Hacker News new | past | comments | ask | show | jobs | submit login
Led by Tesla, EVs drive chip industry's shift beyond silicon (nikkei.com)
182 points by Corrado on Sept 6, 2021 | hide | past | favorite | 130 comments



Good article, but very Japanese centered. This of course is understandable, as it is coming from Nikkei.

There has been a lot of news about how the US is giving up leadership in semiconductors to Asia, but the US has actually been holding our own in Silicon Carbide. A US company called Cree, through their division Wolfspeed is the leader in manufacturing silicon carbide wafers. AFAIK, they are still making the most wafers in the world.

There is another little known US company, ON semiconductor, that is arguably the leader in auto grade silicon carbide chips. They were not the first to make the sic mosfet, but they were the first to make sic mosfets that are qualified for use in cars. On Semi had acquired the legendary Fairchild semiconductor which had a lot of expertise in high voltage high power applications for auto and industrial. This probably helped commercialize silicon carbide.

This is arguable of course, but ON semiconductor insist that they make the highest performance automobile grade SiC chips on the market.

To see how fast this technology is moving, you can watch this video [1] where sandy munro needs a helper to wrestle with the ford mach-e inverter on a table. And then you can see this video [2] where the CEO of ON semi, holds two inverters for a car and an suv in each hand effortlessly.

[1] - https://www.youtube.com/watch?v=mHVV52lPyIs&t=1086s, see around minute 15.

[2] https://www.youtube.com/watch?v=jfq0kDzyPZY - see about minute 46.


I am far from a subject matter expert on this, but I have been doing my best to read up on this lately. It seems like we’re globally diverging on fabrication expertise at a process level i.e. nanometer scale [0]. The bulk of leading edge chip manufacturing (currently around 7nm or below) seems to indeed be slipping away to Southeast Asia, but as you said the U.S. is holding their own at larger process levels (Cree is mentioned in the article as well) and still producing at a volume that may be surprising to many doomsayers. It’s worth noting at this point that, according to the article I linked below, it is estimated that five larger process chips are required for every leading process chip produced. So this doesn’t seem to entirely be an accident on the part of U.S. companies. Sure, some companies like Intel have simply failed to innovate in the past few years, but others seem to have made a calculated decision to avoid massive R&D expenditures that may or may not pay off as we begin to approach the limits of Moore’s law. The reality is that leading edge chip fabs are extraordinarily expensive. I read a podcast transcript recently with a prominent professor who is an expert in this field [1]. He mentioned that he did some consulting for the U.S. Govt., and when they asked him for his estimate for how much it would cost in subsidies to catch up with or outpace the rest of the world in fabs his response made the guy who asked the question practically fall out of his chair. Have we reached a point where quantity (at an average level of quality) is now more important than having the fastest and the smallest IP?

[0]: https://semiengineering.com/can-the-u-s-regain-its-edge-in-c... [1]: https://www.theverge.com/2021/8/31/22648372/willy-shih-chip-...


> seems to indeed be slipping away to Southeast Asia

Do you mean east Asia? Southeast Asia still doesn’t have much fab capacity, and most of what it has isn’t very sophisticated, unless I’m missing something?


Yes, sorry. East Asia is what I meant. To be completely honest, I’m not sure if I read SE Asia somewhere or if I’ve always just thought of Taiwan as SE Asia for some reason. Mercator projections, maybe? Anyway, today I learned Taiwan is in East Asia, north of Southeast Asia.


Taiwan is barely on the edge between Japan (East Asia) and the Philippines (SE Asia). But culturally is very East Asian.


I think this is just some people's reflexive notion that 'China + sometimes Japan' = East Asia & 'Rest of Asia' = southeast Asia.

It was mostly started by China companies like Alibaba treating any international expansion as 'Southeast Asia,' since they did it via HK or SG offices.


The leading fabs are in Taiwan and South Korea.


Yeah, South Korea is roughly at SF's latitude and Taipei is roughly Miami.

Maybe Taiwan can be bucketed into SEA, but SK is definitely East Asia


EUV LLC was heavily funded by DARPA and US industry in the 90s and is why we are able to have export restrictions on (European) ASML machines used by TSMC for the latest EUV nodes.


>There is another little known US company, ON semiconductor,

I chuckled. Only on a Software developer focused forum would ON being called a little known US company. They are really well respected in semiconductor industry. If I remember correctly, ON along with Freescale came from Motorola.


Yup, to me they're known just a bit less than TI and Analog Devices (now that Atmel and ST Micro are gone) and more than Microchip or ST because they serve so, so many high power applications.

I just think those high power applications are invisible, like tens of percent reductions in the power used in an application. You can see them if you measure closely but not if you're casually working with it.


Just introduce fun event: Nikkei Japan called Nvidia "A mysterious AI semiconductor manufacturer" in 2017.


