BRN Discussion Ongoing

[Dolci]

I suspect the Fools will be releasing there updated research paper soon enough. The stage is set for something positive from Sean I hope. Those Fools are not the guys to BRN..


MegaportX.

yeah, to get their free RSU at the AGM... anyway I wouldn't rely too much on them, ..........what is really needed is revenue & that won't show up till next quarterly if there is somthing there, & I would say an IP license that would be a miracle now ..........

 

[Dolci]

Ease up, Dolci
No need for that.

 

[Guzzi62]

Isn’t it strange that the only traction the company is getting is via physical AKD1000 chips, the same chips that Vianna and Hehir said were “too narrow” in their application. No one knows where the chips are coming from. Were they in stock from the TSMC run way back? I’m starting to think that turning a ground breaking nascent technology into an IP only business plan was not the greatest idea. AKD2000 seems to be smoke and mirrors no matter how good the simulation software is. I think potential customers are baulking at the idea that they have to design and fab a chip themselves in order to incorporate our IP. It’s a big ask. Technology - priceless. Execution - woeful. Let’s hope a rabbit gets pulled from a hat. Onward.

There is some truth in that statement.

How many companies design their own chips?

Answer by Goggle:

While not every company designs their own chips, a significant number of large tech companies like Apple, Amazon, Google, Microsoft, Meta, Tesla, and Baidu have started designing their own chips, particularly for specialized applications like artificial intelligence and data centers, making the number of companies designing their own chips relatively small but growing in recent years; most companies still rely on established chip manufacturers like Intel, Samsung, and TSMC for their chip needs.

Key points about chip design:

• Large tech giants leading the trend:
  Companies like Apple, Amazon, and Google are actively designing their own chips to gain more control over performance and efficiency for their specific needs.
• 
  
• Specialized chip design:
  Many companies are focusing on designing chips for specific applications like AI, cloud computing, and autonomous vehicles.
  
  
• 
  
• Cost and complexity:
  Designing custom chips can be expensive and requires significant technical expertise, which is why many companies still choose to buy off-the-shelf chips.
  
  
• 
  
• Manufacturing reliance:
  Even companies designing their own chips often rely on established chip manufacturers like TSMC to produce them.

Where BRN fits in I don't know to be honest.

I can understand why it must be very difficult to sell an IP.

The buyer must have a specific need for it and in huge numbers or they will loose money on the exercise.

This is how I understand it?

 

[Pom down under]

ah, one took the bait too early, as your 30% is just a drop in the Ocean where it is going .......

I know and I can’t wait until we hit the $10 mark, just hope I’m still alive

 

[CHIPS]

We all always forget about Tata Elxsi and what they wrote in their annual report. They will work with Akida for their healthcare products!!!
The money will come and everything will be fine. For all of us!

 

[manny100]

There is some truth in that statement.

How many companies design their own chips?

Answer by Goggle:

While not every company designs their own chips, a significant number of large tech companies like Apple, Amazon, Google, Microsoft, Meta, Tesla, and Baidu have started designing their own chips, particularly for specialized applications like artificial intelligence and data centers, making the number of companies designing their own chips relatively small but growing in recent years; most companies still rely on established chip manufacturers like Intel, Samsung, and TSMC for their chip needs.

Key points about chip design:

• Large tech giants leading the trend:
  Companies like Apple, Amazon, and Google are actively designing their own chips to gain more control over performance and efficiency for their specific needs.
• 
  
• Specialized chip design:
  Many companies are focusing on designing chips for specific applications like AI, cloud computing, and autonomous vehicles.
  
  
• 
  
• Cost and complexity:
  Designing custom chips can be expensive and requires significant technical expertise, which is why many companies still choose to buy off-the-shelf chips.
  
  
• 
  
• Manufacturing reliance:
  Even companies designing their own chips often rely on established chip manufacturers like TSMC to produce them.

Where BRN fits in I don't know to be honest.

I can understand why it must be very difficult to sell an IP.

The buyer must have a specific need for it and in huge numbers or they will loose money on the exercise.

This is how I understand it?

I think its slow market adoption of AI at the Edge, especially AKIDA style rather than IP that is the issue.
AKIDA is not a traditional AI Edge solution and no one ever got fired for choosing to go with IBM or NVIDIA. ..
Just need that 1st decent deal to drop.

 

[Papacass]

We all always forget about Tata Elxsi and what they wrote in their annual report. They will work with Akida for their healthcare products!!!
The money will come and everything will be fine. For all of us!

You’re right. Plus they definitely have the resources to produce a gazillion chips. Megachips is the other obvious one. PVDM said years ago words to the effect that shareholders did not fully understand the worth of the Megachips deal. Here’s hoping. Mario Kart anyone?

