Hacker News Reader

Comment by fencepost | Hacker News Reader

Doing a quick bit of searching based on the 4680 makes me think that there has been or will be a change from NMC811 to LFP chemistry in the 4680, including one article talking about changing to US and European-based in-house manufacturing and reducing dependence on China.

I'm no fan of Tesla, but this looks like the collapse of the contract with the supplier for the battery chemistry they've moved away from, aka "no [more] big deal."

2023 article confirming NMC chemistry: https://iopscience.iop.org/article/10.1149/1945-7111/ad14d0

5/2025 article discussing change to LFP: https://roboticsbiz.com/teslas-4680-lfp-battery-explained-ch...

3/2025 article comparing BYD's LFP and Tesla's NMC/NCM: https://www.sciencedirect.com/science/article/pii/S266638642...

View on HN

This makes sense but does it make sense to manufacture LFPs as 4680 cylinders instead of rectangular blocks/blades, given https://en.wikipedia.org/wiki/BYD_Blade_battery ?

Which cell technology to use depends on the application. Tesla actually uses BYD blade batteries in some of their vehicles sold in Europe. The main issue with prismatic cells is that to be safe, they must be made with LFP chemistry, which hurts energy density. LFP is also worse at charging/discharging when cold, though battery management software mostly solves that.

Cylindrical cells make sense for higher performance NMC and NCA chemistries, as they can be cooled more easily (coolant lines can run in the voids between cylindrical cells), and any single cell failure is less likely to cascade to other cells. Batteries with cylindrical cells were easier to repair, but nowadays cells are welded together instead of bolted, so that's no longer an advantage.

> The main issue with prismatic cells is that to be safe, they must be made with LFP chemistry, which hurts energy density

This is obviously untrue. Tons of other chemistries have used prismatic cells with good safety as well. You think Macbooks and iPhones use LFP or cylinder cells?

> Batteries with cylindrical cells were easier to repair,

It can be just as easy to repair a prismatic battery as a cylinder battery. It all comes down to the layout of the battery. And as you mentioned, how the battery is constructed, if the battery is structural, etc.

Since this is a discussion about electric vehicles, I thought it could go without saying that I was talking about batteries in such vehicles, not batteries in consumer electronics that are 1,000 times smaller.

To use an analogy: If someone stores a gallon of gasoline in a single-walled plastic container, that's probably OK. But storing 1,000 gallons of gasoline without certain safety measures is unsafe. So it goes with battery capacities.

But it still is untrue even in the discussion of electric vehicles. Tons of EVs have been made safely with chemistries other than LFP with prismatic cells. In fact most non-LFP EV batteries are pouch or prismatic, not cylinder.

This isn’t true: “most” non-lfp batteries in EV context are Tesla. Which means they are cylindrical.

Or are you counting OEMs rather than actual vehicles?

Apart from Tesla very few EVs ever used cylindrical cells.

But that's usually just the packaging... if you open that up you'll find... cylindrical cells

No?

BMW is moving the other way, from prismatic cells to cylindrical

https://phdenergy.com/beyond-teslas-4680-why-bmws-gen6-4695-...

My point was about cylindrical cells with higher energy chemistries like NMC and NCA. Rivian uses cylindrical cells for their non-LFP batteries. Lucid uses 2170 cells. As far as I can tell, those three (Tesla, Rivian, Lucid) are the only US car manufacturers who have not had battery recalls due to fire risk.

GM, Hyundai, and Nissan all used pouch cells with higher energy density chemistries, and had recalls due to battery fire risks. Ford also recalled tens of thousands of their plug in hybrids due to battery fire risks, though they haven't found a solution yet beyond limiting the max charge of the battery. These batteries are also NMC pouch cells.

I'm sure it's physically possible to make safe, reliable pouch or prismatic cells using higher energy chemistries, but so far it has been risky for those who have tried.

> You think Macbooks and iPhones use LFP or cylinder cells?

https://old.reddit.com/r/spicypillows/

And I can find tons of videos of cylindrical cells catching fire and exploding. What's your point exactly?

Yep. Prismatic cells have poorer packaging-to-material ratio (circles are optimal). They offer better thermal properties, but thermals are not the main limiting factor anymore.

And the US automakers tried prismatic cells before. Chevy Volt used them in 2012!

Chevy sells EVs with prismatic and pouch cells. I don't recall any they've widely sold that used cylinder cells. Most automakers use prismatic cells on their cars, even non-LFP variants.

I’m not sure id call Chevy widely sold, yet. Their full year sales don’t even match a single quarter of Tesla.

I can go to any major metro area in the US and go buy a GM EV today. That's widely sold.

That’s “widely available” — no one is buying them, so not “widely sold”

Also GM had to replace batteries in 142,000 Chevy Bolts & Volts due to fire risk, so I'm not sure that should count as an example of a successful use of non-cylindrical, non-LFP batteries.

To be fair everyone had to recall those LG batteries and unauthorized SK clones [*]. Porsche Taycan, Ford F-150, its recalls all the way down.

https://www.wardsauto.com/news/ford-terminates-ev-battery-su...

https://www.autoevolution.com/news/ford-recalls-f-150-lightn...

SK settled for $2B case that revealed they stole LG battery technology https://cleantechnica.com/2021/02/15/itc-sides-with-lg-chem-...

How are circles optimal???

Circle packing = 90%

Blade battery packing = 100%

it's not just a basic geometry problem, it's an engineering problem. You need to account for far more details.

With cylindrical cells, you have more leeway in the overall shape of the battery pack and can fill nooks and crannies with them.

It's more complicated. You can't pack the battery at 100% anyway, because you need cooling. Cylindrical cells are also more rigid, so they need less supporting material, or they can even be a part of the support structure itself.

Yeah, it's typical of Elektrec to interpret anything Tesla-related in the worst light.

I personally fact checked articles about "Autopilot disengages milliseconds before collision" and the one related to Benavides v. Tesla case.

In the first case they jumped to conclusions. In the second case they "forgot" to mention any details that contradict their narrative: that there were two cases separated by years, that the police has received all the information they needed in the first case, that the driver was pressing the accelerator.

My understanding is Fred (Electrek) has been nothing but negative about Tesla since he got upset that they haven't released the roadster yet. Something to do with having enough referral points to get one. It has become his whole identity and it is sad to see.

Pretty sure those comments are tongue in cheek. And it’s not that he’s dying to buy one, Tesla owes him one, and doesn’t have to fulfill it by never making any

HN audience loves to undermine one of the only places actually objectively calling out the mega corps

I wouldn't speculate about his reasons, but many of his articles are objectively bad journalism.

I havent seen an article on that site be remotely positive to tesla in a long time if ever. Its weird, the obsession with Tesla hating.

he also sold all is tesla stock, so I assume there is some incentive too feel better in case tesla fails.

Does having larger LFP cells, like the 4680 format, allow this chemistry to be used in higher-performance models? Right now it looks like only the RWD basic models use LFP cells everything else uses NMC.

This is more recent and informative.

https://youtu.be/-sN2_Lp10lw?si=blBPFQ9jy2L8rCNp

Reply on news.ycombinator.com