to use abinit code for phonon calculation. Let me know the possibility of using your platform and send sample script for abinit code to do that.
@AkinolaSamsonOlayinka, yes we can set that up. Could you please send us link for the code you mentioned?
I had a quick question
if I add more nodes, does my process run parallel on all of them?
how to use SGE?
I was wondering if I cancel a run do I get my credits back?
I really need to cancel it right now as I didn't choose the correct machine to run it on
You mean if you use cluster? Yes. For cluster case you submit your job to job scheduler SGE by writing a SGE script. MPI/SGE will take care to make sure it runs in parallel.
My reccomendation ---- first you should try using bigger machince --- meaning add more cpu to single node. Instead of adding nodes to cluster.
I dont think we have any cluster example with QE on our site. We have some cluster examples with Elk and Lumerical. You can look up the format of SGE script in those examples. Or if you have your single node job running properly, you can contact us and we can write a SGE script for you to submit to cluster.
Just like you use Sun Grid Engine in traditional cluster. You submit job by "qsub job'. You check status by "qstat job".
Get your single node/multiple CPU job running properly. We caan then help you move it to cluster case.
@MatthewFu, We just released the version 2.0. https://kogence.com/app/docs/Blog:Kogence_HPC_Grand_Central_Release_2.0_is_Here!
Now you can send jobs without writing any job submission scripts. You can do that from kogence on the browser. Just check mark the bix "run on cluster" on Cluster tab and on Stack tab of your model. Please let us know if you need any further help.