I saw your recent post on running Llama 3.3 70B on a m2 pro 64 gb. Do the many variants of apple silicon alternatives with varying numbers of cpus, gpus, and neural engines matter that much for how fast these llms can generate tokens, answer questions? More hw is always better, but what can we say how performance scales with the many different choices?
64gb ram is crucial, after that, need 1+ tb storage, and then?
I don't know. I believe memory bandwidth matters, and I got the impression that the M4 series isn't yet as good as the M2 was on that front, but I'm half-remembering things I've heard here.
define political violence acts? Taking pictures of someone doing something that might be awful to expose that they are doing something awful is what I imagine these people do. That doesn't feel like what I'd call political violence. P.V. would be trying to shoot someone, maybe bringing up a crowd to chase someone away because you don't like what they are saying. You know, threatening.
Trespassing is what they are doing if they come in and take pictures when they are supposed to be there on private property. Can you explain what you mean here?
Or maybe it's kind of like there were more than 4 different plugs for mobile phones to charger (lightning for apple, several variants of USB, some weirdo proprietary, then usb-c). A lot of people like me have 3 there way splitters (micro-usb, lightning, usb-c), so I can charge any phone.
Gas stations depend on cycling people through quickly during rush hour. A half dozen people sitting around for half an hour at a time won't spend enough to make up the difference in volume of traffic a normal station gets, and they don't typically devote space for lounges or other amenities unless they're nice long haul truck stops.
The model doesn't have to be like for like though. Set up a restaurant instead of a convenience store (think American diner style) and offer a discount on your charge for every $10 spent, or a similar integration.
Now you have a reason for someone to sit and spend money, and likely enough as EVs hit critical mass to justify costs. There are things to work out on speed of food vs charging times, potentially a queue and valet setup - but napkin maths it seems viable.
Add on to that that many gas stations make more money from going in and buying something in the store. If you stay at an EV charging station for longer, you are going to be more likely to buy something. This is a reason there are Tesla superchargers now at places like sonic in the west. Also see Buckees.
I imagine battery technology improving fast enough that it stays economically challenging to build out charging to anything like gas station level scale.
We will come to think of stopping to fill up your car routinely as quaint as lamp lighters.
Their strategy was to build them where there weren't other dcfc chargers, opening up new areas especially around outdoors places. Their chargers are much better maintained in general than the infamous Electrify America (they really struggle to fix EA chargers when they have problems, and they have problems) - there's a long story behind this.
Rivian started before the plan to use superchargers was even envisioned, so today some of those holes that Rivian was filling aren't so important because tesla superchargers are nearby.
Rivian has done a great job keeping their chargers working, have the software be reliable, just running a good service. Many and maybe most DCFC charger networks (DCFC meaning high power CCS chargers not including Tesla superchargers) really have struggled to match the uptime and reliability tesla SC. Tesla's magic ability it: (1) fix them when they break, quickly. (2) don't just build 4 chargers in a spot, it's almost the same cost to build 8 or 10. (3) string them together along main travel routes to make it possible to travel.
The entire rest of the the charging industry basically fails at those 3 things. Most of them can't manage to fix a broken charger, sometimes in months. One major reason is they often don't have standardized hardware so there might not be a spare hardware to fix something there.
I've come to notice that there is no incentive for 3rd party ev-chargers to be dependable.
Telsa and rivian have a vested interest in keeping the chargers working. Their sales depend on the chargers working. Even a little bad press is a multiplier against sales.
They also have significant engineering effort throughout the charging ecosystem, from the batteries, on-board infrastructure, standards, mapping, strategic coverage for sales and more. Lots of engineering support and problems solved quickyly
Meanwhile I kind of suspect 3rd party charging systems are probably like 3rd-party public telephones or atms -- a rent-seeking opportunity from someone who will not shell out for quick detection and fixing of broken chargers.
I noticed this years ago - chaging a non-tesla EV was a crapshoot. evgo was expensive and pretty reliable, but only ever had two fast chargers. blink was always broken - completely undependable ev chargers. chargepoint seemed ok, but only had l2/slower chargers.
There is basically no downside to these folks letting chargers be offline for a while.
I wasn't aware that EA was owned by VW, I thought it was a separate business. Suddenly all the complaints from friends with EVs make tons more sense.