 

[IloveLamp]

Nice to have ARM as a partner

Arm on LinkedIn: We’re proud to partner with OpenAI, SoftBank Group Corp., and SoftBank…

We’re proud to partner with OpenAI, SoftBank Group Corp., and SoftBank Corp. on Cristal intelligence - a new initiative to develop Advanced Enterprise…

www.linkedin.com

 

[Earlyrelease]

Isn’t it strange that the only traction the company is getting is via physical AKD1000 chips, the same chips that Vianna and Hehir said were “too narrow” in their application. No one knows where the chips are coming from. Were they in stock from the TSMC run way back? I’m starting to think that turning a ground breaking nascent technology into an IP only business plan was not the greatest idea. AKD2000 seems to be smoke and mirrors no matter how good the simulation software is. I think potential customers are baulking at the idea that they have to design and fab a chip themselves in order to incorporate our IP. It’s a big ask. Technology - priceless. Execution - woeful. Let’s hope a rabbit gets pulled from a hat. Onward.

Papa you raise a good point and one that hasn't been discussed in a long time. That being how many chips did they make and now how many are left. Whilst I know this discussion could be taken down a negative path I want to focus on the positives and likely next steps, well from an ill informed retail investor. The first question arising from how many were left is we don’t know how many were produced and without speculation as we did to death back then that was never disclosed. So what next if no IP and no chips what do we have. Maybe that was why the sudden capital raise was it for another batch of Ak1000, this time maybe produced in the good old USA to prove it can be made away from TSMC and to get other chip makers to see our product whilst being able to continue to provide the “show and tell” market hoping they buy the IP as intended. If that is the case and there is not yet an IP buyer for AK 2000+ could it be they make both?. Yes purely tantalising thought but based on nothing but hope but love to hear thoughts of others. And finally for those neigh sayers I am still holding from the original listing and have only bought and still buying as my pennies allow but as a spec stock which I like many others think there is too much smoke for there not the be fire…. Eventually.

 

[Bravo]

........dont worry, plenty of time for that super of yours as I said before to you mid-twenties are on the cards ...lol

Community Alert!!!!

 

[7für7]

What would I do without the positive posts? Thank you all!
🥹🥹🥹🥹

 

[Frangipani]

Hi Terroni,

your post reminded me of two comments by Laurent Hili I had spotted before going on holidays, but forgot to post a screenshot of:

View attachment 69143

OBPMark

(On-Board Processing Benchmarks)

obpmark.github.io

Also, Laurent Hili’s comment that I posted earlier today points to Akida technology being envisioned for a lunar lander initially - an ESA mission to the Moon, not to Mars.


I believe the Argonaut lunar lander is what we might be looking at - ESA recently announced that it is targeting 2031 for the lander’s first mission to the Moon’s surface.

Argonaut: Europe's lunar lander programme

The European Space Agency (ESA) lunar lander programme, Argonaut, represents Europe’s autonomous, versatile and reliable access to the Moon. The programme is a key part of ESA’s lunar strategy and will support future robotic and crewed missions, contributing to international efforts to establish...

www.esa.int

SCIENCE & EXPLORATION

Argonaut​

46942 VIEWS221 LIKES
ESA / Science & Exploration / Human and Robotic Exploration / Exploration
Argonaut is Europe’s autonomous access to the Moon, allowing us to play a major role on the surface of our natural satellite. The lunar lander is being designed for a series of missions with many options for its payloads – from cargo and infrastructure delivery to scientific operations, a rover or a power station, Argonaut is being designed as a versatile access to the Moon.


Argonaut elements
Argonaut will launch on an Ariane 6 rocket in a direct flight to the Moon. An Argonaut mission from launch to landing could take from a week to a month, depending on orbits and mission design. No area is off-limits for Argonaut, the spacecraft will be able to land at any region on the Moon.

The Argonaut spacecraft has three main components: the lunar descent element that takes care of flying to the Moon and landing on target, the cargo platform element that is the interface between the lander and its payload, and finally, the element that mission designers want to send to the Moon.

Adaptability is key in the Argonaut design, the cargo platform element is designed to accept any mission profile: cargo for astronauts near the landing site, a rover, technology demonstration packages, production facilities using resources on the Moon, a lunar telescope or even a power station.

Argonaut is continuing Apollo and Artemis tradition to name lunar missions after Greek mythology. Argonaut is the name given to the sailors of the Argo ship who took Jason on the quest to find the golden fleece. ‘Argonaut’ means ‘sailors of the Argo’ and the individual missions using ESA’s lunar lunar delivery service will be named after the individual mythical Argonauts.

Space agencies have much in store on the Moon for humankind and Argonaut is offering autonomous European access as well as strong possibilities for partnership. As part of the Artemis programme, ESA is participating in NASA’s Orion service module that ferries astronauts, and Europe is supplying modules to the international Gateway in lunar orbit. Argonaut could be an independent complement to the Artemis programme offering cargo delivery and more.