> To see how fast this technology is moving, you can watch this video [1] where sandy munro needs a helper to wrestle with the ford mach-e inverter on a table.

My take away from watching that clip wasn't him being confused with the function of the inverter but rather being puzzled by how every step of the assembly eschews the principles of Henry Ford and even the rules he said he wrote while at Ford.

Maybe you linked to the wrong time?


What I meant is he was wrestling with the inverter physically, not intellectually. I.e. the inverter (after they assembled the whole thing) was very large and heavy compared to the one presented by the on ceo.

I guess I should have referred to the end of the video.


Ah... I completely misunderstood what you were saying.

Is Tesla using silicon carbide chips in their inverters? While not as small as the ones the CEO is showing off, they're significantly smaller than what Ford and Nissan are using.


Correct me if I'm wrong but I don't believe Cree manufactures any chips in the US?


I don't think cree manufactures any chips period. They make wafers, which other companies use to make chips. They make their wafers in the US, in north carolina. They are building another large factory in upstate new york.


Why is SiC a good thing?

1. It has a higher bandgap energy than silicon, which means it can tolerate more heat without breaking down. It also means the junctions can be physically smaller.

2. Smaller junctions mean the transistors can switch faster, which means less time spent in the linear region, which means they generate less waste heat than silicon transistors.

3. SiC also has lower ON resistance, which means less waste heat still.

4. SiC conducts heat better than silicon, so removing its [lesser] waste heat is easier and requires a smaller heat sink apparatus.

All of this makes SiC a fantastic technology for power electronics. (It's less important for signal electronics, but even signal electronics waste small amounts of power so reducing that further is still relevant.)

For EVs, SiC inverters mean less mass for an EV to carry around. Less mass per se doesn't matter as much as you might think for an EV because with regenerative braking you get back some of your acceleration energy when you decelerate. However, air resistance matters enormously in an EV because you can't get back the energy you lost pushing the car through the air. And "less mass" also means "smaller" and "smaller" means "more opportunities to make the car aerodynamically efficient."


Thank you for answering everything I ever wondered about SiC. Are there any other semiconductors that would be even better than SiC in all four points?


Diamond is even better in terms of bandgap and thermal conductivity (one way jewelers verify a diamond is real is by checking to see if the gem has ridiculously high thermal conductivity). But diamond semiconductors are much too expensive ATM.

Gallium Nitride is the closest competitor to SiC now; it's better than SiC for high frequency applications but worse for high temperature applications and GaN is somewhat harder to manufacture. But that's changing rapidly.

https://www.arrow.com/en/research-and-events/articles/silico...


> But diamond semiconductors are much too expensive ATM.

It is also difficult to dope diamond n-type, as most of the donors are not shallow.

Interestingly: The conduction band in bulk diamond is above the vacuum level, meaning that conduction electrons "want" to fall out of the material. (Surface physics gets in the way.)


I'd like to point out this is not about increased range, for example, if the ECU already works at 95% efficiency, that's not a very high number for brushless ECU, just for example, so 5% of energy is wasted, if we can reduce that in half, the efficiency will only increase to 97.5%, that's a very small improvement on range. BUT, now the ECU only needs to dissipate half wasted heat, this could be a big net win.

A more close to everyday life example would be chargers for notebooks, we now have smaller chargers thanks to GaN FETs, SiC is like GaN.


This isn't quite right...

In a car which can accelerate up to speed and brake again every 30 seconds while going through a city with a lot of stop lines, the efficiency of the power components really matters.

If you loose 5% in the motors, 5% in the power electronics and 5% in the battery in each direction, then each accelerate-brake cycle is losing you ~30% of the energy you started with. At town speeds, friction and air resistance losses are tiny, so that 30% is probably actually most of the energy you are using!

Add that to the fact that certain motor control techniques such as field weakening involve every Amp of real power that comes from the battery passing many times though the inverter, since there is a higher circulating current between inverter and motor coils.


You're going to lose a lot more than 5% in the motor. You will also lose more to rolling resistance, various hystereses, etc...

Even assuming 100% inverter efficiency you will lose at least 20% in the regen part of the cycle alone.


Motors can get well above 95% efficiency, even at a wide range of speeds and torques.

Eg. See https://cdn.motor1.com/images/mgl/y83JG/s3/tesla-motor-effic...


They can when they are being used to produce kinetic energy. Using a motor as a generator however has much lower efficiency.


Absolute nonsense. Motors are do not care whether they are generating or consuming electrical power. It is a fundamental fact of induction that rotation in the motor creates a reverse voltage- motors literally would not work if they did not have very similar efficiency as generators.

The only differences between a motor and a generator in practice are cost optimizations and fixed-frequency optimizations. Industrial generators run at fixed speeds, unlike EV motors, so they can be built more cheaply. EV motors are not any less efficient when used to recapture power.


