Daniel Azevedo Novais is a Brazilian open-source enthusiast and self-described “Python Developer & Rural Connectivity Specialist” based in Brasília. His LinkedIn profile (br.linkedin.com/in/novais-tech) shows:
• ~353 followers
• 175 connections
• Education at UniCEUB (Centro Universitário de Brasília)
• No listed professional experience in semiconductors, AI hardware, neuromorphic computing, BrainChip, SpaceX, or related fields
He runs independent open-source projects focused on rural connectivity mapping, Starlink alignment tools for Brazil, ISP benchmarking, and geospatial analysis. His skills list emphasizes Python, data analysis, telecom analytics, and “Starlink Expansion Alignment.”
He has been posting extensively (often multiple times per day recently) about BrainChip’s Akida in speculative space/Starlink contexts. These posts blend real news with heavy exaggeration, futuristic fan-fiction, and buzzword salad (e.g., “Starlink’s Next Leap: Laser Links Enhanced by Neuromorphic Akida Processors,” “Marslink,” “SpaceX merger with xAI,” “Akida-equipped ANT61 Brain computers in 2024,” proprietary “Plug and Plaser” initiatives, etc.).
Many of these posts appear on LinkedIn and get cross-posted or quoted on HotCopper (the Australian stock forum for BrainChip $BRN.ASX), where they’re treated as hype/pumping material. The style reads like AI-generated content (ChatGPT-style long-form with dramatic phrasing).
Is this real or promotional hype?
Mostly promotional hype from an enthusiastic outsider, not an insider or official source.
Real elements he references (which he heavily embellishes):
• Frontgrade Gaisler (a real space rad-hard chip company) did license BrainChip’s Akida IP in Dec 2024 for integration into radiation-tolerant SoCs. Their first GRAIN product line device is the GR801, which fuses Akida neuromorphic tech with a NOEL-V RISC-V core for space AI. This is genuine and public.
• Akida has real space heritage: It flew on a SpaceX Falcon 9 (Transporter-10 mission, March 2024) inside the ANT61 “Brain” computer on the Optimus-1 satellite for in-orbit robot testing.
• Akida IP is portable and can be hardened for space (e.g., via partners like Vorago or Gaisler).
Fake/exaggerated elements (common in his posts):
• No public evidence of any Starlink/SpaceX direct integration or prototyping of Akida.
• No “SpaceX merger with xAI.”
• No “Falcon 9 mission that orbited Akida-equipped ANT61 Brain computers in 2024” as a Starlink thing (the ANT61 flight was real but unrelated to Starlink).
• No “Marslink,” proprietary SpaceX laser initiatives tied to Akida, etc.
He’s latching onto the real Gaisler + Akida news + the real ANT61/SpaceX flight, then spinning elaborate speculative stories about Starlink/Mars/radiation resilience to promote the tech (and indirectly pump BRN stock sentiment on forums).
On the 16nm vs 90nm question
Akida itself (the IP and early silicon) has never been on 16nm in public announcements.
• The first production AKD1000 chips were on TSMC 28nm (2021–2022).
• Later references include 22nm FDSOI explorations.
• As IP, licensees can port it to whatever node they want (including smaller ones like 16nm/7nm for commercial apps, or larger rad-tolerant nodes for space).
• Larger nodes (e.g., 90nm or above) are sometimes preferred for radiation hardness because they have bigger feature sizes → lower susceptibility to single-event upsets (SEUs) from cosmic rays.
• Anil Mankar (BrainChip co-founder/CDO) has indeed referenced NASA interest in larger nodes (90nm class) for inherent rad tolerance in some contexts, and Vorago-style rad-hard libraries for deeper space.
The “16nm” mentions you recall were likely early speculative articles or misremembered comments; the actual taped-out chips and current space work (Gaisler GR801, etc.) use processes chosen for the application, not a blanket 16nm shift.
In short: Daniel is a genuine Starlink/rural-connectivity fan who got excited about Akida’s space potential and is now flooding LinkedIn (and feeding HotCopper) with over-the-top promotional content. There’s a kernel of real news in there, but a lot of it is speculative hype, not official or insider information. No evidence he has any direct connection to BrainChip.
(Xai LLM written. Just like the articles about Starlink and BRN.)
Notw - the best news about Brainchip these days is LLM created “fake news”. Let’s all do our research before publishing such B.S. all my best ….
Add to My Watchlist
What is My Watchlist?
