andrewboudreau

Software Developer

[andrewboudreau @andrewboudreau]

@MatthewBerman I think it might be time to do another cursor /windsurf video using Gemini 2.5

[andrewboudreau @andrewboudreau]

My extreme sport is resizing windows from the top right corner.

[andrewboudreau @andrewboudreau]

@blizaine @suno_ai_ @pika_labs Can you sync up Eifel 65 blue da ba dee?

[andrewboudreau @andrewboudreau]

@scottmonaghan @bradanlane I refuse to vote for this

[andrewboudreau @andrewboudreau]

@ThePrimeagen The reason your content is good is because you still work in industry

[andrewboudreau @andrewboudreau]

@scottmonaghan Yeah, chatgpt

[andrewboudreau @andrewboudreau]

@cctvidiots 17

[andrewboudreau @andrewboudreau]

@scottmonaghan @BennettTomlin It's spelled Del Taco, pick me up on the way. 🚀🌮

[andrewboudreau @andrewboudreau]

@techsavvytravvy Yeah, I mean do you even know web development if you do not have a jQuery for in your repos??!? 😉

[andrewboudreau @andrewboudreau]

@PhysInHistory Math

[andrewboudreau @andrewboudreau]

@scottmonaghan @BennettTomlin Your bring a whole new meaning to airdrop for scottcoin, lol

[andrewboudreau @andrewboudreau]

@scottmonaghan @BennettTomlin Myth Busted, https://t.co/JxZev4gnwp

[andrewboudreau @andrewboudreau]

@SmokeMonsterTWI Am I remembering my childhood wrong or is that movie about a bunch of cannibals?

[andrewboudreau @andrewboudreau]

@davidfowl We recently switched from on startup to it being in the deployment pipeline. It's much nicer for the team, makes multi region active active a little easier, need to account for slightly longer gap between db and app update though.

[andrewboudreau @andrewboudreau]

@dotNetkow @romero I'll have to check that out, not related but related Game Engine Black Book: DOOM: v1.1 https://t.co/c2HZiAM89M

[andrewboudreau @andrewboudreau]

I'm happy with that response, https://t.co/iR2JcMCVss

[andrewboudreau @andrewboudreau]

@realGeorgeHotz https://t.co/ynfRdDWnIP

[John Carmack @ID_AA_Carmack]

GPUs are built with more memory bandwidth, but higher latency and lower capacity than CPUs. AI accelerators could usefully make a different memory trade — the bandwidth of a GPU (or more), but the capacity of a CPU, in exchange for even higher latency.

For inference, all the weights, potentially a couple TB, can be accessed in a completely linear manner. It could even be a single transaction, streaming at a constant speed into an L2 ring buffer for a GPU core to chase calculations in, akin to racing the beam on old CRT game architectures. You could build a memory system out of masses of dirt cheap RAM, fully in parallel.

Even for training, the memory access patterns can be just a forward read of the weights, a reverse read , and a staggered reverse write of gradients and weights. You could have minimum transaction sizes in the megabytes, and first byte latencies in the many microseconds.

[andrewboudreau @andrewboudreau]

@scottmonaghan I had to unmute and follow the replies to see if anyone explained why it didn't shoot.

[andrewboudreau @andrewboudreau]

@davidfowl Used to run #VisualInterdev and it wasn't until preview net8 that I caught myself thinking "holy crap, they did it". Also waving my arms around showing-off the buffet of simplicity. Next we need fine-tuned LLM with all the latest Microsoft docs. 🙏, Also gpt4 can elevate your css