As an Australian by birth, I keep an eye on the ASX with the help of my mother, who dabbles in resources and tech, and called me today to ask about something she’s been in for a few years, much to the deteroration of her profits..

Brainchip (BRN.ASX) is, in her words, “Some sort of AI thing, let me see if I can figure it out..”

The website says the following:

“[Product name] delivers a sensor-agnostic, host-flexible solution that enables real-time detection, classification, and understanding of the physical world.”

Oh boy.

My mother is confused by this, naturally, but more confused that a company that recently raised $35m is down to $13m in treasury and is now contacting her, asking her to buy into a $2m financing at a near 52-week low.

“It’s a $300m company, so why do they need $2m, and at that price?”

“Mum, how much are you into this bullshit for,” I said.

If you hang around the pennies long enough, you start to see certain species repeat. There’s the lifestyle junior, the zombie miner, the biotech with a petri dish of hope, and then there’s the tech aspirationalist. They’re usually harmless; a couple of engineers, a neat idea, a chip or a circuit, and a fantasy about licensing revenue that never quite arrives.

Every now and again, a dreamer gets a real audience, likely by being in the right sector at the right time. Money pours in, the cap balloons, press releases multiply, and suddenly you have a three hundred million dollar story built on the promise of a disruption that always seems to be happening next year.

That brings us to BrainChip.

To be clear, BrainChip is not a scam, per se. They have real engineers, real patents, real silicon, and real partners. They also have a real idea. Neuromorphic computing is not pseudoscience. It is a legitimate approach to replicating the way the human brain processes events and patterns. But they don’t seem to be the guys who will commercialize it.

The idea is appealing. Instead of shoving AI processing into massive cloud servers running hot enough to grill steaks on, you bring it all on device. No latency, no bandwidth choke, no privacy issues. It’s the holy grail for cars, sensors, robotics, wearables, and a hundred other markets. If it all worked as advertised, you’d be printing money.

But here’s the part nobody seems to want to admit: there is a Grand Canyon between having an idea and turning that idea into a scalable commercial business. And BrainChip, for all its glossy decks and swirling TAM charts, is still standing on the wrong side of that canyon.

It’s not outright bullshit, but it is absolutely classic small-cap AI puffery.

The Brainchip investor deck is actually a marketing deck, not an operational update. It’s full of sweeping TAM numbers, no revenue disclosures, no customer traction proof, no product performance benchmarks, and no timelines. All of the information is possible, but almost none of it is verifiable.

Let me go point by point so you get a real investor-focused breakdown.

WHAT THE DECK ACTUALLY CONTAINS

It’s 100% high-level marketing positioning:

• Buzzwords
• Big TAM/SAM charts (pg 4–6)
• Partner logos (pg 10)
• Generic AI megatrend talk (edge AI, LLMs, neuromorphic computing)
• A leadership slide
• A disclaimer that basically says “don’t trust anything in here”

What it does not contain:

• Financials
• Revenue
• Contracts
• Design wins
• Unit shipments
• Roadmap
• Case studies
• Benchmarks against competitors
• Costs or burn rate
• ARR or licensing traction
• Chip availability

That’s a huge red flag for anything calling itself an Investor Update.

THE BIGGEST RED FLAGS

A. Zero numbers that actually matter

A real investor update shows:

• cash position
• runway
• margins
• customer logos tied to real contracts and details of those contracts
• quarter-over-quarter traction
• roadmap with dates

This deck has none. It only has market size projections, which are meaningless without execution.

B. TAM (Totally Accessible Market) worship (pg 4–8)

A deck that keeps repeating $163B and $84B markets is trying to hide the fact it has no customers.

Let’s be clear, I project that, by 2030, Charlize Theron will determine she should be romantically involved with me, and that’s key because getting Charlize Theron to make out with you is, in my mind, a billion dollar opportunity.

Am I properly equipped to have any sort of success in penetrating that market?

I mean, no, not at all. But WHAT A MARKET.

Every AI startup says this stuff:

• “Edge AI is exploding”
• “LLMs on device are coming”
• “Autonomy + agentic LLMs = explosive growth”

All technically true.

All strategically useless without:

• product specs
• reference designs
• customer traction
• revenue

Not that this information doesn’t exist – you can dig in on the website and track it down – but when you do, the info is… weak.

Think a 14 second video with 6 seconds of titles and fuzzy images without context weak.

C. Neuromorphic computing hype without substance (pg 6)

They show a YOLE forecast claiming neuromorphic computing goes from $28M → $8.3B by 2034. Cool story, only neuromorphic has been “about to break out” for 20 years.

No benchmarks. No proof their product, Akida, outperforms anything.

And no comparisons to those that do, namely:

• Syntiant
• SiFive
• Nvidia Jetson
• Edge TPU
• Qualcomm AI engines
• Mythic (RIP)

A company confident in its tech always shows benchmarks.

D. Partner slide is highly ambiguous (pg 10)

Absolute logo salad.

This slide includes:

• ARM
• Nvidia
• Mercedes-Benz
• NASA
• ESA
• AFRL

But nowhere does the deck say:

• whether these are paying customers
• whether it’s research partnerships
• whether they are simply members of the same consortium
• whether they bought anything

Small AI companies frequently imply major partnerships by putting logos in proximity. This is classic “guilt by association” marketing and you don’t do it if you have REAL DEALS.

E. No new product news

For a company with “next gen Akida IP” they:

• don’t describe it
• don’t show specs
• don’t show power comparisons
• don’t show latency
• don’t show competitive matrix

If you’re trying to license IP, you always show something measurable to licensees.

SO DO THEY HAVE ANYTHING?