Leading Moon exploration with strong partners​

Argonaut

The lunar lander is being designed with versatility in mind as a strong part of ESA’s lunar strategy and Argonaut could be included in the Artemis programme to deliver cargo, rovers and more, or as stand-alone scientific missions.
The Apollo astronauts never lived and work through the lunar night – a night on the Moon is 14 days long and temperatures on the surface plummet to a chilling –150°C.

One capability of the lander will be to bring a sizable share of the food, water, air, and equipment for a crew of four women and men working on the Moon through the night. The Argonaut lunar descent element will be able to bring up to 2100 kg worth of cargo platform element and payload to the Moon – equivalent to a camper van. The European lander can deliver much more than simple cargo: it could function as a survival kit for the explorers of our new frontier.

Sustaining Moon science for decades​

What is ESA’s Moonlight initiative?
Access the video

For scientists, the Moon’s qualities of being interesting, close, and useful are an enticing motivation to understand our place in the Universe. Argonaut will allow going beyond short excursions with just a handful of instruments. Driven by scientists’ priorities, the European lander would allow samples to be analysed from previously unexplored and hard-to-get-to regions of the Moon.

Argonaut will use ESA’s Moonlight navigation and telecommunication capabilities around the Moon, allowing for fast communication with the Gateway and Earth to return scientific or operational data, as well as location finding for the automated landing – Argonaut will be able to land with an accuracy of less than 100 m.



Terrae Novae destinations: Moon

The teams in ESA, international partner agencies, European industry, and in the scientific institutions are on this journey together that will bring benefits in the form of inspiration, innovation, and economic growth to all Europeans.
Argonaut was approved at ESA’s Council at ministerial level in 2022 and is now in development. Five missions are foreseen in the next decade, fitting in with ESA’s Terrae Novae strategy for human and robotic exploration.

Technical details​

Launcher	Ariane 64
Launch Site	Kourou, French Guiana
Mass on Earth	10 000 kg
Mass on the Moon without cargo	1600 kg
Delivered mass (CPE+ payload)	up to 2100 kg
Size	4.5 m in diameter, up to 6 m tall
Landing accuracy	50–100 m
Mission types	Multiple and diverse: Cargo logistics Science & exploration missions technology demonstration packages Power generation, storage and distribution systems in-situ resource utilisation plants and more...

A CAD model for illustration purposes of the Lunar Descent Element is available here. Please consider mentioning ESA if any material is made using this CAD model

---

Update:

I just noticed that Alf Kuchenbuch has meanwhile edited his post, adding (“That is my dream.”), and that Laurent Hili posted an identical comment twice in response to a) that post resp. b) Alf Kuchenbuch’s comment under yesterday’s post by Steve Thorne, making it crystal-clear that it will still take time until Alf’s dream might come true (and at the same time also stresses that it is by no means 100% guaranteed - he literally says BrainChip “could” play a role, not “will” play a role and that it is one of the technologies they are seriously looking at - I am not sure whether he wanted to express the alternatives they are looking at are non-neuromorphic or whether that could also point to one or more neuromorphic competitors being evaluated in parallel by ESA…)

“We are actively working on it to make it a reality (…) #brainchip could play a role and is one of the technology [sic] we are seriously looking at.”

Sounds to me as if either someone from ESA or BrainChip requested Alf Kuchenbuch to edit his post in order to clarify we’re not there, yet, and to stop the rumour mill…


View attachment 69145



View attachment 69144

In September , I speculated that Laurent Hili from ESA may have been hinting at the future Argonaut Lunar Lander, when he posted on LinkedIn that neuromorphic computing was being considered by ESA for several use cases, including a lunar lander.

Today I noticed Alf Kuchenbuch celebrating a 30 January LinkedIn post by ESA, in which they announced they had signed a contract with Thales Alenia Space in Italy to lead European aerospace companies in building the Argonaut Lunar Descent Element, ESA’s first lunar lander, to be delivered in 2030.

#Argonaut: Europe's first lunar lander. | European Space Agency - ESA

#Argonaut: Europe's first lunar lander. ESA’s Argonaut will establish Europe's independent access to the Moon, with regular missions planned from the 2030s. Thales Alenia Space in Italy leads the European consortium to build this versatile lunar descent module.👉https://lnkd.in/e3dYDA7V | 23...

www.linkedin.com

Argonaut: a first European lunar lander

The European Space Agency (ESA) has signed a contract with Thales Alenia Space in Italy to lead European aerospace companies in building the Argonaut Lunar Descent Element, ESA’s first lunar lander.

www.esa.int

SCIENCE & EXPLORATION

Argonaut: a first European lunar lander​

30/01/20255734 VIEWS102 LIKES
ESA / Science & Exploration / Human and Robotic Exploration

The European Space Agency (ESA) has signed a contract with Thales Alenia Space in Italy to lead European aerospace companies in building the Argonaut Lunar Descent Element, ESA’s first lunar lander.

