Dreamscale Labs (YC F26)
Joined November 2023
- Tweets115
- Following220
- Followers162
- Likes1.3K
Future of hci
We’ve raised $1.3M from YC & Afore Capital so you never have to type or speak again.
Typing is slow. Talking into your computer is awkward. Subvocal fixes both
Our glasses give you the speed of speech without you making a sound or touching a keyboard.
It tracks your internal speech muscles with non-contact sensing and decodes silent, almost invisible mouthing into text at more than 200 words/min.
Pre-orders are now live at:
subvocal.co/preorder?utm_sou…
Flux!
Dreamscale now supports Flux 3 Action!
It's a 7B world action model released by Black Forest Labs (@bfl_ai), which had hit SOTA on NVIDIA's RoboLab-120 benchmark.
Run and fine-tune F3A on your robots with a simple interface at @DreamscaleLabs!
Adding @DreamscaleLabs labs to the list 😁
Recent launches by @physical_int, @dreamlabs_ai, and @perceptroninc are making API-based, on-demand generalist robotics models a real option.
This is a win for the whole space.
Companies can specialize and build faster without rebuilding the intelligence stack in-house.
Robots brains in space 🛰️ 🌩️
My favorite company in the current @YCombinator batch is the @DreamscaleLabs (YC F26) team. Four college dropouts Eric, Chris, Kevin, and Antoine working on cloud inference for robotics. This is almost all done at the edge right now, but will be a ginormous market in a few years as humanoids and other robots ramp up and will wanna run real-time inference that is far more powerful than can be done on the edge. Go team! 🤓👌
@eternaI_entropy, @KDog112358, @pentestduck, @AntoineNeedGPUs
Over wifi - average sf flat. 🐑⚡️
What does it take to run π0.5 in the cloud with zero stalls and imperceptible latency?
For us:
• Custom kernels
• Custom network transport
• Dedicated post-training to account for latency
For you:
• 1-line CLI command
• Serverless compute that spins up in 2 seconds
• Don't pay for idle usage
See for yourself: (1x video in thread)
Volume is the way to make cloud economics work for robots.
128x robots on a single node (h100) is just the beginning.
Excited to announce that we're partnering with HUD @hud_evals to make it easy to evaluate robot AI models in the cloud!
Through our collaboration, we show:
- Up to 128x VLAs running in parallel in simulation
- Running compute-intensive World Action Models over the cloud
- Achieving real-time cloud inference on physical robots
Today, we (@dreamscalelabs) are launching Dreamscale. The fastest cloud inference platform for robotics.
Dreamscale lets you run VLAs and WAMs from the cloud with ultra-low latency. We optimise across the full-stack, from the kernels to network for real time control.
We’re making shared endpoints generally available today on dreamscalelabs.com
1/N
The robots I dream of do not just interact with the physical world, but the personal one.
Think care robots that need to remember the smallest details about health, mood, and hazards in the home, they need to be conversational and invigorating, and they need to know when to act.
5/N
Antoine Nguyen retweeted
Scaling laws apply to JEPA models too!
From 22M to 8B
Introducing RoboJEPA, the first scaling law for multiembodiment world models.
We scaled JEPA predictors up to 8B params, showed scaling laws extrapolate and correlated compute scaling with planning performance.
We release all checkpoints and the code.
robojepa.github.io
Antoine Nguyen retweeted
We just ran one of the largest robot models that exists today, all without a local GPU.
And it’s running real-time at <50 ms overhead.
Scaling laws are becoming real for physical AI. Perceptron’s Isaac 0.5 is one of the largest open-source models, sitting at 36B total parameters.
@perceptroninc joins other frontier robotics labs like @DynaRobotics and @GeneralistAI in showing the need to massively scale pre-training data, from a 1000s to millions of hours of video. Specifically, they show that the more egocentric video the model is trained on, the less teleop data that it needs to achieve the same loss.
At @DreamscaleLabs (YC F26), we’ve noticed an important truth here: data and training compute do not just scale in isolation. The model sizes must increase too, and GPT-6 Astra has given a glimpse into what a trillion-parameter robotics model could do.
We’re building to serve that future - one that is not constrained by the bottlenecks of on-board compute, but rather, allowing large models to run fast over the cloud.
Reach out if you’re interested in our work.
Antoine Nguyen retweeted
@DreamscaleLabs (YC F26) is the first inference provider to serve NVIDIA’s DreamZero robot model over the cloud!
And we beat @nvidia's own reported inference speed by 8%, with just half the silicon.
DreamZero is a model that pioneered the new wave of “World Action Models” that has taken over generalist robotics today. It promises to generalize better to new tasks, learn physics from human video, and transfer well to new embodiments.
Normally, DreamZero requires 2x GB200 superchips (each costing $60K) or is painfully slow (waits 5.7s before moving 1.6s), which made it infeasible to deploy to robots in practice.
Now, with Dreamscale’s real-time cloud inference infrastructure, we’ve demonstrated fast, smooth motion on actual robots with sub-second latency.
As scaling laws become more relevant in robotics, frontier physical AI models are growing more computationally expensive than ever. With Dreamscale, we are finally making World Action Models (WAMs) accessible to startups and robot developers across the world.
Reach out if you are interested in cloud robot inference.
Robot brains can (and should) live in the cloud 🐑 ⚡
OpenAI's Astra showed us truely competent spatial reasoning arises out of enormous scale and data.
Skyrocketing dram costs, onboard energy limits and most painfully the need for generality afforded only by scaling are forcing us off board onto datacenter grade compute.
So we're now announcing @DreamscaleLabs , an applied research company engineering this future into a possible reality, gutting the latency and amping up the reliability necessary for safe remote operation.
Onwards to Physical AGI 🚀
@pentestduck @KDog112358 @eternaI_entropy
Today we're incredibly excited to announce @DreamscaleLabs (YC F26)! We're an applied research company putting robot brains in the cloud.
Robot AI models are growing larger and on-board compute is getting increasingly inaccessible due to cost. We're taking a contrarian bet that the future is moving inference off the robot and into the cloud, while still meeting real-time deadlines required by robot motion.
We're building the real-time inference infrastructure that enables frontier physical AI models to run smoothly on your robot. We optimize for low latency across the full stack: from the model, down to the GPU kernels, and over the network.
Our mission is to unlock a future of safe and intelligent robots doing meaningful work, and we think this is only possible by letting you deploy and run robot models in the cloud.
Reach out to us at [email protected] or visit our website at dreamscalelabs.com to learn more.
@KDog112358 @AntoineNeedGPUs @eternaI_entropy