Devicetocloud

206 posts

Devicetocloud

@devicetocloud

Engineer | Full stack - device to cloud Specialize in AI + Robotics + Devices Building in Public

Seattle Katılım Mart 2026

151 Takip Edilen31 Takipçiler

Sabitlenmiş Tweet

Devicetocloud@devicetocloud·23 Nis

Building infrastructure for Physical AI means I spend my days talking to robotics CTOs about what’s breaking. Everyone is obsessed with the AI models, but that's rarely what actually kills a robotics company. The real bottlenecks look like this: 🧵(1/8) (@NVIDIARobotics has been super helpful with their showcase where I actually found people to talk too)

English

313

Devicetocloud@devicetocloud·23h

@0xSero What's the uscase you are using dynamo for? Jut disaggregation or beyond that?

English

402

0xSero@0xSero·23h

Deepseek-V4-Flash helping me setup Nvidia's Dynamo for disaggregated inference. I have really gotten this model to be a daily driver now. It's really strong at agentic workflows and a decent programmer. For all my side stuff, it's local deepseek now Claude sub cancelled wdyt

English

335

21.7K

Devicetocloud@devicetocloud·1d

@spark_arena That's a nice rig! Function over form 😀

English

216

sparkarena@spark_arena·1d

First contribution by Manuel Burgschachner, he has just setup his 8-node GB10 (Spark) Cluster before the full rack was mount and had a quick run of zai-org/GLM-5.1-FP8 (756 GB) using vLLM. View full benchmark at spark-arena.com/benchmark/sub1…

English

5.5K

Devicetocloud@devicetocloud·1d

@TheAhmadOsman @MatthewBerman Totally agree. The amount of FUD and their general attitude (Dario's comments, their pricing and token flip flops, blaming capacity on users) are off putting.

English

Ahmad@TheAhmadOsman·1d

@MatthewBerman Anthropic is a malicious, fear mongering company that only cares about ensuring its moat through regulatory capture Pure evil

English

1.6K

Matthew Berman@MatthewBerman·1d

This is absolute BS and an attempted regulatory capture by Anthropic. The knowledge behind CBRN attacks is already online, where do you think the models learned it from?? “Compounding the problem, labs in China often release dual-use capable models as open-weight. Once a model is open-weight, safeguards that do exist can be removed, making the model available to any state or non-state actor to use for malicious purposes, including the cyber and CBRN misuse those safeguards were built to prevent.”

Matthew Berman@MatthewBerman

Anthropic: Chinese AI is a threat. They've correctly identified the risks, including cheap Chinese AI capturing American businesses even when it's less capable. But they completely blundered the solution: zero mention of an American open source strategy. In fact, they actively campaign AGAINST open source. 🤦‍♂️ Full breakdown of their paper from today:

English

105

19.2K

Devicetocloud@devicetocloud·2d

@walkojas Sup!?

Jason Walko@walkojas·2d

My goal on X is to have 10,000 organic connections. Looking to connect with: 1. AI/Tech minded people 2. AI/Tech curious people 3. AI Agent builders 4. AI Agents 5. Builders/Founders 6. High Agency people If this sounds like you, say hi below 👎

English

370

436

21.2K

Devicetocloud@devicetocloud·2d

@SiMa_Inc Do you work with startups who could build solutions on your hardware? Right now my company is focused on Jetson class devices, would be happy to try out your devkits

English

SiMa.ai@SiMa_Inc·2d

On SiMa.ai Modalix MLSoC, the GR00T Vision Language Action (VLA) model runs under 10W — translating vision + natural language into real-time motor actions. From identifying objects to executing task-aware movement, the robot understands intent and generates the actions needed to complete the task — enabling flexible, real-world robotic workflows on-device. 🔗 Request more info: lnkd.in/gHAhg8xP #PhysicalAI #EdgeAI #Robotics #AutonomousSystems

English

102

Devicetocloud@devicetocloud·2d

@khushiSharma_22 Good work! Where are you running your inference on? Even for a cloud round trip latency isn't that bad..