A Moon exploration scenario

ESA’s Argonaut represents Europe’s autonomous and versatile access to the Moon, supporting international exploration endeavours on the lunar surface. From the start of the next decade, the spacecraft will launch on regular missions to the Moon. These could deliver infrastructures, scientific instruments, rovers, technology demonstrators and vital resources for astronauts on the lunar surface such as food, water and air.

Argonaut will be able to survive the harsh lunar night and day for five years, providing a key capability for sustainable lunar exploration.

A mock-up of the Argonaut lunar descent element on show at the LUNA facility inauguration

Argonaut is a cornerstone of ESA’s lunar exploration strategy and is designed to work seamlessly with ESA’s Lunar Link on the Gateway and Moonlight communication and navigation systems.

Argonaut is one of Europe’s contributions to international lunar programmes, particularly NASA’s Artemis programme and commercial lunar lander services, contributing to establishing a permanent and sustainable human presence on the Moon.

The spacecraft for an Argonaut mission has three main components: the lunar descent module that takes care of flying to the Moon and landing on target, the payload, and the cargo platform that acts as the interface between the lander and payload.

Argonaut elements

Thales Alenia Space in Italy will be leading the European consortium to build the lunar descent module; the rest of the core team includes Thales Alenia Space in the United Kingdom and France, and OHB.

The team will deliver the Argonaut Lunar Descent Element in 2030 for the first operational mission, ArgoNET, expected in 2031.

By the end of 2026, the industrial consortium in charge of using the first Lunar Descent Element will be selected.

"The Argonaut contract signature is a pivotal moment for Europe’s lunar exploration ambitions," says Daniel Neuenschwander, ESA Director for Human and Robotic Exploration.
Argonaut mission patch

"This first-of-its-kind European lunar lander demonstrates ESA’s dedication to advancing our industrial capabilities in deep space exploration. Argonaut will enable Europe to contribute meaningfully to international partnerships, while paving the way for a sustainable human presence on the Moon. Europe is on its journey to the Moon and has broken the ground towards European autonomy in exploration," he adds.

 

[Frangipani]

So Sriniketh seriously saw some secret silicon?

“ Recently discovered BrainChip’s Akida 2.0 neuromorphic chip and I’m blown away! ”

Wouldn’t we all be, if that had really happened?!

The above post appears to be the (over?-)enthusiastic ASU Master student’s (ChatGPT-aided?) application for a summer internship with BrainChip on a LinkedIn form factor.

 

[Frangipani]

On 28 January, Jörg Conradt, who leads the Neuro Computing Systems Lab at KTH Royal Institute of Technology in Stockholm, gave a presentation titled "Energy-smart Neuromorphic Sensing and Computation for Future Space Applications” at the AI for Space Applications Workshop that took place at KTH’s Digital Futures hub.
The workshop was held as part of the ASAP project:

View attachment 77025




View attachment 77021

While the first presentation slide didn’t look too exciting from a BRN shareholder’s perspective…

View attachment 77022

… we actually did get a mention later on, when Jörg Conradt touched upon a number of neuromorphic startups…

View attachment 77023

…although he didn’t get it quite right. Not only did he mix up the names of our company and its neuromorphic processor - he also misspelled and pronounced the latter as “Aikida”. And since when does BrainChip sound like a place name?

To be fair, though, he actually gave us an “out of this world” mention by pointing to what he mistook to be our company’s name and telling his audience (from 14:50 min):

“In fact, this company develops IP, and a very big Sweden-based space company has recently teamed up with Aikida [sic] to develop hardware - computing hardware - that they can send into space, that is potentially very robust against radiation and other effects.
So we have a common project starting on what to do with that hardware, and that’s my point of contact into space.”

It sounds as if Jörg Conradt’s lab at KTH will somehow be involved with whatever Frontgrade Gaisler has planned for Akida…
All the more reason he ought to familiarise himself with A & B ASAP.

Encouragingly, Akida made another appearance shortly after (this time with impeccable orthography) - if only for a split second, before Jörg Conradt moved on to the next slide. It was in the context of a research project titled Neuromorphic Edge Computing for Urban Traffic Monitoring in the city of Stockholm, funded by Digital Futures, a cross-disciplinary research centre established by KTH Royal Institute of Technology, Stockholm University and RISE Research Institutes of Sweden. Akida was listed alongside SpiNNaker 2 and Loihi 2 under “neuromorphic chips” and even had the honour of providing an exemplary image for that category.

View attachment 77027


View attachment 77030

View attachment 77031

Jörg Conradt didn’t specify whether or not the different neuromorphic processors were going to get benchmarked against each other, but I assume that’s the plan.


Here is some further information I found about said project that runs from January 2024 to December 2025 and “estimates a 100x reduction in power and a 20x reduction in installation cost.”

