Got tired of running out of RAM when analyzing large datasets, and I don't like paying Amazon or Google to rent their clusters, so I ordered pieces to assemble a new laptop from a "for parts" base Getac V110 G5 (rugged not because I want to pretend I'm a militant gundam pilot but because I drop things a lot, plus a laptop I can use outside sounds nice) with 32GB ram, all it needed was a drive and a charger so I got a 2TB SDD.
If it works it should be pretty solid, and I want to try a new #Linux distro. Been using #!++ for years, now leaning towards Void Linux, but I'm wondering if anyone has any suggestions for a distro optimized for data analysis, if that makes any sense. Ideally that would mean not just package availablility, I know how to download and compile from source if I need, but how it works at the base level. I know a lot of them use the same kernel since they're mostly forks of Debian, Ubuntu, or Arch, but some like Garuda have features that effect how it utilizes RAM etc.
I was looking into NixOS too but it seems like that's more for people who love hacking their own OSs. Would also try out a BSD or Redox OS but worry about the learning curve and dealing with incompatiblilty
Holly A. Gultiano /
npub12g…h55ru
2023-12-15 03:00:55
Author Public Key
npub12gvxrj7dnx4ylr6y5uuatp7wlga4n8ctr8t72trvtaaml8a2r3yqzh55ruPublished at
2023-12-15 03:00:55Event JSON
{
"id": "a5722211d93bfdc9c5a5c2c512457a2b58eeb2fb77b5c84754035d2768ea7ed7",
"pubkey": "521861cbcd99aa4f8f44a739d587cefa3b599f0b19d7e52c6c5f7bbf9faa1c48",
"created_at": 1702609255,
"kind": 1,
"tags": [
[
"t",
"linux"
],
[
"content-warning",
"Computer question, Linux distros"
],
[
"proxy",
"https://synapse.cafe/users/axoaxonic/statuses/111582200140273818",
"activitypub"
]
],
"content": "Got tired of running out of RAM when analyzing large datasets, and I don't like paying Amazon or Google to rent their clusters, so I ordered pieces to assemble a new laptop from a \"for parts\" base Getac V110 G5 (rugged not because I want to pretend I'm a militant gundam pilot but because I drop things a lot, plus a laptop I can use outside sounds nice) with 32GB ram, all it needed was a drive and a charger so I got a 2TB SDD. \n\nIf it works it should be pretty solid, and I want to try a new #Linux distro. Been using #!++ for years, now leaning towards Void Linux, but I'm wondering if anyone has any suggestions for a distro optimized for data analysis, if that makes any sense. Ideally that would mean not just package availablility, I know how to download and compile from source if I need, but how it works at the base level. I know a lot of them use the same kernel since they're mostly forks of Debian, Ubuntu, or Arch, but some like Garuda have features that effect how it utilizes RAM etc. \n\nI was looking into NixOS too but it seems like that's more for people who love hacking their own OSs. Would also try out a BSD or Redox OS but worry about the learning curve and dealing with incompatiblilty",
"sig": "04c7ff082973e09f070199488eb6ffd817bc9368ba859b3154d224fc7b73b356c439035b77cbff4199b31b9fa9affc09a91b211ad05f67791b7467e3cc0b49fe"
}