From the most recent financials:

• Cash inflow from customers in the current quarter of US$1.4M was higher than the prior quarter (US$0.14M).
• Total payments to suppliers and employees of US$4.4M in the current quarter were lower than the prior quarter (US$4.9M) due to the payment of employee short-term incentive payouts in Q1-25.

So, yeah, there’s revenue, and more than the previous quarter. But with a $300m valuation?

BrainChip also has Akida, a neuromorphic chip architecture that works on event-based processing. Okay, cool, that’s legit. They have a dev board, the kind of thing engineers use when they want to tinker, test models, or evaluate how a system handles edge workloads. They have published patents. They have an engineering team with real semiconductor experience. They have some partners, including universities and industry players who have at least expressed interest in the space. They also show up to the right conferences. CES, Embedded World, RISC-V gatherings. They’re not hiding. They’re out there trying to be part of the conversation.

That’s the good news.

Now let’s talk about the holes.

You know what’s missing from the glossy investor deck? Anything that looks like actual commercial proof. There is no breakdown of licenses signed, deployments made, or customers paying real money. There is no timeline for production volume, no evidence that an automaker, a defense contractor, a consumer electronics company, or a robotics manufacturer has chosen Akida for a shipping product.

Instead, you get TAM charts. Trillions in theoretical AI demand. Graphs that slope up and to the right like a ski jump on Viagra. Market research projections that conveniently stretch to 2034. And the constant carrot that edge AI will boom, that LLMs will move on device, that agentic AI will transform mobility.

All of this is possible. But so is Charlize.

The deck also plays the classic logo salad trick. Mercedes-Benz. NASA. ESA. ARM. Nvidia. University of X. Defense lab Y. Are these customers? Partners? People who once emailed an engineer? Nobody knows. The deck doesn’t say. A real company with real commercial traction would state precisely what these logos represent. Instead, they are simply placed on a slide, inviting you to infer whatever story you want to believe.

I mean, you can dig through their news releases if you want to extract that info, but when I did, I didn’t find a lot of dollars flowing in.

I did find a LOT of POTENTIAL MARKET talk.

For example,. this is the SECOND PAGE of their deck.

Really bro? You’re coming in hot with ‘we talked to some people at conferences?”

Maybe page 3 gets to the raw meat for the masses?

Nope. You could summarize this slide as “the potential market, how we think we can penetrate it, and what we’ll do down the road if all goes well.”

FLUFF.

Hey, surely page 4 of the investor deck shows something more than the potential market… right?

Jesus Christ on a hot dog stand, guys.

How about page 5?

I won’t bore yuou with it, but there’s more of this on page 6, and page 7.

Page 8 gets to the executive bios, then the logo salad, and that’s it for the deck.

To be clear, if this were a pitch deck for a straight startup with a business plan and a theory, fair enough, project away and get fluffy.

But it’s not. These guys were at a stock high in 2022.

The biggest problem is not that BrainChip is lying. The problem is scale.

They have a $300 million dollar market cap on the ASX, which prices them as a company far closer to commercial maturity than they are.

Three hundred million dollars suggests revenue around the corner. It suggests design wins, production deals, or licensing partners with open wallets. It suggests moat, traction, and competitive advantage. It implies the business is past the scientific validation stage and nearing the revenue expansion stage. But BrainChip is not anywhere near that point.

That’s why the stock is low.

It’s also why nobody is banging down their door to oversubscribe the cheap financing, or take them out at this ‘cheap’ price.

Ask yourself why, with the AI sector BOOMING the last few years, why Brainchip stock has only gone down. Do you really think the rubber is going to hit the road NOW?

Neuromorphic computing is still an emerging field.

Ask Mythic, which burned through tens of millions before collapsing. Ask Intel, whose Loihi chip sits mostly in the research world. Ask IBM, whose TrueNorth chip was cutting edge and went nowhere. The idea has been floating around the semiconductor world for 20 years and no company has cracked the mass market.

That’s not a BrainChip problem. That’s an industry problem. But the valuation implies BrainChip will be the one to break through where everyone else stalled.

If they had tens of millions of revenue, sure. If they had dozens of licensing deals, sure. If they had a major OEM ready to ramp production using Akida, or $150m in cash ready to be spent on execution, absolutely. But there is no evidence of any of that. The company keeps talking about potential, not delivery. It keeps highlighting the market, not its own numbers. It keeps pointing to ecosystem partnerships rather than proprietary wins.

This is classic pre-revenue deep tech behavior, but the cap is post-revenue fantasy. That’s the disconnect. And disconnects like this are where retail holders get burned.

Why are they raising $2m at the 52-week low? Because they burn through $3,8m per quarter, and have three quarters of runway left.

Note: They will not be profitable by the time that runway ends. They’re grabbing $2m now because they NEED it, and the AI space is wobbling right now.

If the AI sector continues to crumble, they might not find another $2m, cheap or not.

Let’s be fair though. The tech might be good. The team might win a design-in. The market could shift to edge AI faster than expected. They could become the go-to IP vendor in neuromorphic compute the way ARM did for low power processors.

That’s the bull case. It’s just not the current case.

The current case is this: BrainChip is a promise story valued as an achievement story. It’s a technology in search of adoption. It’s a niche product in a massive narrative. It’s a company that wants to be judged on potential but is being priced as though the potential has already been fulfilled.

And when you’re overvalued WHILE AT THE 52-WEEK STOCK LOW, that’s a bad scene.

Investors need to stop asking whether neuromorphic computing is a big deal and start asking whether BrainChip is the one that will monetize it. Big TAM does not equal big revenue. Logo slides do not equal customers. PowerPoints do not equal product shipment. Engineers do not equal execution.

Sell, mum.

— Chris Parry

FULL DISCLOSURE: No commercial dog in the fight.