Edge computing for urban traffic monitoring — Digital Futures

Embedding low-power event cameras with low-power neuromorphic local computation provides a uniquely scalable solution for the growing need to monitor urban traffic and resource flows in real time...

www.digitalfutures.kth.se

View attachment 77028

View attachment 77029
View attachment 77032

I wonder whether this urban traffic monitoring project with the help of event-based cameras was somehow inspired by the 2023 tinyML Pedestrian Detection Hackathon submission (utilising Akida) from Cristian Axenie’s SPICES lab at TH Nürnberg. It doesn’t seem that far fetched to make a connection - after all, they know each other well: Jörg Conradt was Cristian Axenie’s PhD supervisor at TUM (Technical University of Munich).

In the light of my recent post about Jörg Conradt (KTH Stockholm), his upcoming talk at the University of Cambridge (and online via Zoom) titled “Low-power embedded event-based vision processing for low-latency robotics” could be rather interesting…

talks.cam : Low-power embedded event-based vision processing for low-latency robotics

 

[rgupta]

Papa you raise a good point and one that hasn't been discussed in a long time. That being how many chips did they make and now how many are left. Whilst I know this discussion could be taken down a negative path I want to focus on the positives and likely next steps, well from an ill informed retail investor. The first question arising from how many were left is we don’t know how many were produced and without speculation as we did to death back then that was never disclosed. So what next if no IP and no chips what do we have. Maybe that was why the sudden capital raise was it for another batch of Ak1000, this time maybe produced in the good old USA to prove it can be made away from TSMC and to get other chip makers to see our product whilst being able to continue to provide the “show and tell” market hoping they buy the IP as intended. If that is the case and there is not yet an IP buyer for AK 2000+ could it be they make both?. Yes purely tantalising thought but based on nothing but hope but love to hear thoughts of others. And finally for those neigh sayers I am still holding from the original listing and have only bought and still buying as my pennies allow but as a spec stock which I like many others think there is too much smoke for there not the be fire…. Eventually.

As per my memory brainchip paid 8 million plus to socionext for akida 1000 and we know only about 1st product batch as market announcement. So how were the arrangements and what were the arrangements no one knows, but I assume there must be some agreement on batch production of those chips and that is how we are keeping getting akida 1000.
But frankly speaking how many akida 1000 you reckon brainchip may had used till date, coz the revenue is not showing about use of akida 1000. The only bulk use case may be edge boxes but still we get no revenue from them either.
Dyor

 

[Frangipani]

Interesting that the image of Intel Loihi appears in a USB form factor. Wasn’t aware Intel reduced Loihi down to this size. Last I checked it fit in a brief case

FYI: Loihi has been available for research in a USB form factor (aka Kapoho Bay) for more than five years… In early 2022, the NASA Ames Research Center even launched one of them into Low-Earth Orbit in that silvery spacesuit, aboard a 3U CubeSat.

https://niceworkshop.org/wp-content/uploads/2019/04/NICE-2019-Day-4c_Loihi-Overview.pdf

https://ntrs.nasa.gov/api/citations/20230011534/downloads/BrainStack_SSC23_Presentation.pdf

 

[TheUnfairAdvantage]

CyberNeuro-RT

www.quantumventura.com

An IoT, AI/ML-driven, highly-scalable, real-time network defense & threat intelligence tool with CPU, GPU or low-power neuromorphic chip deployment

At-the-edge Neuromorphic Processing

• Two offerings from the leading neuromorphic developers: Intel and Brainchip
• Small form factor, magnitudes less power consumption than GPU
• On-chip learning for deployment network specific attack detection

Intel Loihi

Brainchip Akida

FYI: Loihi has been available for research in a USB form factor (aka Kapoho Bay) for more than five years… In early 2022, the NASA Ames Research Center even launched one of them into Low-Earth Orbit in that silvery spacesuit, aboard a 3U CubeSat.

https://niceworkshop.org/wp-content/uploads/2019/04/NICE-2019-Day-4c_Loihi-Overview.pdf

View attachment 77153

https://ntrs.nasa.gov/api/citations/20230011534/downloads/BrainStack_SSC23_Presentation.pdf

View attachment 77155

Thanks Frangipani,

Upon review of your post I concur that Intel have in fact presented to the market a Loihi option in the form of a USB “for research” only. How that compares to Akida I’m not sure. I am not still not aware if it commercially available. I should have been more specific in my statement. Apologies. My post was referring to a commercial offer of service.

Therefore, if Cybernuro is making an offering of cyber security to the market now, which commercially available low power neuromorphic chip in USB form factor would they make that offering on?

 

[Bravo]

Goodness me! What have we here?!!!!

The below article includes quotes from Steven Meier of the Naval Research Centre.The importance of what this might mean to us can be further ascertained in light of the collaboration between the AFRL and the NRL, as is outlined in this extract from the 2023 Space Symposium..


