I've generally tried to avoid Nvidia cards because binary blob drivers are a pain (especially as a FLOSS developer I occasionally need to build newer kernels). I believe the recent firmware changes mean the nouveau driver can now control clocking but I've no idea what the status is for CUDA which I assume you need to run the models.
Now I'm all for smashing atoms and the LHC did a grand job with the Higgs. However are we sure just smashing things harder is going to be as revelatory as other things we could spend the money on? What other grand physics instruments could we build? For example LISA will be a massive step change in our gravitational detection capabilities?
So this row wasn't over something like a Hamas flag or something like that, just a patch showing the kids cultural heritage? You see kids with Welsh flags and dragon logos all the time here. I'm fairly sure I've seen some of the Ukrainian kids with yellow and blue on.
There are still some things meson can't do which we need the configure script for. The meson upstream are slowly working through our feature requests 😉
Indeed - the general configure, build install steps are fairly universal and the configure script doesn't have to cover from autoconf. We still have that and Makefiles as a wrapper around a meson based setup to keep the process familiar.
There are various insulation schemes you can apply for but they are usually limited in who can apply. You probably don't want to pay for Lord Farqwa upgrading his 16 bed converted wheel house pad.
Also older housing stock. My house is over 120 years old so that does limit done of what you can do to improve things. That said with decent windows and loft insulation there is something to be said for the thermal properties of a bunch of stone.
I would support additional taxes on Landlords inversely proportional to the energy rating of the house. Currently there is no incentive for landlords to improve the energy efficiency of their properties. At least when you "own" a house the action to improve it is on you.
I have an binding for my terminal, my Emacs and a general fuzzy selector for apps or SSH hosts. I generally operate with everything full screen with windows sorted across 10 named workspaces across 3 monitors.
I've just started a Voyager run through with my kids. Dare I say S1 of Voyager is a much better introduction to "modern" Trek than the fairly uneven S1 TNG? I think Janeway is underrated as a Star Fleet captain.
AntennaPod is great but most ads are directly injected into the files these days. The best you can do is skip them when they come up. AntennaPod allows you to set custom skips for the start and end of individual podcasts though which is useful.
I have a stack of old phones in my drawer of tech I need to go through and check which ones work on postmarketOS. I think I have an old Pixel or two as well as Nexus phones.
I'm startng to think I should run local models for completion engines but I don't have a GPU I can use. What's the best option for accelerating these models? Are there PCI cards that give a better bang for buck for running models?
I've generally tried to avoid Nvidia cards because binary blob drivers are a pain (especially as a FLOSS developer I occasionally need to build newer kernels). I believe the recent firmware changes mean the nouveau driver can now control clocking but I've no idea what the status is for CUDA which I assume you need to run the models.
They do look pretty affordable though 😀