Donut Lab alleged Solid State Battery launched at CES 2026

[ejeffrey]

So the thing that behaved like an off the shelf Li-Ion battery run beyond safe limits was in fact an off the shelf lithium ion battery run beyond safe limits?

/Shocked pikachu

[Siwastaja]

So the thing that behaved like an off the shelf Li-Ion battery run beyond safe limits was in fact an off the shelf lithium ion battery run beyond safe limits?

Exactly what sensible (and great and nice and modest) people like me predicted early on:

But the charge curve, the amount of internal resistance, all mean their sample could be just a state-of-the-art, COTS NCA/NMC high-power cell, overdriven to 4.3V and 11C rating, because in such test, they don't have to care about aging (because it's tested only for a few cycles, not 100000). Heck, 10 year ago I tested COTS LCO and NCA cells charged to 4.45V instead of rated 4.20V, seeing +20% capacity, and no damage / no capacity degradation after 10 cycles, after which I stopped. I could have had VTT to rerun the same test and claim that I have a new battery tech with 20% better capacity, and higher 4.45V voltage, definitely not based on lithium and cobalt, and capable of 100000 cycles. This kind of stunt is easy enough to arrange.

[Randy222]

So the thing that behaved like an off the shelf Li-Ion battery run beyond safe limits was in fact an off the shelf lithium ion battery run beyond safe limits?

/Shocked pikachu

Yeah, but theirs had 3 more protons in it. Special.

[ftg]

The latest available video in the donut circus is actually pretty interesting.


If they truly can tailor the voltage by stacking layers in the pouch cell, it will be interesting to see how well they stay in balance.
Ofc the voltages shown (3.30V, 4.97V and 7.8V) could be one normal lipo, two discharged lipos in the same pouch and two lipos with a decent charge in one pouch.
It's not like the nickel strips would show that badly from the middle of it.

[voltsandjolts]

Is there a drilled hole under each cell, with wires taking the different voltages from psu's under the bench to the "battery" tabs.
At this stage I don't know what to believe.

[tszaboo]

The latest available video in the donut circus is actually pretty interesting.
If they truly can tailor the voltage by stacking layers in the pouch cell, it will be interesting to see how well they stay in balance.
Ofc the voltages shown (3.30V, 4.97V and 7.8V) could be one normal lipo, two discharged lipos in the same pouch and two lipos with a decent charge in one pouch.
It's not like the nickel strips would show that badly from the middle of it.

Everything, except giving it to a third party to make a full test report.
No, instead we show some voltage levels, that may be a battery or a capacitor or a bunch of coin cells in series. 
Measured with a Chinese 100 EUR multimeter. Without load.

[mikeselectricstuff]

Does anyone have any interest in anything these Donut clowns have to say any more?
Their scam is over, time to move on.

[Siwastaja]

Does anyone have any interest in anything these Donut clowns have to say any more?
Their scam is over, time to move on.

I have been dissatisfied with the lack of shitstorm. Scam is indeed over but usually the best part is what follows after that  - and that's been lackluster, so far.

Of course when lawyers and cops get involved, it can delay the shitstorm by 2-3 years. And at that point, most people have lost their interest, meh.

[Randy222]

If any of us didn't give the nutLabs any money, then all the complaints/criticism here is just for fun.
If anyone did invest, then I guess we are all laughing at you, not with you.
.... and for whatever the nutLab folks are trying to do, wouldn't a decent contract in some AI datacenter already provide them the technical solutions for a product they seek to build?

It would just be better if the dude in the video was dressed like this clown, would be so much more interesting/appropriate.

[EEVblog]

Does anyone have any interest in anything these Donut clowns have to say any more?
Their scam is over, time to move on.

I haven't been keeping up, does anyone have a TLDR on what's happened recently?
Are they still a thing?

[ejeffrey]

I haven't been keeping up, does anyone have a TLDR on what's happened recently?
Are they still a thing?

They have released capacity testing data, but its a different cell design than the one they did charge/discharge rate on. They only discharged it at 0.1C and to a very deep (for lithium ion) depth of discharge to hit the capacity claim -- something like 2.2V.  They still have no life cycle data and nothing is production ready. There were a bunch of fraud allegations, but any investigation is going to be slow, and they are out of the spotlight now. 

Basically its going out with a whimper.

[ftg]

Similarly, now the energy density numbers have been published. Also a destructive test with a nail penetrating the cell (nothing happens) and a disassembly of a cell showing the internal construction.
But these are done with the "version 1.5" cell, not the "version 1" that was used before.

This project included independent energy density measurements of the energy storage devices
supplied by the customer. Cell weight and dimensions were measured, and cell capacity was
determined at an ambient temperature of +25 °C within the recommended voltage range using a
0.1C discharge rate.
Based on the measured weight, dimensions, and capacity, the gravimetric energy density was
determined to be 409.3 Wh/kg, and the volumetric energy density was determined to be
804.7 Wh/L

But they did finally find a scale.
The destructive testing was done by Intertek, not VTT.

[thm_w]

Link: https://pub-01c7ea9972fc472b971b53522518cfc0.r2.dev/VTT-CR-00391-26_Energy_Density_Test.pdf

The SiCore SA03? pack is quite close to that spec here, advertised as 400Wh/kg and 872Wh/L: https://www.unmannedsystemstechnology.com/company/amprius-technologies/
That or they just got a sample preproduction pack from another manufacturer.