EXTRACT

AFRL showcases key programs at 2023 Space Symposium

The Air Force Research Laboratory, or AFRL, will feature several space-related technologies and programs during the 38th Space Symposium at the Broadmoor, Colorado Springs, Colorado, April 17-20, 2023.

www.dvidshub.net

Pentagon​

Power generation challenges could overshadow Stargate AI initiative​

By Courtney Albon
Saturday, Jan 25, 2025

President Donald Trump and AI industry leaders and investors on Tuesday unveiled a five-year, $500 billion initiative to boost AI development in the U.S. (Julia Demaree Nikhinson/AP)
While the Defense Department is likely to benefit from OpenAI’s announcement this week that it would invest half a trillion dollars to build new artificial intelligence data centers around the country, Pentagon officials warned that the U.S. lacks the energy resources and computing power to support the new infrastructure — and solving that problem won’t be easy.
OpenAI announced the project, dubbed Stargate, on Tuesday, pledging an initial $100 billion — plus another $400 billion over the next five years — to build new AI infrastructure across the U.S. and create “hundreds of thousands of American jobs” in the process. Early funders include Softbank, OpenAI, Oracle and MGX — a technology investment firm based in the United Arab Emirates — and OpenAI will partner with Oracle, Microsoft, Arm and NVIDIA on technology development.

During a press conference Tuesday at the White House, President Donald Trump called the effort a “monumental undertaking” and said the White House would support the project, in part, through issuing emergency declarations, though he didn’t expand on details.
The Defense Department has an ambitious vision for using AI across a range of military missions, including data collection, intelligence analysis, campaigning and logistics. But running those tools and applications takes more computing power and space than DOD has access to.
Roy Campbell, deputy director of advanced computing in the Office of the Undersecretary of Defense for Research and Engineering, said Thursday that many times, bases outside of the U.S. don’t have the computing power they need to retrain new AI tools.

“In some cases, for you to be able to handle a situation a forward operating base can’t handle, you have to kick that back to [the continental United States] and use the DOD supercomputing centers that we have there,” he said during a panel at the Potomac Officers Club’s annual Research and Development Summit in McLean, Virginia.
Jeff Waksman, who’s leading an effort in the Pentagon’s Strategic Capabilities Office to develop a mobile nuclear reactor, said the strain that technologies like AI and high-power computing place on the electric grid raises questions about who should have access to data and how to mitigate the risk of blackouts.
“This is not a problem that industry or the DOD can figure out by itself. It’s about the nation’s grid as a whole,” said Waksman, who spoke on a panel with Campbell. “It’s probably the most underrated challenge of this huge $500 billion announcement.”
Waksman’s nuclear reactor program, known as Project Pele, offers one answer to that challenge: using nuclear power to source energy for AI computing.

The effort, initiated in 2019, aims to demonstrate the first-ever U.S. prototype of a portable nuclear reactor within five years. The mobile reactor, which the department estimates could deliver one to five megawatts of electrical power over a minimum three-year operating life, would support DOD’s growing energy needs by providing power to austere locations.
The Pentagon broke ground at the Project Pele test site at the Idaho National Laboratory last September and plans to begin assembling the reactor — built by BWXT Advanced Technologies — as soon as next month. The department aims to demonstrate the technology in 2026.
“It’s going to be the first generation portable nuclear reactor built anywhere in the world, outside of China,” Waksman said. “It’s very much not a paper project anymore.”

Another potential solution to the AI power problem is making processors more effective at crunching data. Steven Meier, associate director of space technology at the Naval Research Center, said his lab is exploring the use of more efficient neuromorphic processors that can be 100 times more efficient than a standard processor. Essentially, neuromorphic processors take up less space, work faster and use less energy.
“There’s huge gains to be made in terms of neuromorphic processors making AI and [machine learning] more accessible on autonomous vehicles of all shapes and sizes,” Meier said at the conference.

Power generation challenges could overshadow Stargate AI initiative

Pentagon officials said the U.S. lacks the energy resources and computing power to support the new infrastructure — and solving that problem won’t be easy.

www.defensenews.com

 

[Quercuskid]

Up and running!

 

[Bravo]

EXTRACT FROM BELOW ARTICLE

The Stargate Project: A $500 Billion Investment in U.S. AI Infrastructure​

ByteBridge
·
Jan 22, 2025

https://medium.com/m/signin?actionU...a6---------------------clap_footer-----------

Introduction​

The field of artificial intelligence (AI) has seen unprecedented growth and investment in recent years, with the latest development being the groundbreaking Stargate Project announcement. This report provides an overview of current AI infrastructure investments and collaborations among leading tech companies, with a special focus on the newly announced Stargate initiative. This research was solely done and written by Kompas AI

The Stargate Project​

Overview​

The Stargate Project represents a groundbreaking AI infrastructure initiative:

• A $500 billion investment announced in January 2025
• A collaborative effort between OpenAI, Oracle, NVIDIA, and SoftBank
• Focused on developing next-generation AI infrastructure across the United States

