We're a group of makers shipping together. We help each other stay accountable and reach our goals.
Apply for access
Siavosh Zarrasvand
@siavoshzarrasvand
Latest, fully maxed spec. Always.
I've got an M1 32 GPU cores 64 GB RAM and 4TB memory.
The M1 32 GPU cores is an M1 Max. If you look at benchmarks, the M1 Max beats normal M3s and M3 Pros, only losing out to M3 Max.
Generally, I find in hardware, the mid layer priced options are always worst "pound-for-pound" over a cycle as you end up having to replace your device sooner. I paid £7k for my M1 in 2021 but I can use it for another 1-2 years. I ran LLMs on it without any problems.
Meanwhile, had I bought an M1 Pro with 32GB RAM and 4TB memory for £5k, I would have had to replace it by now...
@siavoshzarrasvand okayyy but my question is about Mac apps. 😅
Lol, my bad!
On that note:
vscode
warp
beekeeper studio (db tool)
vlc
frostwire
expressvpn
insomnia
proxyman
Haha no worries, I was just a bit confused about whether I'd written something unclear. Thanks for your list!
I was drunk mate, literally, I was a bit wasted. But I answered in all seriousness. You can see I didn't even have typos in my response, just in my reading ;-)
After trying Daisy for a day I decided to just go bare bone Tailwind + Web Components. Removed Daisy.
understandable. how do web components work for you? have never used it myself
Very well actually. Syntax-wise, they are very similar to React or Vue components - so it was easy for me to pick up. I find their rendering cycles much easier to predict than React Components.
After trying Daisy for a day I decided to just go bare bone Tailwind + Web Components. Removed Daisy.
Daisy it is, thank you for highlighting the thing about importing components.
Don't think there are that many drug traffickers here.. ;-)
I don't. I quit day job. Get 1-2 years of savings and just do it.
It's 12k. Not sure I want to risk Apple not taking it back.
And even if you return it, a beast like that costs to return.
If you don't damage it, they have to take it back and you don't have to worry about shipping just fill out the form and get it for free.
My use case would be to just run the models. So think running 100B+ parameter chat-LLMs locally.
Lol, my bad!
On that note:
vscode
warp
beekeeper studio (db tool)
vlc
frostwire
expressvpn
insomnia
proxyman
Haha no worries, I was just a bit confused about whether I'd written something unclear. Thanks for your list!
I was drunk mate, literally, I was a bit wasted. But I answered in all seriousness. You can see I didn't even have typos in my response, just in my reading ;-)
LMAO 😂