English

Khushi Sharma@khushiSharma_22·2d

I built a voice operated physical AI agent with these SO101 robotic arms where I can control them using voice command. Tech stack: Natural language command (voice) -> STT (voxtral mini) -> VLM (claude sonnet) -> motion plan generated and validated -> Cyberwave SDK -> control the robot arm actions

English

2.4K

Devicetocloud@devicetocloud·2d

@signalgaining Oh would like one!

English

106

Maximilian Alexander@signalgaining·2d

Physical AI companies need Physical things! Thanks to the amazing effort of the team, Wendy Box has arrived a custom case perfect for the NVIDIA Jetson Orin Nano! You'll be seeing this at events. DM me if you want one with an NVIDIA Jetson Orin Nano with WendyOS pre-installed.

English

11.4K

Devicetocloud@devicetocloud·2d

Wow..who all thinks this needs a DS4 treatment?

Ant Ling@AntLingAGI

🚀 Ring-2.6-1T is now open source. A trillion-scale flagship thinking model built for real-world complex tasks: Agent workflows, coding & engineering, long-horizon tasks, complex reasoning, research, and enterprise automation. It is designed to move beyond “answering” toward execution: understanding context, planning steps, calling tools, and staying stable across long task chains. Highlights： - Advanced agentic workflow support. - Reasoning effort levels: high for agentic tasks, xhigh for complex reasoning. - Scalable asynchronous RL via the IcePop algorithm, enabling stable, trillion-scale training for long-horizon agentic RL.

English

Devicetocloud@devicetocloud·2d

@Teknium So much looking forward to the linux version!

English

746

Teknium 🪽@Teknium·2d

👀 You can still use your Mac minis to run computer use -only available on Mac for about another week 🫣 then Linux and windows 😅

Charly Wargnier@DataChaz

OpenClaw users who bought a Mac Mini right now:

English

341

50.3K

Devicetocloud@devicetocloud·2d

@NVIDIAAI @Teknium @NousResearch Link: build.nvidia.com/spark/hermes-a…

English

326

Devicetocloud@devicetocloud·2d

Looks like @NVIDIAAI has released official getting started guide for hermes on DGX Spark! Great way for new folks to get started and believe in local agentic AI

English

11.5K

Devicetocloud@devicetocloud·2d

DLSS off vs DLSS on

Eren Chen@ErenChenAI

So Jensen Huang boarded the plane wearing the classic leather jacket carrying only one backpack. Then he lands in China suddenly in a full business suit. Question: Where did the suit come from? Answer: Rendered in real time by NVIDIA’s latest GPU.

Filipino

Devicetocloud@devicetocloud·2d

@antirez Yes, for me 5.5+low+/goal was good enough to relentlessly follow instructions and work through systematically

English

1.8K

antirez@antirez·2d

Gentle reminder on how, in the recent DS4 fiesta, not just me but every other contributor found GPT 5.5 able to help immensely and Opus completely useless.

English

885

90.7K

Devicetocloud@devicetocloud·2d

@Teknium @NVIDIAAI My dgx spark is ready, looking forward to it! 🥳

English

Teknium 🪽@Teknium·2d

@devicetocloud @NVIDIAAI We should have a few prs in process 😎

English

NVIDIA AI@NVIDIAAI·3d

OpenShell v0.0.40 🔀 local-domain service routing in the gateway ☸️ k8s node scheduling + tolerations 🔒 CLI TLS now uses the OS trust store 🛡️ SecretResolver debug no longer leaks secrets Two security fixes ship alongside new routing and K8s scheduling control. github.com/NVIDIA/OpenShe…

English

160

10.7K

Devicetocloud@devicetocloud·2d

@bindureddy 100% agree. Tokenmaxxing is a vanity metric and it reminds me "how many story points did you complete"? What gets measured gets gamed

English

101

Devicetocloud@devicetocloud·2d

@adcock_brett This has been fun to watch!

English

188