Key Partners and Roles​

• OpenAI:
  - Contributing AI technology expertise, advanced models, and ethical AI development
  - Developing large language models and reinforcement learning algorithms
  - Ensuring responsible AI deployment and addressing bias in AI systems
• Oracle:
  - Providing cloud infrastructure, data center capabilities, and database management systems
  - Implementing role-based access control systems for secure data handling
  - Optimizing database performance for AI workloads
• NVIDIA:
  - Supplying cutting-edge GPU technology and AI accelerators
  - Designing custom ASICs and FPGAs for AI-specific tasks
  - Developing scalable software frameworks for distributed AI computing
• SoftBank:
  - Leading the investment strategy, international coordination, and business development
  - Facilitating partnerships with global tech companies and research institutions
  - Identifying potential applications across various industries

Technological Advancements and Projected Milestones​

The Stargate Project aims to achieve several key technological breakthroughs:

1. Development of exascale AI computing systems
   - Milestone: Achieve 1 exaflop performance by 2027
   - Challenge: Addressing energy efficiency and heat dissipation
   - Solution: Implementing advanced cooling systems and optimizing power usage
2. Creation of advanced neural architectures
   - Milestone: Deploy models with 1 trillion parameters by 2028
   - Challenge: Managing computational resources and training time
   - Solution: Utilizing distributed training across multiple data centers
3. Implementation of quantum-inspired algorithms
   - Milestone: Demonstrate 100x speedup in optimization tasks by 2029
   - Impact: Enhancing problem-solving capabilities in finance, logistics, and drug discovery
   - Challenge: Bridging the gap between quantum principles and classical hardware
   - Solution: Developing hybrid quantum-classical algorithms
4. Establishment of a nationwide, high-bandwidth, low-latency network
   - Milestone: Complete network infrastructure by 2030
   - Expected benefits:
   1) Enabling real-time collaboration on large-scale AI projects
   2) Facilitating seamless data sharing between research institutions
   3) Supporting edge AI applications with minimal latency
   - Implementation strategy:
   1) Leveraging existing fiber-optic infrastructure
   2) Deploying advanced 5G and future 6G technologies
   3) Establishing regional AI computing hubs connected via high-speed links

Addressing Exascale AI Computing Challenges​

• Scalability: Developing AI frameworks that efficiently utilize millions of cores
• Fault tolerance: Implementing robust error detection and correction mechanisms
• Data management: Creating advanced data pipelines for efficient processing of exabytes of information
• Algorithm adaptation: Redesigning AI algorithms to exploit massive parallelism

By integrating the expertise of each partner and addressing these challenges, the Stargate Project aims to revolutionize AI infrastructure and accelerate technological advancements across multiple domains.

Investment Allocation​

The $500 billion investment will be distributed across various phases and components:

40% ($200 billion) for hardware infrastructure:​

• Advanced GPUs and TPUs from NVIDIA and Google
• Custom ASICs for specific AI applications
• High-performance FPGAs for reconfigurable computing
• Exascale-capable CPUs and high-bandwidth memory systems
• Cutting-edge networking equipment for low-latency communication
• State-of-the-art cooling systems and power supply unit

30% ($150 billion) for software development and AI model training:​

• $50 billion for large language models and natural language processing
• $40 billion for computer vision and image recognition systems
• $30 billion for reinforcement learning and autonomous systems
• $20 billion for AI in healthcare and personalized medicine
• $10 billion for AI applications in finance and climate change

15% ($75 billion) for research and development of new AI technologies:​

• Explainable AI (XAI) and AI ethics
• Quantum AI and integration with classical systems
• Advanced neural architectures and neuromorphic computing
• AI for edge computing and Internet of Things (IoT)
• Cognitive AI and artificial general intelligence (AGI) research

10% ($50 billion) for talent acquisition and workforce development:​

• Global recruitment initiatives for diverse AI talent
• Partnerships with top universities and research institutions
• Establishment of AI-focused training centers and bootcamps
• Development of online learning platforms for AI education
• Creation of mentorship programs and internship opportunities

5% ($25 billion) for regulatory compliance and ethical AI governance:​

• Implementation of robust data privacy and security measures
• Development of AI auditing and certification processes
• Creation of ethical AI guidelines and governance frameworks
• Establishment of AI ethics review boards and oversight committees
• Collaboration with policymakers on AI regulation and standards

This detailed and comprehensive report was exclusively prepared by Kompas AI. AI-powered writing tools help generate similar high-quality reports quickly.