That is true in theory. In practice, when using the motor as a generator, the operating conditions are different, and so are the efficiencies. The load side is also different, the motor controller will not act the same way as the inertia of the car.

For example, even the simple inductance and resistance of the motor means that at the same speed and at the same load angle you will see a lower voltage at the poles of the motor when acting as a generator. When your motor has high efficiency, these differences can lead to a large absolute difference when running as a generator.


You could say the same thing about how speakers/microphones are equivalent. But it should be clear that some of those devices are tuned for outputting sound vs inputting sound, or vice versa.


> If you loose 5% in the motors, 5% in the power electronics and 5% in the battery

You realize SiC is not the savior for all your problems right?

I'm not saying efficiency doesn't matter, I'm saying this is not gonna improve range significantly, there is a difference.


The GP's point is that this is not about maximum range, it's about increasing range in city transit that is normally much smaller.

Both gains seem relevant to me.


EVs typically have more range in city transit than on the highway.


Totally! X^n with 0<X<1 quickly goes to zero as n increases. You get about 6 of those cycles if X=.9 before you lose 50% of what you started with initially.


You mean ESC, not ECU.

Generally cooling ESCs is not that big of an issue, just stick them in the same cooling system as the batteries.


arguably the same thing. If you want to get pedantic, "ESC" refers to a specific subset of drive topologies that are all fairly lossy on the D axis, meaning the efficiency drops at low speeds or under heavy acceleration.

"Drive controller" or "motor controller" are common terms, and while I have seen acronyms like DCU, nothing really sticks. MCU is already taken as an acronym, which doesn't help.


ESC doesn't have any connection to efficiency as a function of load. All circuits will be less efficient as you output more amps, ie, at low speeds where the voltage is low but the power stays the same, or under heavy acceleration.

Nowadays ESC simply refers to any circuit you attach to motor and provide with a target speed/torque/power. In common practice yeah ESC/drive controller/motor controller are used interchangeably.

ECU is specifically something else. It's "Engine Control Unit". It refers to the controller for a combustion engine, not a motor.


To make things more confusing, ECU can also stands for Electronic Control Unit


I'm pretty sure Tesla refer to them as ECU.

Trust me, I'm more customed to use the term ESC since I play RC models and I don't drive a electric car, or any car.


This is the key observation - hitting breakpoints in efficiency that let us remove active cooling, silicon area or heatsinks reduce weight, cost & complexity.


Yep. High efficiency is also the key to the crazy high performance we see in high end EVs these days. Going from 96% to 98% drivetrain efficiency only gives you 2% more battery life but it doubles your power output.


You are absolutely right there. I first read that in an article an EE wrote. He said what you care about with power inverters isn't the efficiency but the inefficiency. Small gains in efficiency you wouldn't care about translate in to large reductions in heat and smaller cheaper power devices. Power devices don't follow Moore's Law and aren't cheap.


The article doesn't pass the smell test for me.

Claim 1. Smaller SiC inverters enabled Tesla to implement sports-car-like streamlined design.

Traditional Si inverters are not that big. Certainly not big enough to single-handedly change the car design [1]

Claim 2. Hyundai will use SiC chips to increase its typical EV range by 5%.

The article claims that SiC will reduce wasted energy by half (or more). If we go by this figure, Hyundai's EVs need to waste 10% of its energy on Si chips. This seems quite high to me. No doubt there are losses in the motor, power electronics, and battery, but losing 10% of the EV's total energy on Si based chips seems incredibly inefficient to start with.

[1] https://www.youtube.com/watch?v=mHVV52lPyIs&t=909s


> If we go by this figure, Hyundai's EVs need to waste 10% of its energy on Si chips.

Most of it is not lost in the silicon, but in the interconnects (wiring and connectors) and motor(s).

SiC is better (more robust) than Si with high voltage, high current, and/or high temperature. Specifically, kilovolt-level, kiloamp-level, and temperatures above 150 celsius.

If you double your battery voltage from 400V (current standard) to 800V, you halve your resistive losses in the "wires" for the same electrical power demanded. Silicon MOSFETS that can withstand a kilovolt and a kiloamp are very big and very expensive.

You get more power delivered for the same power leaving the battery.

Size: SiC's very high thermal conductivity means less of a margin of safety for secondary breakdown. SiC tolerates higher current density (amps/square mm) than Si, so smaller devices are possible, saving mass.

Its temperature stability allows it to be sintered to the package heatsink--further improving heat removal. Tolerance for high temperatures means you need less mass and volume for heat dissipation structures.

Other efficiencies: With very high currents (kiloamp) and medium frequency switching, SiC IGBTs provide more benefits, because rather than being directly proportional to current as with MOSFETs, losses are proportional to the log of current.

