Skip to content

Brief · 23 June 2026

What changed

A new video from a Chinese OEM shows its MOYA humanoid robot walking, gesturing, and responding with a claimed 92% human likeness rating, marking the first public demo of the platform in over a week. (https://www.youtube.com/watch?v=KdeO-D0tZD0)

One number

92%

Human‑likeness rating claimed for MOYA, suggesting the robot looks and moves almost like a person.

source ↗

Still vapor

NVIDIA touts its Rubin liquid‑cooled data‑center design as “running hotter to use a lot less water,” implying massive power savings. In practice, higher inlet temperatures raise fan and pump loads, and the water‑use reduction is modest compared with the extra cooling infrastructure required.

The MOYA demo is the day’s only concrete hardware shift. The robot walks on a two‑legged gait, lifts objects, and engages in short dialogues, all while a side‑by‑side comparison claims 92 % human likeness. The OEM has not disclosed the compute stack, but the visual fidelity suggests a high‑end edge AI accelerator—likely an NVIDIA Jetson Thor or a custom ASIC—paired with advanced sensor fusion. No power or price numbers were released, leaving buyers to guess the total cost of ownership. Compared with Boston Dynamics’ Atlas or Unitree’s G1, MOYA’s visual realism is a step forward, but the lack of performance data makes it a risky spec for a production line.

Meanwhile, NVIDIA’s water‑saving claim for its Rubin design received a skeptical note from the community. Running hotter does reduce chiller load, yet the overall system still consumes comparable power, and the promised “pretty much all” water elimination is unverified.

The catalog remains static: 51 rigs verified, 20 NVIDIA units, and no new entries in the last 30 days. Operators should treat the MOYA demo as a proof‑of‑concept rather than a ready‑to‑buy solution, and watch for a formal spec sheet before allocating budget.

Sources are listed below for verification.

Composed by the MadCoolStuff editor pipeline · Groq · openai/gpt-oss-120b · 2026-06-23

Tags

What we read