[ftg]

And the latest episode in this circus is that the new "1.5" battery cells contain only trace amounts of lithium and sodium.
Tested at Intertek instead of VTT.

Report:
https://pub-3f15a9a74c904192ade59d264aec1584.r2.dev/IWTN-W000020339RLM002.pdf

[tszaboo]

And the latest episode in this circus is that the new "1.5" battery cells contain only trace amounts of lithium and sodium.
Tested at Intertek instead of VTT.

Report:
https://pub-3f15a9a74c904192ade59d264aec1584.r2.dev/IWTN-W000020339RLM002.pdf

I'm not that familiar with this whole spectroscopy method. Would lithium compounds show up as lithium content, or not?

[TimFox]

A discussion of the spectroscopy technique from Perkin Elmer, the manufacturer of the equipment referenced in the report:
https://www.perkinelmer.com/library/icp-oes-tips-and-tricks.html
It is essentially optical spectroscopy of the light from a plasma containing the test object.
Optical spectroscopy usually deals with the atomic, not molecular, content of the object.

[thm_w]

https://electrek.co/2026/09/23/donut-lab-intertek-lithium-test-solid-state-battery/
The test is telling you the percent elemental lithium by weight. Though not the form it was originally in.

Obviously very low values and not typical of a lithium-ion battery, but as said in the article the samples could have come from anything.

1 ppm w/w = 1mg per 1kg

[Electrodynamic]

Beyond the typical nerd speak, there is one thing we do know and there's more than one way to skin a cat.

It begs the question, if I had similar tech would I disclose all the IP?. Well no that would be stupid and everyone would be trying to reproduce it for their own self interests. It would seem to me the recipe for a successful company would be to keep everyone guessing. I mean if joe blow on the corner could figure out what you're doing and replicate it you wouldn't be in business long. Here in Canada we call this "capitalism".

I think it's cool that so many people have a melt down just because they cannot figure it out. Newsflash, as we get older we tend to understand that little makes any sense. Let's be honest, most are just pissy because they have no idea how this could work. The thing to remember is that there are about 8 billion people on the planet and if you think your the smartest there is a 99% chance your deluded.

For example, a 6 year old just set the world record(s) for being able to solve a Rubicks cube with an average of 4.27 seconds.
https://www.msn.com/en-in/news/other/meet-lian-yunzhi-the-6-year-old-who-broke-the-rubik-s-cube-world-record-twice/ar-AA2cRL3S?ocid=BingNewsSerp
Think about that, by the time most people here could pick up the cube and look at all the faces this 6 year old would have already solved the problem. Sorry, your not the sharpest tool in the shed and not by a little but more than you can even imagine.

Newsflash, it's going to get much worse. Moreso than you can even imagine.

[SteveThackery]

Newsflash, it's going to get much worse. Moreso than you can even imagine.

Oooh, mysterious!😳 But presumably not more than you can imagine, because you're smarter and know secret stuff that we don't know. Yes?

[Bud]

Think about that, by the time most people here could pick up the cube and look at all the faces this 6 year old would have already solved the problem.

It is unclear from the article how much time the girl spends looking at the faces before starting. Secondly, it is unknown how the cube is prepared (the initial state). Third, spending 3 hours daily can simply make you memorize solutions for specific initial states. Forth, it is 3x3 cube.

[Cyclotron]

Think about that, by the time most people here could pick up the cube and look at all the faces this 6 year old would have already solved the problem.

It is unclear from the article how much time the girl spends looking at the faces before starting. Secondly, it is unknown how the cube is prepared (the initial state). Third, spending 3 hours daily can simply make you memorize solutions for specific initial states. Forth, it is 3x3 cube.

A kid that lived across the street from me could toss a ball in the air and hit it with a bat every single time at 2 years old.

Dang I'm old, that kid just graduated high school.  And I want a donut. 

[mikeselectricstuff]

Think about that, by the time most people here could pick up the cube and look at all the faces this 6 year old would have already solved the problem.

It is unclear from the article how much time the girl spends looking at the faces before starting. Secondly, it is unknown how the cube is prepared (the initial state). Third, spending 3 hours daily can simply make you memorize solutions for specific initial states. Forth, it is 3x3 cube.

Normal cube competition rules are you can spend as long as you like examining the cube,  put it down and the time is between picking up & putting down after solving

[booscrawl]

Quantumscape released a video of a nail going through their prototype cell in 2024 and nobody made a peep.

https://youtu.be/z84EnVzTHIA?t=154

I own a little bit of their stock. Less than you would need to buy a car with their batteries in it...

[Siwastaja]

Quantumscape released a video of a nail going through their prototype cell in 2024 and nobody made a peep.

Because everyone else also has released similar nail penetration test results. It's one of the most standard, most repeated tests, so there is absolutely no wow factor in being able to do the same everyone else.

[scopeguy27]

The claims sheet reads like a checklist of red flags: order-of-magnitude jumps in energy density, no independent cell teardown, and a launch timed for CES headlines instead of a datasheet. Battery chemistry is a brutal, incremental business - when someone skips the boring middle part (pilot lines, cycle-life data, safety certification) and goes straight to hype videos, what changed is the marketing, not the physics. I'll believe it when a third-party lab tears a cell down and repeats the cycle testing.