Alternatively, doubling the battery voltage again to 1.5kV (and so halving the currents) might allow Hyundai to use more aluminium for the "wires", with a cost saving over copper.


>Alternatively, doubling the battery voltage again to 1.5kV (and so halving the currents) might allow Hyundai to use more aluminium for the "wires", with a cost saving over copper.

This seems dubious, you're not going to be able to use aluminum for motor windings which is where the bulk of the copper in an EV is going to be. As for conductors between the battery, HV auxiliary devices, inverter, motor, charger, etc they could already use aluminum, it's not like they're space constrained on the gauge of cable between components. Aluminum is already more mass efficient than copper, you might need a thicker gauge to match the same resistance but that thicker aluminum is still lighter than copper and way cheaper. The problems with aluminum are oxidation and brittleness. Maybe eventually auto manufacturers will shift more towards aluminum bus bars between components with no relative movement but I don't see a big link here (haha, busbar puns) between ampacity and the choice between copper and aluminum.


Resistive losses are proportional to current squared, not linear.


Active work has been going on in this area for years, for really high power applications.[1] Locomotives. Power substations. Wind turbines. At the low end, silicon carbide gates have been used for wall wart power supplies, in low volume. Looks like it's finally making it into automotive, which makes sense.

[1] https://www.nist.gov/system/files/documents/pml/high_megawat...


>Applications for HV SiC

>Data center distribution - MV AC or DC to the rack, (13.8kV distribution to match gensets

What the hell are they smoking? MV to the row, maybe, but are they seriously proposing MV terminating to a power supply in every rack??? That seems like it'd be an installation and maintenance nightmare. I feel like it'd be much more sensible to at least limit in-rack power to 480V max. Stick with something that doesn't require DC techs to use a hot stick to plug in a new rack.


No point gaining a few points in efficiency if the cars themselves keep getting bigger and bulkier.


For electric cars though, you make more money as a carmaker if their batteries are smaller, which happens when your consumption is lower, so it encourages slicker designs, unlike ICE cars where somehow Americans still think 20mpg is acceptable in 2021, and the carmaker makes as much money or more selling you a 20 or a 50mpg car.


Unfortunately, and also "Led by Tesla" - the EV metrics that people have been trained to look for are now: range, 0-60 times, and such.

Things like weight, battery capacity, or chemistry - all which have the largest impact on emissions - are ignored. In fact, judging by the chemistries that Tesla is now exporting from China, there will be a very serious effort to hide battery chemistry as much as possible.


The biggest emission reduction is when a person switches from using an ICE car to a EV car. Getting the price of a new EV car to be below a similar ICE car will cause that to happen very quickly. These new battery chemistries using iron instead of nickel are a key to reducing price and will lead to much quicker overall emission reductions. Also price is a close proxy for total energy. A ton of iron (~$200 a ton) surely takes a lot less energy to produce than a ton of nickel (~$20,000 a ton). I have no idea how much less but I would guess somewhere between 10 and 50 times less.


> The biggest emission reduction is when a person switches from using an ICE car to a EV car.

Not at all. The biggest emission reduction is when a person doesn't buy a new car at all. The greenest car is always the car that won't be made, regardless of propulsion system.

A Tesla Model 3 is actually a more polluting vehicle than a Toyota Corolla up until the breakeven point of 21,000 KM (and change). [1] And that's assuming a US energy grid.

If that same Tesla Model 3 is being driven in say, China, the Toyota Corolla is greener than the Model 3 until the 125,000 KM mark because the energy grid powering that Tesla is so absurdly dirty that just straight up burning fuel in an ICE is better basically for all its life.

What we really need to move away from is the "well if everyone just gets a Tesla, the ice caps will un-melt". That's a great Musk-enrichment plan, but still awful for the environment. The energy grids powering the EVs are not going to go green anywhere near as fast as if we focused our efforts on mass transit and better urban planning.

The issue with the battery chemistries is just a cherry on the fake-green sundae. You are correct in that those batteries may reduce emissions, but the trade off is increased weight (so you're just really shifting emissions around throughout the lifecycle) and a crushing loss in efficiency in cold weather climates. There is a reason why these are coming up during a time of scarcity. You don't switch back to battery technology from decades ago due to progress, you switch because key materials are not scarce and make it impossible for you to keep up the production necessary to prop your stock price.

---

[1] https://www.reuters.com/business/autos-transportation/when-d...


The biggest emission reduction would be if every person committed suicide today, so you can always reduce emissions more with more radical plans until you get to that point.

Using nuclear power and wind with electric vehicles gets you pretty low environmental pollution for society's personal transportation needs. I think it would be easier to green the electric grid than to make everyone in America use mass transit and rebuild our cities.

