My friend, Matt, uses RStudio to further his doctoral research in statistics. The problem is all Matt has access to is a laptop that’s several generations old. The last code he ran took 3 days with the laptop doing nothing else.
I have 2 desktop computers that are way more powerful than his laptop and both have rtx graphics cards.
Is there a way I can setup both of my desktops to be extra compute for Matt’s equations? Like a render farm for 3D images?
Both systems run Windows 11. One has a ryzen 7 5700x with 32gb ddr4 and an RTX 3060 Super. The other has a ryzen 5 3500x with 16gb ddr4 and an RTX 2070. I know this will increase my power consumption, but power is cheap and relatively clean where I am.
Please feel free to ask follow up questions. And I apologize if this is 101 level stuff, I tried using different terms to search the forums and didn’t see anything related to this.
Assuming that his computer can access yours, you could run an instance of RStudio Server on your computer:
It's not quite the same as seamlessly accessing your unused compute, but functionally, it will work the same way. He'd kick off his code on your computer via web interface from his computer.
If you used one of our commercial offerings, I'm sure you could figure out how to rig up multiple machines, but I'd start with "one powerful desktop" and see what happens. R itself isn't a particularly parallel language, so unless he's got experience writing that type of code, you're probably not going to get the speed-up you're imagining.
Or am I just giving him remote access?
Yes, I'm talking about him just having remote access to your desktop via the instance of RStudio Server. By running it inside of Docker or similar, you'd be able to give him access to your resources while not giving him access to your general filesystem.