One of the issues was with a Thinkpad and another was with an Ideapad.

Even the ThinkPad range has a lot of variation in it and there are lots of build customisation options. I have never had a major issue with ThinkPads on Linux. I tend to spec them quite highly and keep them for many years.

I just bought a new T16 a month ago and it's great (32Gb RAM, sRGB screen, 2Tb SSD, Ryzen 7 processor). Installed PopOS in about 20 mins and everything just works.

However, as others in this thread have observed, I don't really see the point of trying to run local LLMs on a laptop, unless you want to just play a little. If you want to really play with LLMs wouldn't a separate larger box be more effective rather than trying to do it on a mobile device?

Running LLMs is more of a nice to have. If I can run something like DeepSeek-Coder-V2 even if it's a bit slow, I'll be happy.

If you have a powerful computer at home, you can also offload your ai work to it. It's still local in the sense it's your computer, but it would require network access.

Stephen Toub does an impressive performance deep dive blog post [0] every release. It's over 300 pages long!

[0] - https://devblogs.microsoft.com/dotnet/performance-improvemen...

Yeah, these posts are totally wild. The amount of work that goes into them blows my mind.

I recently read Stoner by John Williams. Would recommend.

Thanks. I have only recently become aware of the Cloudfare tunnels. What's your use case for running them exactly?

you can run a tunnel client on any server and as long as the server has outbound internet access, you can publicly access this server via Cloudflare controlled domain. You can also have standard CloudFlare perks in this way such as CDN and rewrite rules. You don't need to have a reverse proxy server, no need to rely on static IPs or inbound firewall rules in the data center. So you can freely create and destroy droplets they can have different IPs each time.

> many start-ups begin on VPS' and only move to the cloud when scalability becomes a real issue.

This is the idea, yes. But I am hoping we can scale vertically for a long time.

This is my thinking as well! That is why I want to avoid the big 3.

A lot of people have asked about more details. I am going to reply here and hope it gets enough visibility.

Right now, it's a simple web app + db + redis queue. We'll start with just one prod server (and a test/staging server perhaps?) with some kind of a failover setup. My goal is to keep operational complexity to a minimum. I plan to avoid k8s and just use docker compose/docker swarm/dokku.

I have seen enough places burned by going with the big 3 - either through billing, support, or the sheer complexity of managing everything. You might start of with good intentions but you tack on a service because it's easily available and the next thing you know your costs have creeped up and moving things has become a non trivial project that you don't have time for.

I am just looking for a box I can run my setup on with fair pricing, good support, and enough reputation where I have faith they are not going to fold in 6 months. And if things go south, I want to be able to migrate easily.

Do you use all 3? How do they compare in your opinion?

Thanks. I hadn't heard about vultr. How does it compare to Hetzner and Linode?

It’s similar. I liked Hetzner for their affordable bare-metal offerings, but found their interfaces to be fairly clunky. Linode and Vultr (and Digital Ocean) have a nicer web based interface. I think you can find benchmarks online comparing performance of the different VPS providers but IMO the differences are not all that relevant for every day use cases.

It mostly boils down to which service you like more based on entirely subjective variables like company, branding, ownership (Linode was acquired by Akamai).