Model 3 bodies are designed for a million mile lifespan and the batteries for around 300-500 thousand miles. 125,000km (~80,000 miles) is no where near its expected lifespan. The batteries are warranted to 120,000 miles. We'll see how they actually wear out, but I believe Elon Musk actually wants to help humans impact the environment less while still improving the human condition. Building cars to last instead of planned obsolescence would be in alignment with that goal.

People should really move to where they don't need heating if they want to reduce their carbon footprint. That would do more than any mass transit scheme.


There’s a huge, big incentive to increase Wh/kg though which will reduce EV mass significantly. Tesla’s 4680’s will make a decent step forwards there.


The cars getting bigger and bulkier are ICE SUVs and trucks largely, things like Cadillacs, Jeeps, fully loaded trucks being used as commuters, etc. Tesla actually started larger with S/X then moved downwards in size/cost towards 3/Y. Most of the other makers have moderate sized EVs, like Priuses, Volts, Bolts, Leafs (leaves?), they're all modestly sized.


EVs are still significantly heavier than the ICE equivalent though, compare models from European manufacturers that sell the same body in EV or ICE versions. The EV version is typically a few hundred kilogram heavier, or even more if you get a bigger battery.


Weight is certainly a variable one can point to, but EVs still end up VASTLY more energy efficient than ICE vehicles.

If you account for the hundreds of gallons of fuel inserted and burned by the ICE vehicle over time, much of which is then wasted as excess heat energy, the "weight" of an EV's battery starts to look pretty good.


Passenger cars have also gotten significantly larger since the 1980s/90s. I think its mostly to do with enhanced safety features.


It's the irony of the push for more efficiency. In theory you could have the equivalent of a 90's PC now with a fraction of the power consumption, but in practice any gain is used to create more powerful chips and do more calculations. There's a saying that was more about financial budgets, but tl;dr expenses will fill up the budget.

In my country there's irony as well. On the one side, power companies and the government invest a lot in green power production, offshore windmill farms and solar farms and the like. But at the same time, companies like Microsoft and Google come in and have datacenters built that use up all the energy that those 'green' alternatives produce. We would've been carbon neutral years ago if it wasn't for the increase in power consumption that follows production.


It's another face of the same effect that causes the Jevon's paradox. At some point it stops and further efficiency leads to less consumption, but there's nothing that says that point is achievable.


With computers it did. Everybody had a PC that used hundreds of watts, now much of that is running on 4 watt phones or 80 watt laptops.


The main power consumption of electronics is the embodied energy (energy used to build the device), and the number of devices per person has kept increasing worldwide (24 per household in the US! https://www.forbes.com/sites/tjmccue/2013/01/02/24-electroni...)


I bet the sum of those 24 devices still uses less energy than a Pentium IV and a CRT monitor.

IoT will take off at some point, and we may see a period when the many dozens of devices it needs use more power than an early 00's PC, but those will be optimized too (if IoT isn't delayed enough that they are born optimized).

I imagine there is enough demand for computing to eventually go over the 00's level, but I can't imagine its form.


Again, in a life cycle analysis, energy use from utilisation is dwarfed by energy used during production. With IoT the number of chips will keep increasing and that's what determines the total energy (carbon) footprint, which is not the same as your electricity bill.


Can't exactly conflate energy with carbon though, because manufacturers will use green energy at various stages of the manufacturing process.

Just to put numbers on this, according to this article from 2014[0], an iPhone has an embodied energy of 278 kwh, which is about 10 days of energy usage for the average American household.

[0] https://www.theatlantic.com/technology/archive/2014/10/the-e...


You have a point to a degree but a big PC+CRT is going to have those same costs and has a lot more mass to require energy during production, even compared to several mobile devices.


>We would've been carbon neutral years ago if it wasn't for the increase in power consumption that follows production.

What country is this? US power consumption is flat or trending down: https://www.eia.gov/energyexplained/electricity/use-of-elect.... I'd imagine that's the same for most developed countries, so your thesis would be wrong.


That's only because production has been moved to developing countries. Calculating consumption-based energy use (which includes energy "embodied" in things produced elsewhere, but consumed in the US), things look less great. See https://eeb.org/library/decoupling-debunked/ p. 21 (in the context of decoupling of material use and GDP growth)


My cell phone is a perfect example of what you are saying we should have had. I can connect up a keyboard and mouse and monitor. The lcd monitor also uses vastly less power.


Same with houses, new homes have better insulation and more efficient heating but just means people build bigger houses.


The carbon neutral claim is a huge stretch, but otherwise yes.


I think the comments might be missing the bigger picture. It might be a few percent improvement dissipation or range: minimal impact to you, say 2 watts per car.

But if they are shipping 800k cars per year that's 1.6 MW off the global power consumption for those new cars.


You're using the wrong units. 1.6MW global power consumption has no meaning on its own. 1.6MW is the power that two sports cars can generate.

