Ronald Mannak (@ronaldmannak) / X

Ronald Mannak

27.1K posts

Ronald Mannak

@ronaldmannak

Local AI. Building @picogpt Download: apps.apple.com/us/app/pico-pr… Discord: discord.gg/Nrf5y8Uaxw

San Francisco, CA

Joined April 2008

Pinned
Ronald Mannak
@ronaldmannak
Apr 16
Apple Silicon + Gemma 4 fans: this is for you. Pico AI Server now supports continuous batching with MLX-Swift. 43 tok/s on 1 stream. 26 tok/s per stream on 2 concurrent streams. That’s 52 tok/s total. a 21% throughput gain on a six-year-old MacBook Pro M1 Max!
00:00
49K
Ronald Mannak
@ronaldmannak
Mar 3, 2025
The Mac mini is the cutest Mac apple released in a looooong time
Geek
00:00
Geek
699K
Ronald Mannak
@ronaldmannak
Apr 30, 2024
I'm almost too embarrassed to ask, but... who's hiring?
592K
Ronald Mannak
@ronaldmannak
Apr 17, 2025
Replying to @Kalshi_Film and @TheOscarRace
Easy mistake to make, but this is Les Grossman, the studio executive.
64K
Ronald Mannak
@ronaldmannak
Sep 29, 2025
Replying to @Angry_Staffer
Laugh all you want, but I’m pretty sure the guy in the lower right corner will someday strap balloons to his house and float off to South America on a grand adventure.
16K
Ronald Mannak
@ronaldmannak
Jun 30, 2025
Wow. Buried in a Siri article: "Apple this month also nearly lost the team behind MLX." What is going on at Apple? MLX is literally the only technology putting Apple ahead of the competition in the emerging local AI market.
176K
Ronald Mannak
@ronaldmannak
Sep 9, 2025
My biggest takeaways from today’s Apple keynote: - Apple is focused on improving the GPUs for AI inference. The Apple Neural Engine wasn’t even mentioned once. - Expect inference compute to increase 3-4x (maybe more) in the upcoming M5 Apple Silicon chips for Mac. Memory
Ronald Mannak
@ronaldmannak
Sep 9, 2025
GPU cores in the new A19 Pro have specialized “Neural Accelerators”? What exactly is it? It sounds like Apple is adding AI hardware to the this year’s Apple Silicon. Can’t wait for M5 Macs to launch.
191K
Ronald Mannak
@ronaldmannak
Mar 29, 2023
ChatGPT is fundamentally changing the way we interact with computers. We all agree about that. However, I believe many experts are *underestimating* how big the paradigm shift will be. My prediction: LLMs will replace the operating system, App Store and apps. Here’s why 👇
349K
Ronald Mannak
@ronaldmannak
Jun 30, 2025
When you bet your startup on Apple Silicon and read that Apple almost let the MLX team leave
00:00
106K
Ronald Mannak
@ronaldmannak
Oct 31, 2023
Replying to @sdw
Sorry, I can’t hear you. I’m watching the kids through my Vision Pro.
21K
Ronald Mannak
@ronaldmannak
Apr 27, 2024
This is the next big thing (for nerds): your own home cluster of Mac Mini or Mac Studio for distributed local LLM inference. Made possible by the unified memory architecture of Apple Silicon. Should I create a distributed version of @PicoGPT MLX server just for fun?
elegyal ˆ •͈ ·̫ •͈ ˆ ♪
@elegyals
Apr 26, 2024
Replying to @yacineMTB
you might prefer to use ip over thunderbolt, ethernet is too slow
196K
Ronald Mannak
@ronaldmannak
Nov 4, 2025
Your Mac is about to run inference like a datacenter. Coming soon to MLX-Swift: Continuous batching: the fastest way to handle multiple inference streams locally. It starts with regular inference and seamlessly upgrades to batched mode when new requests arrive. The best of both
00:00
61K
Ronald Mannak
@ronaldmannak
May 9, 2025
Replying to @krzyzanowskim
That’s the only change they made? Fascinating
76K
Ronald Mannak
@ronaldmannak
Mar 20, 2024
Run Apple MLX from your menu bar. Introducing Pico MLX Server, a graphical frontend to download and start multiple(!) AI models locally on your Mac. You can use it with any chat client you like (e.g. @PicoGPT) that uses the OpenAI API standard.
105K