Project Timeline​

The Stargate Project is set to unfold over the next decade:

2025–2027: Initial infrastructure deployment and baseline model development​

• Establishment of primary data centers and computing clusters
• Development and training of foundational AI models
• Implementation of core ethical AI frameworks and governance structures
• Launch of talent acquisition and workforce development programs

2028–2030: Scaling of computing capacity and introduction of advanced AI applications​

• Expansion of infrastructure to achieve exascale AI computing capabilities
• Deployment of advanced AI applications across various industries
• Integration of explainable AI techniques in all major AI systems
• Establishment of nationwide high-bandwidth, low-latency AI network

2031–2035: Integration of quantum computing elements and nationwide network completion​

• Implementation of quantum-inspired algorithms for complex problem-solving
• Development of hybrid classical-quantum AI systems
• Full integration of quantum computing with AI infrastructure
• Completion of a secure, nationwide quantum-enabled AI network

This research is solely done and written by Kompas AI

Challenges and Risks​

Several potential obstacles have been identified for the Stargate Project:

1. Data Privacy and Security:
   - Implementation of advanced encryption protocols for data transmission and storage
   - Strict access controls and multi-factor authentication for all systems
   - Regular security audits and penetration testing
   - Compliance with international data protection regulations (e.g., GDPR, CCPA)
2. Environmental Impact:
   - Significant energy consumption from large-scale computing infrastructure
   - Mitigation strategies include:
   1) Transition to 100% renewable energy sources for all data centers
   2) Implementation of advanced cooling technologies (e.g., liquid cooling)
   3) Optimization of AI algorithms for energy efficiency
   4) Investment in carbon offset projects
3. Regulatory Challenges:
   - Varying AI regulations across states and countries
   - Establishment of a dedicated legal team to navigate complex regulatory landscapes
   - Proactive engagement with policymakers to shape AI governance frameworks
   - Development of flexible AI systems adaptable to different regulatory requirements
4. Job Displacement:
   - Empirical evidence suggests significant impact in sectors like manufacturing and retail
   - Mitigation strategies include:
   1) Large-scale reskilling and upskilling programs for affected workers
   2) Collaboration with educational institutions to align curricula with future AI workforce needs
   3) Investment in AI-human collaboration technologies to augment rather than replace human workers
   4) Support for entrepreneurship and new job creation in AI-related fields
5. Advanced AI System Risks:
   - Potential for unintended consequences or misuse of powerful AI systems
   - Risk mitigation measures include:
   1) Robust testing and validation protocols for all AI models
   2) Implementation of explainable AI techniques to enhance transparency
   3) Establishment of an AI ethics board to oversee development and deployment
   4) Regular risk assessments and contingency planning

Impact on AI Industry and Related Sectors​

The Stargate Project is anticipated to have wide-ranging impacts:

1. Acceleration of AI Research and Development:
   - Breakthroughs expected in:
   1) Natural language processing
   2) Computer vision
   3) Reinforcement learning
   4) Creation of new markets for AI-powered products and services
   5) Potential disruption of traditional industries
   
   - Fostering innovation in sectors such as:
   1) Autonomous vehicles
   2) Smart home technologies
   3) Personalized digital assistants
2. Industry Transformation:
   a. Healthcare:
   - Enhanced diagnostic accuracy
   - Personalized treatment plans
   - Improved early disease detection
   
   b. Finance:
   - Advanced fraud detection
   - Automated trading systems
   - AI-powered personalized financial advice
   
   c. Manufacturing:
   - Predictive maintenance
   - Optimized supply chains
   - Enhanced production processes through AI-powered robotics and IoT integration
3. United States as a Global Leader:
   - Establishment of cutting-edge AI data centers
   - Development of advanced computing facilities across multiple states
   - Potential for breakthrough discoveries using quantum-inspired algorithms in:
   1) Drug discovery
   2) Materials science
   3) Cryptography
4. Addressing Global Challenges:
   - Optimizing energy consumption
   - Improving climate forecasting
   - Enhancing environmental monitoring for climate change mitigatio
   - Advancing early disease detection and outbreak prediction
   - Supporting personalized medicine through AI-driven health monitoring systems
5. Economic Impact:
   - Significant contribution to the global AI market (Expected to reach $190.61 billion by 2025)
   - Creation of numerous high-skilled technology jobs
   - Driving economic growth across AI-related sectors

Impact and Future Prospects​

The Stargate Project is anticipated to generate significant economic and technological impacts:

1. Job Creation:
   - Thousands of high-skilled technology positions
   - Development of new technology corridors across multiple states
2. Enhanced AI Capabilities:
   - Improved research and development infrastructure
   - Focus on building state-of-the-art AI data centers
   - Development of advanced computing facilities
   - Establishment of strategic AI research hubs
3. Market Projections:
   - 75% of enterprises expected to adopt AI by 2027
   - Companies implementing AI engineering practices projected to outperform peers by 25% by 2026

These advancements are poised to drive innovation and competitiveness in the AI sector, potentially reshaping industries and technological landscapes across participating regions.

Conclusion​

The announcement of the Stargate Project marks a historic moment in AI infrastructure development, representing the largest private sector investment in AI infrastructure to date. This initiative, combined with existing investments from major tech companies, signals a transformative period in AI technology development and implementation.

References​

1. Trump announces up to $500 billion in private sector AI infrastructure investment