Power consumption (e.g. energy) requires time, so 2W per car at 2 hours of daily use would be 4Wh per car per day or about 1460Wh per year per car or 1168MWh for 800k cars.

That's the electricity consumption of ~106 average US households [0].

I think that gives a much better picture of the amount of energy you're talking here.

[0] https://www.eia.gov/tools/faqs/faq.php?id=97&t=3


To an outsider, “4Wh/day” could also be expressed as “167 mW (amortized)”. Wh/day is an idiom. I think it’s strange to claim that it’s the only “right” way of expressing power consumption. For a good many people outside the field, the first thing we do when we see such a figure is convert it back to its SI units so that we can do our calculations within the system that we’re more familiar with (and something like 90% of the world uses SI).


First of all I was talking about power consumption, which is never expressed in MW or kW.

Secondly, mW is also power, not electricity consumption. 4Wh is power consumption (or electricity use if you will). The "per day"-part was just referring to the duration, hence the use of a different unit. I could just as well have used Joules and skip the "day"-part entirely to confuse the reader as to how I arrived at the final figure.

Edit: just to make this even clearer - do the dimensional analysis yourself and see why I used Wh/day. Or to put it differently: what do think is the answer to the question "say a car saves 4Wh of electricity every day. How much electricity does the car save in a year?"

Correct me if I'm wrong, but I'd say it saves 4Wh/d•365d = 4•365 Wh/d•d = 4•365 Wh (or 5.256 MJ if you insist).

> the first thing we do when we see such a figure is convert it back to its SI units so that we can do our calculations within the system that we’re more familiar with

I bet my firstborn that the vast majority of the population will not do that and are not familiar at all with the actual meaning of the figure.

Had I converted the whole shebang into Joules instead and expressed the total energy savings as 5.256 Terajoule I can guarantee that outside the odd physics major no one would be familiar with what kind of energy use we're talking about.

Look at your electricity bill and tell me what it unit it uses.


4 Wh/day = 4 W*h/(24*h) = 0.167 W. They’re equivalent, by algebra. The only way they can ever mean different things is if you use a custom algebra.


Your solution still answers the wrong question and even results in a contradiction: the initial statement was "say there's a [power] saving of 2W." Your conclusion would be that there's a power saving of 0.167W, which in incoherent with the initial assumption.

Context is very important.


> 1.6MW is the power that two sports cars can generate

If by "sports cars" you mean Bugatti Veyrons, then sure. For more "normal" sports cars, you'd need a bit more than 2.


Only if you completely ignore the tune-up and car modding scene.

On the BMW side we have the G-Power G6M V10 Hurricane CS (based on the M6 Coupé), for Mercedes there's the Mansory Black Edition (based on the Mercedes S-class/S 63 AMG), Underground Racing uses the Audi R8 V10 TT for its 1500 HP machine, the Porsche 911 has a 9ff F 97 A-Max incarnation with about 1400HP and the 9ff GT9 V-Max.

The 9ff GTurbo comes in at a mere 1200HP, based on the 911 GT3 and there's more.

So no, in the car enthusiast scene even 1MW is is not uncommon. And that's not using Lamborghinis or Bugattis. You'd be surprised how many street legal modded M3s, S-class and 911 have 800kW or more max power (at least in Europe).

Edit: shame on me - I forgot the Tesla Model S Plaid. Those have straight up 750kW electric power trains, too.



Is there a market for SiC chips in cloud data centers? I’m unfamiliar with the space, but 50% efficiency gains (if costs can be kept comparable) could be a game changer.


It's worth noting that there's also gallium nitride to consider and while it isn't as good as silicon carbide in high temperature applications or areas needing higher thermal conductivity it's better than silicon carbide as far as breakdown voltage and low resistance are concerned. Maybe eventually we'll see GaN in server power supplies and SiC in VFDs for fans, cooling, and DC infrastructure. There might even be a justification to use SiC solid state transformers to bridge between utility power, generators, battery banks, etc.

It's very important to note that 50% efficiency gains might not be as huge as it sounds given that we're just talking about losses in MOSFETs. This isn't going to be turned into your next CPU or show the same kinds of efficiency gains in logic circuits.


I didn't even think about non-silicon based chips being viable, let alone more efficient. The article mentions that SiC (Silicon Carbide) chips can be 50% more efficient. Diamond based chips can reduce energy loss by one-50,000th of that of Silicon chips.

If we can get these chips manufactured at a reasonable cost it would really help EVs get longer range. And it might even give us multi-day use out of our smart devices (phones, watches, etc.)


> If we can get these chips manufactured at a reasonable cost it would really help EVs get longer range.

How much longer though?

The motors that move the wheels consume way more energy than the microchips on board the car. Even when a car is standing still, the AC or the heating system are also quite energy-hungry compared to a few CPUs.