From wikipedia:
(Electrify America) is a subsidiary of Volkswagen Group of America, established in late 2016 by the automaker as part of its efforts to offset emissions in the wake of the Volkswagen emissions scandal. Volkswagen, as part of its settlement following the "Dieselgate" emissions scandal, invested $2 billion in creating Electrify America.
That's true, but there's also another weird counteracting effect. Lets say you have 4 chargers offering a full 350 kw each. If they get used at full power for ~15 minutes in the month, the demand charge for the month might be ~$10k (probably will be higher, tbh). This will be true even if the stations themselves see ~15% utilization.
But break two of them? Yeah, there might be an occasional line, but your fee drops to $5k and you still produce the same revenue. TBH, some of these stations likely have better ROI when a stall or two are broken.
High utilization sites are completely different, of course.
The stations have all kinds of strategies for avoiding these max demand charges. They can slow down the charge rate to save money for them, but there are lots of places that have local battery storage to reduce the max demand charge likelihood.
You can sell something at a markup but a gas station sells so many $50 fill ups per day compared to an EV station, the gas station probably has about 1000X more revenue.
Sure their margin might be better than a razor-thin gas station margin but for the time being it's a miserable business. Unless there's another incentive, like selling cars.
So yeah. Like the poster said, there is no incentive for 3rd party EV chargers to be dependable.
I think the answer is that there needs to be, like gas stations, substantial retail/dining side-quest stuff set up to get revenue. If you don't have working chargers, people won't buy your hamburgers and pop. Or whatever.
Likely though this market is going to need some regulation/standardization.
In the end though, the demand is sporadic. Supercharger use is only for a minority of use cases (road trips) for a small segment of EV users. I've had my Polestar2 for 3 months and have yet to use one. It will never carry the same revenue possibilities as gas stations, esp as battery sizes go up.
Agree with your thoughts. To add - I've taken my Model Y on many roadtrips and have been to many superchargers. A solid number _are_ at restaurants/shops/etc. Some aren't -- but there are so many chargers that usually you can alter the planned route to include a longer charge wherever you want.
Not only will batteries get bigger, but chargers will get faster. Most of my stops now are 10-15min so often there's not really a need for any side-questing. Tesla recently added a supercharger-specific leaderboard for their in-car Mario Kart clone, which is super cool. I think we'll see some growth there for that kind of thing, but the market is obviously much lower than gas stations/etc
I've liked that 7-Eleven started putting DCFC's at some of their nicer locations in the US. The last time I did a fast-charging session was for a few-minute stop after a weekend spent away from home. The 7-Eleven was a pretty new location, easily right off the highway, and had a decent taco restaurant along with the usual snacks and pizza and roller grill items. Having a quick taco, grab some drinks, and a quick pee break for the kids gave us more than enough charge to get back home.
And the charger had a traditional credit card terminal on it. I didn't need any special app. Plug in, tap my payment, and it was charging.
I was doing a road trip through Ontario and stumbled upon a set of rest stops called onRoutes. These were pretty neat. A food court with a quick convenience store with a gas station and nice restrooms, right off the highway. They had a number of DCFC's which usually seemed in good working order, but this road trip was with an ICE so I can't speak to actual uptime. I did see a lot of cars charging, on my trip, so they seemed good.
More stops need to be like that for the 250kW+ stops on major highways. They were so easy and nice to go though.
For more highway-adjacent sit-down restaurants and areas (think Cracker Barrel kind of stops and those highway exits with lots of sit-down restaurants) I think it makes more sense to have those ~50kW chargers be the norm, maybe a few higher power ones. Have a dozen or so ~50kW chargers that'll have your car to 80-90% in an hour or so, four+ 250kW+ chargers for those quickly getting through, and a bunch of 9.6kW AC chargers for those stopping for a few hours.
I live in Ontario and haven't had a chance to use the ONroute chargers yet but I haven't heard good things. They're run by Ivy who has a bad reputation for reliability and rates.
As much as I dislike Musk personally, I will probably buy the NACS adapter and just use Tesla stations.
> Meanwhile I kind of suspect 3rd party charging systems are probably like 3rd-party public telephones or atms -- a rent-seeking opportunity from someone who will not shell out for quick detection and fixing of broken chargers.
This only makes sense if supply of chargers outstrips demand; otherwise, they're leaving money on the table by not fixing them.
I wonder what's the financial arrangements between the charging system owners and the owners of the parking lots where they're installed.
Why does 3p charging work way better in European countries like France and Germany? Talked to a guy that had a (if I recall correctly) Mercedes EV family van and he was doing a road trip and he said at least in France it was smooth going for travel.
I would think the incentive is aligned because then your not making money but you still have land rent or property taxes to pay but not generating any revenue to cover it.
Pay phones used to actually make money and get maintained by the phone company! A lot of money. And a lot of regulation. Outside of the hood, they started reducing the maintenance expenses in the mid 90s when receiving calls was pretty much banned.
Chargers are worse than ATMs, they are yet another real estate tax scheme. The actual function of charging is often not a serious consideration.
Phones were expensive, in 1960, you’re looking at the equivalent of a $75/mo bill today. The median income then was about $58k 2023 dollars, about 18% less than todays median income.
Also, urban apartment buildings weren’t always wired and while traveling you didn’t always have phones in motels. My grandfather had a bar whose clientele was mostly from a shipyard. When they’d furlough the workers for a retooling or something, he’d close the place and work as musician in a resort town for a few weeks - he’d send a postcard with a pay phone for folks to call at a certain time window.
For me it was that I wasn’t allowed to use the house phone but my girlfriend could use hers so she’d call the number of the pay phone down the street from me so it was free for me and her (but not her parents!)
Europe and China have charging standards that everyone can build to. There is more investment, more charge point operators, more charger manufacturers, and more chargers simply because companies can deploy chargers that will work with all cars.
Eventually North America will achieve standardization by building to CCS chargers with the J3400 plug but that will be a slow and gradual process. There will be many adapters and retrofits in the meantime.
Where Europe dropped the ball is not forcing gas station like behaviour onto the chargers. Now we have a bazillion different charging cards and prices differ wildly depending on the Tarif.
Or with Tesla leave your card in your wallet, you won't need it. The station IDs the car and charges the card on file in your account. No card, no scan, no tap, no need to use an app, just plug in. At least in the US. Hopefully they aren't forcing this step backwards to contactless payment by card on Teslas too when the existing system is so much better.
Yes, but currently that's always the most expensive option.
I am driving a lot from Cologne to Bremerhaven. Depending on the spot where I stop, I would need an EWE, EnBW, Aral, ionity, eon, shell or some smaller card to get somewhat reasonable prices. Annoying.
Get yourself a Tesla account. I think all of the Tesla chargers between Cologne and Bremerhaven are open to any CCS compatible car; you just need the tesla app and register your credit card. Tesla is usually cheaper as well as more reliable.
But there are also services like ChargePoint which unify the payment systems for numerous networks (certainly including Eon and Shell).
Quite a few if you count GFCIs going bad, they need to be replaced periodically. (If you test them you’ll also find that they really do have a failure mode where they keep delivering power but will no longer trip, which is a lot less noticeable than when they trip and won’t reset.)
But to your actual point: I have a feeling they’re just badly engineered. Tesla ones are so much more reliable than third parties. I’m not sure if that’s because they break less or are fixed faster, and I suspect it’s some of both, but the difference is very large.
no - poor design, lack of organization to fix them were issues.
But a key issue seems to be that EA went crazy and made their chargers out of lots of different components in a panic to get started, analogous to the PC world where there are huge numbers of different parts configurations, and even if you have only 6 major charger models, your fix it tech probably doesn't have parts of all 6. It seems obvious this was a bad strategy, but they wanted to get their chargers working as fast as possible. Instead they created a permanent bad quality reputation that may never be fixed for EA.
But it's not just EA, other companies seemed to have no plan for fixing their chargers when they break.
I mean they’re buying electricity for 10 cents and selling it for 60. They sell quite a few kw per hr that way at the fast chargers, and there are frequently lines so it’s not as if supply greatly outstrips demand.
The gross margins on EV chargers are likely pretty good as you say. But is there any evidence that these networks actually turn a net profit? These charging stations are very expensive to install and require ongoing maintenance/repairs.
The only charging network I know that is publicly traded is Chargepoint, and they have never turned a profit. In fact, their losses have increased as revenue has gone up. But, they are not a perfect example since their chargers are independently operated with prices set by the operator (with prices anywhere from free to significantly in excess of retail electricity prices).
I think the answer has to be yes. They can easily pull in a few grand a month from the electricity. They generally exist rent-free. They don’t cost that much to install.
Maintenance doesn’t seem to be even half the issue for Tesla that it is for the rest but they could also just care more. I suspect it’s both.
tesla has a factory that makes their chargers and equipment. They have standardized chargers, mass production can lead to refining the design, fixing common pain points. Most other companies seem to have taken the pc organ bank strategy where they made it out of almost careless combination of available parts - EA is the poster child.
I'm sure careful people tried to make a plan but they all ended up being such a dumpster fire of failures and low quality with no visible plan to fix them it's hard to understand.
Yeah, I assume their chargers are better engineered than the competition. And also that since they've got a national footprint they're better able to respond when one goes down. They DO go down, I've seen it, but with EA it seems there's always at least one inoperable in any bank of them.
You'd think now that Tesla is under 50% of EV sales here (and under 20% globally) that this problem will be solved soon, but I really don't know.
I'd drop Tesla from that argument. No, not that the SC network is bad.
Their CEO alienated 50% of their customer base with politics. This is going to be one of the most interesting case studies in brand management and advertising history.
But you are probably correct. I think making the infrastructure act funding contingent on reliability metrics is pretty key.
You know, if the money doesn't get yanked by the administration in a month.
So what is the incentive for operational gas pumps: Oh right, the convenience store (even more important for EVs who have to idle longer) and brand name.
50% is grossly exaggerated and kind of shows you’re in a bubble. Tesla sells in China, the world’s largest car market and by far the largest EV market, and Europe, and Australia, and Japan etc. even half of americans dont give a fuck despite some democrats claiming that. Most people dont give a fuck about the ceo’s politics except maybe in the bay area.
Tesla's sales in China are shrinking, but for ordinary car sales reasons - the competition has cheaper cars.[1]
The disaster in US automotive is that the manufacturers thought they could increase prices, and discovered that, instead, unsold cars are piling up. Tesla's oversupply can be seen from space.[2] Stellantis overdid price increases so much that the CEO was canned.
The US currently has 3 million unsold new cars.[3] One wonders how many will still start.
Jalopnik has been peddling stories about tesla’s imminent demise due to parking lots full of unsold teslas every year since before the model 3 release lol
Agreed. Tesla fans are so enamored with Elon that there isn’t really going to be anything that changes their mind. The vast majority of people I’ve talked to who said they won’t buy a Tesla because of him are not EV owners at all. Just democrats.
Once you get past the passionate technologist crowd, and there have been signs the initial buyer pool for Tesla was already stagnating a bit in 2023, who are the next natural candidates for buyers?
The mass market is 1/3 right wing, 1/3 left wing, 1/3 centrist roughly.
The 1/3 right wing are the pickup truck buyers. Like a rock ... Strong as I can be.
The 1/3 left wing is now pretty much a no.
The 1/3 centrist is probably price driven. Tesla is still too expensive for them, and if they want a Tesla, I would say they are buying a used one.
Tesla has two cars: a sedan, a crossover, in small and medium. They don't have real pickups, work trucks, station wagons, sports cars, roadsters, delivery vans, real SUVs, city cars/kei cars, or minivans. The cybertruck is a sideshow, the robotaxi is years away given Tesla's roadmap with mundane manufacturing scaling, to say nothing of the AI advance it requires and the alien nature of it to the larger buying public.
Tesla failed to utilize its name brand in ebikes, motorcycles, scooters, or even other battery/motor devices like hand tools, lawn tools, snowblowers, riding mowers. Obviously no heavy equipment aside from some pilot Semis. No RVs.
The Solar business is meh. The grid business likely will be outcompeted by sodium ion/lfp/other schemes. Home storage is slow uptake, and again they'll probably be outcompeted in price.
Tesla has no more tech advantage. Their EV drivetrain may be a percent or two more efficient, but likely not enough to matter. NMC cylindrical cells are about to be economically noncompetitive with 150 wh/kg sodium ion and 200-225 wh/kg LFP, and denser stuff is on the 3 year roadmap. The battery day 4680 tech is basically undelivered or won't scale.
I just don't see the growth path for this company anymore barring miraculous AI combined with miraculous execution combined with miraculous domestic battery supply. About the only thing I can see floating the company it terms of real sales figures would be the Semi, but it sure looks like a soft launch to me that is still a year away, and something that will need either sulfur chemistries, solid state, or dirt cheap high density LFP/Sodium Ion to be economically viable.
I think Tesla has probably missed the boat on the semi, at least in Europe.
Scania has already delivered plenty of heavy goods EVs here in Norway. Delivery companies like the Post office, DHL, ASKO are investing heavily in large EVs but most of that is going to either Swedish Scania. Chinese SAIC is also selling well for urban delivery vehicles.
Their problem is that their valuation had the market predicting that they would generate a MAJORITY of all profit in US automotive manufacturing - they were worth more than every other player combined. This was grossly unrealistic. So expect a down round. Whether they survive that drop to... What... 5-10% of their former value? Depends on what sort of explicit and implicit dental they have accrued to scale.
Tesla was using CATL batteries in part. Tesla uses different batteries in diff cars, also some come from China, some other places.
Tesla said they'd have amazing new 4680 batteries but that didn't really succeed, a reason why the ct has lower range than planned is the new higher density batteries didn't arrive.
Tesla has one advantage left, they can make their cars at mass quantities and make a profit. Most of the western auto companies can't do that, unless they charge very high prices.
The Internet says 50% of US buyers are identified democrats, YMMV for internet stats. That number doesn't seem stupid -- EVs are progressive, environmentalist, anti-establishment.
Tesla's stock is a huge bubble. It's worth the next 10 car companies combined! Yes there's a ton of AI speculation in there, but fundamentally the company's stock is about geometric growth in sales.
The most important customer to a car company is their current ones, since those have a very high chance of being loyal to the brand when they buy again.
Tesla is a company that does almost no advertising. Maybe that's because the press gives them enough hype, but I suspect there is a huge degree of passionate advocacy. I should know, I ... was ... a strong advocate for Tesla. I still am for EVs, but they were one and the same with Tesla for 20 years.
A subtle thing happened with the Tesla owners I know. They went from advocating for their car, to rationalizing it, within the span of six months. I don't think it is a stretch to say that Tesla owners were passionate advocates for their vehicles. All progressive, environmentalist people I know are now hell-nos.
Point is, Tesla is stock bubble. Their sales are already down this year, and I'm picking up the popcorn to see what happens in Q3/Q4. The trade war will probably make Tesla a noncompetitive product in China. So US sales are going to get MORE important, not less.
People are actively selling their Teslas. Sure, it's on the used market, what does that matter? I don't have stats, but how many sales of new Teslas will a robust used market (which also has Hertz dumping cars onto the market) cannibalize? I'd say each used Tesla is half of a new one.
Point is, I don't see Tesla as a stock that can rationalize bad sales. They don't have a brand-spanking new car for a new mass market: the Cybertruck is explicitly a miss from the pickup truck market. The Semi still can't seem to scale production. The robotaxi is, face it, two years away and a fundamental leap in AI from working.
Trump is going to be on television reminding every democrat who helped him vocally, publicly, monetarily, possibly illegally get him in office, and will join an agency that will target progressive government systematically and with high visibility.
All it takes is, what, 1/3 of the 50% democrats of their potential customer base to say "hell no" and that's ... 16% drop in sales. Ever seen a company lose 16% of sales?
> Tesla's Model Y is still the best selling car model in the world in 2024
Do you have a source for that? Tesla is not even in the top 10 largest auto manufacturers, but the larger ones have way more models so I suppose it is possible Model Y is the best selling car (though it certainly isnt in the US).
The geography matters, but just speaking for the US, rational buyers who want an EV will learn their lesson and come to Tesla. Maybe after one last ICE mistake but eventually.
Superchargers are not just a bullet point, having them or not is a dividing line between "it just works" and "life is miserable" for EV owners. And only a small subset of them are open to non-Teslas for who knows how long.
That is simply not true. You are speaking about the past. In the present and the future, especially as superchargers open up to other brands and more NEVI chargers get built in the second half of this decade, I see absolutely no reason people would “come to Tesla.”
I would put money on the opposite, and on people moving to other brands.
Superchargers opening up will not be overnight. It will take years.
Prices will be higher for non Teslas and will require inconvenient steps and even membership plans to get a slight price break. And there will be fewer features. When the chargers are even ready. A few are, but don’t be fooled to think that means bad times are in the past for non-Tesla EVs.
As to other chargers being built, sure, but scale and ubiquity is what matters. Not having one fast charger at the odd Four Seasons resort. I’ll believe it when I see it.
Tesla is not importing to china. Its not debt ridden like the 10 other car companies, and has a growth outlook far beyond them via FSD and renewables. They produce 1.5M cars a year, Hertz is a blip. FSD13 is looking fantastic.
What competitive advantage does Tesla have in renewables?
They produced 1.5 million. Their sales are down, they have no new models releasing, they have competition on all fronts, they likely will lose the Chinese market in the trade war. They need to sell 20 million vehicles to justify their stock.
If Tesla has a kuckass sub 10000$ sodium ion city car (aka the car for 1-3 billion people)... But they have luxury vehicles and that's it.
Fully auto drive is a convergent evolution of infrastructure problem. It is ten years away. Army of Indians won't cut it.
When I look at my Tesla, I see the hard work of hundreds (or thousands) of engineers, tasked with doing something new and complicated. They're collective trailblazers, and they know they're making the product that best it can be, given the constraints.
They were once trailblazing, while their CEO is lighting progressive environmentalism on fire.
Tesla is building a moat. They are defending their market share, preventing subsidies aiding competitors, keeping China out of the US market. That's not trailblazing, its not innovative.
Tesla needs to jettison their CEO. He's toxic and destructive to the mission of the company, or at least the mission it used to have.
It seems gas stations get 70% of the profits from the conveience stores.
EV charging stations are pretty standalone. They could probably do the mcdonalds thing - become real-estate companies. Leasing surrounding land to convenience stores could be a good model.
I don’t think most people make purchasing decisions based on the personal politics of a firm’s CEO. It will be interesting to see how impactful it ends up being.
For a consumer to buy an EV is one that involves other brand factors: politics/environmentalism was a large component of initial buyers. Tesla's inception was around the time of Michael Moore's who killed the electric car.
The core demographic of the electric vehicle in America is still Urban.
Most CEOs don’t attempt to directly influence an election to align with their personal politics by spending a quarter of a billion dollars on a candidate’s campaign, either.
As a cybertruck owner who likes to drive deep into national parks, I'm optimistic about this. Tesla covers the big highways really well with their superchargers, but where they fall short are those adventure/outdoors destinations. E.g. the north cascades loop in WA state. You can find the occasional level 2 charger but they are painfully slow and aren't maintained well. Prior to the CT, Tesla didn't have "adventure-capable" vehicles, so serving those locations was not a priority. Hopefully with Rivian's DNA being more outdoors-oriented, everyone with a beefy EV can benefit.
It's a pleasure to drive and it has been great for what I use it for. Mostly family stuff but the occasional camping trip in the middle of nowhere. A Rivian R1S probably would have been more practical for me due to the 3 rows of seats. But I love the angular look. It's definitely polarizing, no question.
Some downsides: snow/ice can block the headlights, the windshield wiper and washer fluid sprayer are bad, the range isn't good compared to competitors.
I've not been a huge fan of Tesla generally because of the lack of physical buttons, as well as build quality.
I think my next vehicle, will be at a minimum plug in hybrid, and maybe full BEV, ironically, replacing our dual electric tank hot water heaters allows for us to install a charger.
OP here. Yeah I talk about them a lot. All of the other topics I could talk about, I seldom have original content to contribute. All of the interesting things get said pretty quickly for software engineering, living in a west coast city in the USA, not enjoying ads, etc. But for this one, I'm a relative rarity on this site and I'm enthusiastic about the topic, so there are more frequent opportunities for me to say something unique.
Superchargers also have the benefit of only having to support Teslas. It is very relevant to note that the rollout of using superchargers by other brands has been a giant clusterfuck.
Tesla’s magic is excellent, quick, and cheap manufacturing coupled with their own dedicated software for them. And constant iteration.
EA is a pathetic failure because its a hodgepodge of outsourced components (so very difficult to iterate) running on some shitty outsourced windows based software.
That and it seems like VW can't reliably deliver any decent software on any kind of platform. It's incredible how they have so many cars hobbled with some of the worst UXs on the market. It's not even like most of their competition is that competent, they're just exceptionally bad.
Look at a lot of the reviews of the ID series, like the recent ID.Buzz.
There's only 3 plug formats that matter. J1772 for AC charging, the Tesla plug (called NACS), and CCS, the other main standard. It may look messy but it's trivial in practice, once you can use superchargers.
Once you have access to superchargers you can just drive. I've driven across the country 3 times in a Tesla, now I have a different EV with a CCS plug and access to superchargers with an adapter, I'm free to go where I want.
It's not only the plug shape though, it's also the language the plug speaks. Newer Tesla chargers (I think supercharger v3 and up) will be able to talk CCS over the NACS connector for wide compatibility, but the older ones will look the same and only charge a Tesla because they don't communicate with CCS.
That "old superchargers don't work with CCS" is true yet it's not really a problem. First, the software sites that show you chargers know about the different versions. Tesla and even Rivian show you only the superchargers you can use.
The old superchargers (called v2) used the original protocol (CANbus) which is different than CCS's protocol. Newer Tesla superchargers speak the original proto and the ccs wire protocol. All new chargers uses the new protocol. Tesla is slowly replacing their chargers as they age out. At the same time this transition helps reserve a few chargers for teslas (the older v2 ones).
What this means is you use your incar app or a phone app to find chargers, and you only see the ones that you can use. This works for Rivian, Tesla, but also GM and Ford. It's a messy issue for sure, but it's turns out not to be an issue.
Everyone's car already knows what chargers they can use. The card doesn't route you to incompatible chargers. This is a solved problem. It sounds like you just don't have an EV.
It's more a note for any apartment dwellers thinking "I see Tesla chargers all over town, I could buy a 2025 Equinox (or whatever) with NACS and swing through the one on next to my office when I need to. If it's a v3 or v4 that's true, if it's an older station it's still only for Teslas.
I think you're overestimating normal car buyers, as EVs continue to shift from early adopter novelties into the mass market. People will see a charger and expect to be able to pull up and charge there, like they've been doing with gas stations for decades.
SAE is thankfully working on universal plug and charge standards which will be a huge help for the other problem - the mess of different apps for each charging network. Because just providing a credit card reader like every gas pump ever was too hard.
Why do these new machines have 192 cores, not a power of 2? (3* 2^6) = 192. The article mentions 256 and 512 cores coming. I get that various components can limit scaling like say bus speed or capacity. But 192, that's wacky.
> Architectural design: The physical layout of the chip may favor certain core arrangements. For example, a 48-core processor could be designed as a 6x8 grid, which efficiently utilizes the square shape of the silicon die6.[1]
I disagree about trusting the CDC, I have no reason not to trust them.
They'll do their best job to understand threats, avoid making confusing statements. But if there is a threat and people are dying in hospitals in large numbers without much information, they'll do their best to make health safety precautions, but they will probably make some mistakes. This is pretty common in science too.
The CDC never had the trust of many people to start with, but their repeated failures at communication didn't do them any favors. They also repeatedly allowed themselves to be bullied and manipulated. For example, they changed the recommendation to stay home 5 days after testing positive, but that change wasn't based on any new evidence that would contradict their old recommendation.
Yes, good points. Politicians and the public and all of us can have trouble when it's a rush situation with deadly consequences and the experts make their best prognosis based on previous times in some part - then they revise best practices as they gain knowledge.
This happens all the time but it's not usually so visibly combined with conspiracy theories about controlling people and all that kind of crazy stuff. America is so influenced by foolish postings on social media and popular, but not fact-based TV networks.
There was the realization that covid spread in a different way than the flu and the discovery of things spreading through the air was a bit surprising and unexpected, compare with traditional flu issues.
64gb ram is crucial, after that, need 1+ tb storage, and then?