Usage of the SSCC Slurm cluster is growing rapidly: Slurm did more computing in February than Linstat ever has in a single month. But there’s still plenty of room for SSCC researchers to think bigger. Slurm users should not hesitate to put unused capacity to work. If your job will run faster with 128 cores, use 128 cores. If your job can use 20 servers, use 20 servers. If you aren’t using Slurm yet, think about how more computing power could enhance your research. Whether that means running models faster, converting tasks that run sequentially into tasks that all run at the same time, using bigger data sets, running bigger simulations, or something else entirely, we’ll be happy to help you see what Slurm can do for you. Just sign up for a workshop on running big jobs in general or Slurm in particular, set up an appointment with one of the SSCC’s statistical consultants, or read the new Guide to Research Computing at the SSCC.
To make submitting jobs to Slurm even easier, we’ve created a Slurm Assistant to help. Just fill out a simple web form describing your job, and it will craft the ssubmit (Slurm submit) command you need.
We’ve also made some major changes to how Slurm manages GPUs (the Slurm cluster includes four powerful A100 GPUs as well as two smaller Tesla T4s). If you’re interested in using GPUs, (re)read the GPU section in the Guide to Research Computing at the SSCC.
The SSCC’s summer training schedule is now available. As always it will include data wrangling classes in R, Stata, and Python. Data wrangling is a crucial skill: we’ve recently seen several problem sets grad students were asked to complete as part of job applications, and they had more questions about data wrangling than statistics or analysis. (No, we can’t help with such problem sets.) We’re also teaching our regular workshop on regression diagnostics, an area many of us need to review from time to time.
A workshop we don’t teach often is Structural Equation Modeling in Mplus. This is a rare opportunity for researchers interested in SEM and we encourage you to take advantage of it.
Starting on April 10th, you’ll need to use multi-factor authentication if you log into Linstat from off campus without using VPN. You can do so using Duo, just like you do with your NetID, SSCC VPN, or Silo. If you have never set up Duo for use with SSCC resources, please visit the SSCC MFA portal before April 10th. If you use a token, contact the Help Desk and we’ll assist you.
This change will affect very few SSCC members but substantially improve the security of Linstat.
The Board of Regents has decided to ban the TikTok app from University-owned computers just like Governor Evers banned it from state government-owned computers. UW–Madison employees, regardless of full, part-time, or hourly status, who have the TikTok app installed on a university-owned device, must uninstall the app as soon as possible. This change only applies to the app, not the website itself. For assistance, contact the Division of Information Technology Help Desk or the SSCC help desk.
Note that an exemption is allowed for university-owned devices used as part of an active academic research project where the TikTok app is part of the project. Those individuals may continue to use the TikTok app on university issued devices. To seek an academic research exemption, please contact firstname.lastname@example.org.
The Division of Information Technology and the SSCC have implemented a technical solution that blocks the future use and installation of the TikTok app on university-managed devices.
Questions about the TikTok app ban can be found on the TikTok policy page.
If you are using a wireless keyboard or mouse and notice that the devices are acting oddly, like specific keys not working or the keyboard and mouse not responding, that is usually a sign that the batteries need to be replaced. It is best to replace the batteries when oddities start happening to avoid disruption or accidents due to keyboard or mouse glitches.