The power electronics in an electric drive train also have transistors. That's where you can make real efficiency gains in an electric car. Not in some microcontrollers.


Yeah but in most EVs you need to transform the battery DC current to AC for the motor.


I was under the impression the Model 3 uses permanent magnet DC motors, with the increased efficiency accounting for some of the range gains.


No electric motors are really DC driven. They do differ how the AC is derived, though, mechanically or electrically.


To say a word about the physics: The fundamental force at play is the electromagnetic force. If you run a current in a coil in the magnetic field, it will rotate so that the coil’s field comes into alignment with the outer field.

Once it’s there it is happy to stay. You’ve got to change something to get the motor running continuously. AC is quite clever about it because it just changes the field electrically and this is much more reliable than changing it with moving parts.


I've got an idea for a DC electric motor that has no apparent AC. Can anyone spot any hidden AC here?

Take a disk on a shaft. Use a ratcheting mechanism on the shaft so that it can only rotate in one direction. Attach permanent magnets around the rim of the disk.

Attach more permanent magnets to bimetallic strips positioned outside the rim of the disk. Near each bimetallic strip place a resistor.

The angular spacing of the bimetallic strips and magnets around the outside of the rim should not be that same as that of the magnets on the rim.

Turn the disk by moving one or more of the outer magnets closer to the rim to create asymmetrical torque on the disk. You move the outer magnets by using the resistors to heat the their bimetallic strips. You move an out magnet back to its original position by letting it cool back down.


That sounds like a https://en.wikipedia.org/wiki/Stepper_motor with extra steps.

You could argue that a stepper motor is also just pulsed DC, though it starts looking more like AC as you increase the drive efficiency.


Your resistors would still need a varying voltage applied to them in order to spin the rotor. If you view your "pulsed DC" as not being AC then a simple brushless DC motor would meet the same definition. I believe if you want a truly DC motor without a commutator generating AC like a brushed DC motor you would need a homopolar motor.

https://youtu.be/TzsUhiOej-s


When the resistors are heating, there must be more power going through them then when they’re cooling, right?,


Why would you use bimetallic strips instead of, say, a solenoid? The whole point of the discussion is to avoid thermal losses...


The comment I was replying to said DC motors actually use AC internally. My comment was an attempt to design a DC motor that definitely uses no AC.

Solenoids have an inductor with a changing magnetic field that stores energy and eventually gives it back. I want to avoid the possibility that that changing magnetic field will induce a current somewhere that is in the opposite direction of the normal current in that place thereby producing AC.

By just using the DC input to make heat, there is no chance of inadvertently getting AC.


> You move the outer magnets by using the resistors to heat the their bimetallic strips. You move an out magnet back to its original position by letting it cool back down.

You are switching the resistors on and off with a pulse pattern that will look a whole lot like you're driving a stepper motor. That's what we call AC.

DC means that the voltage stays constant over time. AC means it changes. Switching DC on and off creates AC.


DC is any current that flows consistently in one direction. AC is current whose flow direction switches back and forth.


The voltage driving the resistors? Just terribly slow, matching the terribly slow speed of the motor.


That would be variable DC, not AC.


I think Tesla has been migrating from induction motors to permanent magnet motors (which usually have higher efficiency and produce less heat), but they all run off of 3-phase AC from the inverter.

Sometimes the inverter/motor combination is referred to as a brushless DC motor because it takes DC as input before converting it to AC for the motor to use.


Model 3 uses two types of motors in the dual motor model and performance model. AC in the front and DC in the rear.


I think you're confusing induction versus synchronous reluctance permanent magnet motors with AC versus DC.

The motors Tesla uses are all AC, but you could also call them DC if you're referring to the whole motor/inverter assembly, which inputs DC from the battery.

DC motors aren't used much in EVs because the brushes require periodic maintenance and doing regenerative braking or running in reverse are both rather difficult whereas with a motor that runs off of 3-phase power, those features can be implemented in the software of the motor controller. They tend to be rather low efficiency, too. That said, there are some really powerful and cheap series wound DC motors that have been used to great effect in vehicles like the White Zombie.


Articles refer to the rear motors as DC similar to the DC motor used in the leaf. Here's an example:

https://www.motortrend.com/news/electric-cars-explained-gear...


I'm pretty sure the author of that article is just confused. The synchronous reluctance motors in the Model 3 run on AC just like an induction motor, but can be thought of as a DC motor if you consider the whole inverter/motor assembly, just like an induction motor.


That's certainly true, however the cooling required for the microchips also typically takes a lot of energy, as well as the microchips themselves. Reducing the energy usage, reduces the heat output, and thus also the cooling requirements.


> And it might even give us multi-day use out of our smart devices (phones, watches, etc.)

Not going to happen, they'll just find a way to use more power instead. You probably noticed that for years now, phones and the like have had a fixed battery lifetime of about a day of medium use, even though battery and chip technologies have improved. I'll admit that in terms of efficiency - performance per watt - things have increased, but the goal of the phone manufacturers is not to make things last longer, it's to min/max and find a balance between battery life and power.

With current day technology you can easily make a phone that lasts for a month on a single charge, but it'd be bulky compared to what you can do with it.


Yes, you're probably right. For proof we just need to look at RAM & CPU requirements for most software. As both of those components increased, the software grew to consume them. sigh one can wish, can't one?


Once we all own nothing and are happy though, perhaps it will be in the phone company's interest to make their phones last longer so they can get more rental income off one asset.


SiC is not a good fit for anything under ~100-200V, it's just not what it is designed for.

GaN makes more sense for consumer electronics, that's why it's popping up in a lot of devices, even if prices right now are too "high end" for most users.

As for phones, watches, etc..:

1- These technologies won't be as dense as standard silicon for quite some time (if ever.

2- We've had big swings in both efficiency and battery capacity several times already, but it works in the same way internet speed works. Power-hungry features grow with increased efficiency in the same way websites grow in size with increased internet bandwidth.


Silicon Carbide is extensively used in Power Electronics. It's just that there is a very small group of people that are having to deal with kilovolts on a daily basis, so I suppose it is not perhaps widely known.


For some strange reason, my high school had a subscription to the EPRI journal [1]. For this budding nerd, reading about SCRs the size of hockey pucks, with optical fiber triggering, was quite entertaining.

[1]: https://eprijournal.com/


As far as I understand it, these SiC chips are things like discrete MOSFETs designed to work at hundreds of volts and handle huge power running through them. I think the likelihood of this scaling down to millivolts is quite unlikely. The more likely applications for this are power stations and hooked up to solar power plants rather than in your iPhone I would expect.


You're correct that SiC is for medium to high voltage. But there's also GaN, which is for low to medium voltage, and already works in the millivolt range. Both GaN and SiC will have wider ranges in the future as the technology continues to improve.


All EVs are already highly efficient, any further increases is going to be extremely marginal.


There is still plenty to be gained from squeezing that last 1% out. Tesla is probably cell limited right now, so if your car is 1pct more efficient you can pull less cells in it and mthen make 1pct more vehicles.


For pure range, sure. But think about things like heat sinks or other cooling devices. Or the weight of these components themselves. Swap over to this new SiC, you get more direct efficiency, AND you don't need to carry as much supporting hardware!


EVs tend to be efficient at high power. They can be 94-98 percent efficient at 100kW but normal driving is something like 4kW at 45mph.

If you reduce loads by 40 to 50 watts, you can gain a mile of range. With inverter losses in the 1-2 kilowatt range at high power there is room for improvement.


Presumably you mean to rather than by?


> And it might even give us multi-day use out of our smart devices

I don't think you understand how software development works. Q: How much hardware do you need? A: How much you got?

The latest generation only gets the battery longevity because software developers still have to support the previous generation. As soon as enough e-waste happens, the gains disappear.

I might be a cynic but there sure seems to be a lot of empirical data to support my disorder.


> but electric vehicles are helping chip away at its dominance in the pursuit of energy efficiency.

I suspect the authors did not intend that pun ("chip away", get it ;-) )

Is there a name for such unintentional puns?


There is not much they can show to really lead besides switching to superior switches.

I admit Tesla's inverter design is very conservative, but it's very thorough engineering.


I'm really surprised you don't see the inverter and the motors closer coupled in Tesla's designs.

For example, by having MOSFETs spliced into the wire wound onto the stator. That minimises wire length (and therefore resistance). The heavy motor windings are a great heatsink for brief acceleration, and both will need liquid cooling for prolonged 100 mph driving up a hill.


Almost all high power switches are made to be soldered directly to cables.


The article mentions GaN not widely used because it is integrated with Si. Cree’s process is GaN on SiC, so you’d get better heat dissipation than in Si, but it costs more. There are also done processes integrating a diamond heat spreader under the GaN channel.


@madengr it seems you were secretly banned (shadowbanned)

https://ibb.co/NK6Yhj5


There's nothing secret about it:

https://news.ycombinator.com/item?id=23559004


@madengr it seems you were secretly banned (shadowbanned)

https://ibb.co/NK6Yhj5


Thanks God the beaches are running out of sand!


So wrong!

Non-silicon is 10x-100x more expensive and 10x-100x less fit for most applications.

Non-silicon (aka compound semiconductors) are only 1% of all semiconductor revenue and volume. Switching from majority semiconductor to tiny minority semiconductor is NOT going to solve any supply problems!!

SiC and GaN are STILL made on silicon and most of the circuitry is still silicon. Only the power transistors are using non-silicon.

Mostly a nonsense article.


So obvious you didn't read the article.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: