And if Tesla said something about battery swaps to their customers that was credible, the news would spread, starting most likely with their own discussion forums. Besides that, I'd dispute that Tesla owners are really getting any sort of inside info, particularly since all the Tesla owners that you know appear to be wrong on this subject.
I work for a large annual convention, and at our most recent event we took our first baby steps into using pallets. We got a bunch of plastic pallets and a pallet jack to move them.
For us, it's about doing our load-in and load-out at the venue quickly. Before, we'd have one or two trucks doing the rounds between our storage space and the venue. The trucks would arrive, and then we'd load everything into the trucks one piece at a time (using boxes, at least), and then the trucks would go off and we'd sit on the loading dock waiting for them to return. With pallets, the trucks arrive, we stick the pallets in with the pallet jack, the trucks (very quickly leave), and then while waiting for the truck to return we're loading stuff onto pallets and wrapping it for the next truck. By the time the truck has returned, the pallets are ready to load.
We're not fully converted to using pallets, but the first steps have already shown us how much extra speed we can get out of them, mostly just by letting our people work on loading the trucks before the trucks even get to the dock.
There are dozens of languages that compile to the.NET CLI, including BASIC, C++, Ruby, PHP, Java, JavaScript, Python, Lisp, Pascal, Perl, Scheme, etc. C# is the most popular language to compile to the CLI, yes, but almost any other common language out there can be used too.
Sort of: they're in a flat pack (kind of like a big skateboard) that is bolted to the bottom of the car. The battery pack does give the car a good deal of structural support and rigidity, but it can be removed relatively quickly and swapped out with another.
The batteries are permanently embedded in a giant battery pack that takes up most of the bottom of the car. The battery swap station is replacing that, not the individual cells inside it.
The battery stations and superchargers are meant to charge in between cities when driving between them. The assumption is that your normal charging happens at home or at your destination.
It seems like they're going to be doing part or all of the battery swap manually, so the improvement from 3 minutes down back to the target of 90 seconds is more about getting everything automated again rather than simply improving the process. It's not practical to roll out large numbers of battery swap stations all over the world if they need a pit crew at each one.
From their press release, it sounds like the culprit is the additional armour that was added to the car to avoid damage to the battery packs from road debris. The original swap demonstration was fully automated, but then they went and stuck a bunch of other stuff on the underbody, invalidating their existing automation work for it.
You can buy hardware with lower-power Intel chips from NewEgg, you just can't buy the CPU by itself. It's not in your price range, but here's a Zotac mini computer for $380 (which includes the RAM and SSD): http://www.newegg.com/Product/...
They also sell numerous tablets/laptops/etc with that processor.
For an Intel motherbaord/CPU/RAM combo, you're looking at around $40 for the motherboard (all these prices from NewEgg US), $30 for 4GB of RAM, leaving you up to $130 for the processor. That puts it at an Intel i3-4360, a high-clocked dual-core Haswell part.
For AMD's part, assuming the same CPU budget, you're looking at an A10-6800K.
Right off the bat we can notice a disparity in TDP, being 54W on the Intel chip and 100W on the AMD chip.
In terms of performance, AnandTech Bench conveniently has both of those chips in their system. The benchmarks show that the Intel chip is faster (sometimes substantially) in almost every benchmark, be it single or multithreaded... And all that while using much less power.
After looking up those results, though, I realized that that AMD chip was a Richland, while there is the newer Kaveri cores available. There is also a Kaveri CPU at the same $130 pricepoint (but with a lower clockspeed/model number), the A10-7700K, with a TDP of 95W. That one is unfortunately not in Bench, but looking at other sides indicates that it manages to narrow the gap substantially, while still generally being slower than the i3 chip. But in order to do this, it uses 10W more power at idle and 26W more power at load.
Intel's got Haswell chips at decent clockspeeds down to roughly the $40 pricepoint, where I expect they'd still compete favourably with AMD. Below that is nothing, and I suspect there might be a point somewhere between $40 and $130 where AMD makes more sense.
All this said, I'm disappointed in AMD. I don't want Intel to beat them, I want AMD to put out parts that are competitive in price, performance, AND power consumption. Most of my early CPUs were AMD chips. I've owned a K6-2, a Duron, and an Athlon XP, and all were fantastic. The Athlon 64 was also great. But ever since then, AMD has been behind, and the lack of competition has certainly not helped the market. I keep hoping that AMD will put out something new that wows me, a completely new architecture that shakes things up like Conroe did for Intel or ClawHammer did for AMD. But thus far they seem to keep iterating on the same non-competitive designs.
Unfortunately, it seems like ARM is more likely than AMD to bring competition to the market as they keep slowly creeping up the TDP ladder, and we're just now starting to see them going head-to-head with Intel in the PC space with lower-end Chromebooks. nVidia's Denver core is supposed to be competitive with Haswell in terms of performance/power, but I can't find any benchmarks that directly compare them.
It's not entirely clear to me that ARM chips offer better power efficiency than Intel chips when comparing modern parts at similar performance targets. There's not traditionally been much in the way of comparison points between them, because only very recently have ARM chips and Intel chips begun overlapping in terms of power envelopes.
I'd be interested in seeing a comparison between nVidia's Denver cores (or the A15) and Intel's new Core M parts. I believe they have similar TDPs.
So, if NewEgg doesn't sell it, then Intel doesn't make it? Intel's lower power parts are packaged to be soldered directly onto the motherboards of tablets and notebooks, places where people don't replace CPUs. NewEgg doesn't sell those parts. It's also worth noting that TDP is going to be the maximum power draw, but isn't going to tell you how much power it takes to accomplish a given task. Two CPUs with the same TDP can exhibit dramatic differences in battery life in mobile products.
Your implication was that AMD offers better power efficiency than Intel chips. This is false. Intel's parts offer either substantially more performance at similar power consumption, or substantially less power consumption at similar performance. This is true from tablets through servers. But AMD is still competitive on price. This has led to them holding onto some market share, but it means that they have minimal presence in the mobile space, since power efficiency is a big deal in tablets and ultrabooks.
i3 processors start at a TDP of 11.5W and are almost as fast as those 25w amd chips despite using less than half as much power. AMD chips have not been able to come remotely close to the performance per watt of Intel's chips since Conroe launched in 2006. They compete on performance per dollar, not power efficiency.
Lots of Canadian companies trade with or actually operate in Cuba, and none of them are facing any sort of issues in the US. I realize that the Helms-Burton act does enable the US to sanction such companies, but it seems that in practice the sanctions are not applied.
I founded SuprnovaRadio (or one of it's incarnations) back in 2003, good times. Still own the domain that hosted the site and streams (novasearch.net) that it was hosted on back then too, although it hasn't pointed to a server in ages.
While I'm generally also skeptical, when JPEG2000 was released, decoding images in JavaScript wasn't an option. As such, there's not really any barrier for individual websites to switch, if they're heavily image-driven.
A bigger roadblock might be that these days, bandwidth (and storage) is cheap, and so savings in image size are less relevant than they used to be.
OVH has been doing this for more than a decade. They credit it as one of the reasons they're able to undercut competitors by so much, by eliminating most of their cooling costs. They get their power usage efficiency, which is the ratio of IT equipment power consumption versus facility power consumption, under 1.1 for their newer datacentres.
China's launch record is quite good, but when they fail, huge numbers of people die. China doesn't use launch abort systems like the US does. When a US rocket fails, it is destroyed immediately to minimize damage. When a Chinese rocket fails, people die. One launch in the 90s killed over 500 people when a Long March III basically bombed a town and they couldn't stop it due to the lack of any abort system. Atypical to be sure, but they've had other launches that have killed people, and these are just the ones we know about.
Basically what I'm saying is that China really needs to put launch abort systems in their rockets.
What is a phone encrypted/locked with? A password.
And what is the U2F protected by? Nothing. Anybody who gets hold of the dongle can use it, at least getting into the system protected by a mobile app would require them to steal the device *AND* get the password. And not all phones are locked with a password. There are phones locked with biometrics, or patterns that couldn't quite be called a password.
As a company, I wouldn't rely my security on unlock passwords.
So you wouldn't rely on a system that requires a device be stolen and then its password cracked, but you WOULD rely on a system that only requires the device be stolen with no password required?
How often do you enter your unlock password when other people could, in theory, watch you?
Once every few weeks, maybe every few months. I'm not using a password as the primary means to unlock my device.
With remote-wipe you can never be sure whether the attacker didn't crack the phone
With a dongle, you can be 100% sure that the thief has instant and complete access, because there isn't even a password to protect it.
and now just sent a fake "I'm wiped" message.
Do you have any evidence that it is possible to spoof the success of remote wipe on a modern smartphone running the current operating systems? On top of this, there is also the possibility of de-authorizing the device on the server-side with the 2FA provider.
Do you want to copy supid strings from your phone to your computer?
Considering how much more secure the system is than a USB dongle that anybody can grab and plug in, yes. Very much so.
U2F protects from those too by also authenticating the server.
It authenticates that somebody plugged in my dongle. It doesn't authenticate that I was the one who did it.
The smartphone can be lost/forgotten, but at least smartphones tend to be encrypted/locked with the option to remote-wipe. A U2F dongle that is lost would seem to offer no such protection.
The apps for 2FA services tend to offer a rotating key, so it's not a fixed password that can be guessed.
Best on the market? Errm, it has a bunch of deal-killer restrictions. It requires that the device that you're trying to log in on have USB ports (sorry smartphone/tablet users) and you need to carry around a physical token for you to lose/forget instead of having an app on your smartphone. And while it doesn't require any software be pre-installed on the computer (since the device basically simulates a keyboard), it still requires that the system be configured to let random keyboards/USB devices be plugged in.
Slashdot Mirror
And if Tesla said something about battery swaps to their customers that was credible, the news would spread, starting most likely with their own discussion forums. Besides that, I'd dispute that Tesla owners are really getting any sort of inside info, particularly since all the Tesla owners that you know appear to be wrong on this subject.
I work for a large annual convention, and at our most recent event we took our first baby steps into using pallets. We got a bunch of plastic pallets and a pallet jack to move them.
For us, it's about doing our load-in and load-out at the venue quickly. Before, we'd have one or two trucks doing the rounds between our storage space and the venue. The trucks would arrive, and then we'd load everything into the trucks one piece at a time (using boxes, at least), and then the trucks would go off and we'd sit on the loading dock waiting for them to return. With pallets, the trucks arrive, we stick the pallets in with the pallet jack, the trucks (very quickly leave), and then while waiting for the truck to return we're loading stuff onto pallets and wrapping it for the next truck. By the time the truck has returned, the pallets are ready to load.
We're not fully converted to using pallets, but the first steps have already shown us how much extra speed we can get out of them, mostly just by letting our people work on loading the trucks before the trucks even get to the dock.
There are dozens of languages that compile to the .NET CLI, including BASIC, C++, Ruby, PHP, Java, JavaScript, Python, Lisp, Pascal, Perl, Scheme, etc. C# is the most popular language to compile to the CLI, yes, but almost any other common language out there can be used too.
I'm not sure. The additional armour that they added consisted of three components, some of which aren't located under the battery.
Sort of: they're in a flat pack (kind of like a big skateboard) that is bolted to the bottom of the car. The battery pack does give the car a good deal of structural support and rigidity, but it can be removed relatively quickly and swapped out with another.
The batteries are permanently embedded in a giant battery pack that takes up most of the bottom of the car. The battery swap station is replacing that, not the individual cells inside it.
The battery stations and superchargers are meant to charge in between cities when driving between them. The assumption is that your normal charging happens at home or at your destination.
Why would Tesla owners know anything more about Tesla's future plans than anyone else?
It seems like they're going to be doing part or all of the battery swap manually, so the improvement from 3 minutes down back to the target of 90 seconds is more about getting everything automated again rather than simply improving the process. It's not practical to roll out large numbers of battery swap stations all over the world if they need a pit crew at each one.
From their press release, it sounds like the culprit is the additional armour that was added to the car to avoid damage to the battery packs from road debris. The original swap demonstration was fully automated, but then they went and stuck a bunch of other stuff on the underbody, invalidating their existing automation work for it.
You can buy hardware with lower-power Intel chips from NewEgg, you just can't buy the CPU by itself. It's not in your price range, but here's a Zotac mini computer for $380 (which includes the RAM and SSD): http://www.newegg.com/Product/...
They also sell numerous tablets/laptops/etc with that processor.
For an Intel motherbaord/CPU/RAM combo, you're looking at around $40 for the motherboard (all these prices from NewEgg US), $30 for 4GB of RAM, leaving you up to $130 for the processor. That puts it at an Intel i3-4360, a high-clocked dual-core Haswell part.
For AMD's part, assuming the same CPU budget, you're looking at an A10-6800K.
Right off the bat we can notice a disparity in TDP, being 54W on the Intel chip and 100W on the AMD chip.
In terms of performance, AnandTech Bench conveniently has both of those chips in their system. The benchmarks show that the Intel chip is faster (sometimes substantially) in almost every benchmark, be it single or multithreaded... And all that while using much less power.
After looking up those results, though, I realized that that AMD chip was a Richland, while there is the newer Kaveri cores available. There is also a Kaveri CPU at the same $130 pricepoint (but with a lower clockspeed/model number), the A10-7700K, with a TDP of 95W. That one is unfortunately not in Bench, but looking at other sides indicates that it manages to narrow the gap substantially, while still generally being slower than the i3 chip. But in order to do this, it uses 10W more power at idle and 26W more power at load.
Intel's got Haswell chips at decent clockspeeds down to roughly the $40 pricepoint, where I expect they'd still compete favourably with AMD. Below that is nothing, and I suspect there might be a point somewhere between $40 and $130 where AMD makes more sense.
All this said, I'm disappointed in AMD. I don't want Intel to beat them, I want AMD to put out parts that are competitive in price, performance, AND power consumption. Most of my early CPUs were AMD chips. I've owned a K6-2, a Duron, and an Athlon XP, and all were fantastic. The Athlon 64 was also great. But ever since then, AMD has been behind, and the lack of competition has certainly not helped the market. I keep hoping that AMD will put out something new that wows me, a completely new architecture that shakes things up like Conroe did for Intel or ClawHammer did for AMD. But thus far they seem to keep iterating on the same non-competitive designs.
Unfortunately, it seems like ARM is more likely than AMD to bring competition to the market as they keep slowly creeping up the TDP ladder, and we're just now starting to see them going head-to-head with Intel in the PC space with lower-end Chromebooks. nVidia's Denver core is supposed to be competitive with Haswell in terms of performance/power, but I can't find any benchmarks that directly compare them.
It's not entirely clear to me that ARM chips offer better power efficiency than Intel chips when comparing modern parts at similar performance targets. There's not traditionally been much in the way of comparison points between them, because only very recently have ARM chips and Intel chips begun overlapping in terms of power envelopes.
I'd be interested in seeing a comparison between nVidia's Denver cores (or the A15) and Intel's new Core M parts. I believe they have similar TDPs.
So, if NewEgg doesn't sell it, then Intel doesn't make it? Intel's lower power parts are packaged to be soldered directly onto the motherboards of tablets and notebooks, places where people don't replace CPUs. NewEgg doesn't sell those parts. It's also worth noting that TDP is going to be the maximum power draw, but isn't going to tell you how much power it takes to accomplish a given task. Two CPUs with the same TDP can exhibit dramatic differences in battery life in mobile products.
Your implication was that AMD offers better power efficiency than Intel chips. This is false. Intel's parts offer either substantially more performance at similar power consumption, or substantially less power consumption at similar performance. This is true from tablets through servers. But AMD is still competitive on price. This has led to them holding onto some market share, but it means that they have minimal presence in the mobile space, since power efficiency is a big deal in tablets and ultrabooks.
i3 processors start at a TDP of 11.5W and are almost as fast as those 25w amd chips despite using less than half as much power. AMD chips have not been able to come remotely close to the performance per watt of Intel's chips since Conroe launched in 2006. They compete on performance per dollar, not power efficiency.
Hey thanks for not making a huge deal out of castros death
I think I'm missing part of this joke, because Castro isn't dead...
Lots of Canadian companies trade with or actually operate in Cuba, and none of them are facing any sort of issues in the US. I realize that the Helms-Burton act does enable the US to sanction such companies, but it seems that in practice the sanctions are not applied.
The problem is that the US courts ruled that US law does apply in Ireland because Microsoft has a presence in both countries.
"Our words are backed with NUCLEAR WEAPONS!"
- Ghandi
I founded SuprnovaRadio (or one of it's incarnations) back in 2003, good times. Still own the domain that hosted the site and streams (novasearch.net) that it was hosted on back then too, although it hasn't pointed to a server in ages.
While I'm generally also skeptical, when JPEG2000 was released, decoding images in JavaScript wasn't an option. As such, there's not really any barrier for individual websites to switch, if they're heavily image-driven.
A bigger roadblock might be that these days, bandwidth (and storage) is cheap, and so savings in image size are less relevant than they used to be.
OVH has been doing this for more than a decade. They credit it as one of the reasons they're able to undercut competitors by so much, by eliminating most of their cooling costs. They get their power usage efficiency, which is the ratio of IT equipment power consumption versus facility power consumption, under 1.1 for their newer datacentres.
Sorry, yes. I mean a flight termination system, not a launch abort. Either way, they (and the Russians) really need to use them.
China's launch record is quite good, but when they fail, huge numbers of people die. China doesn't use launch abort systems like the US does. When a US rocket fails, it is destroyed immediately to minimize damage. When a Chinese rocket fails, people die. One launch in the 90s killed over 500 people when a Long March III basically bombed a town and they couldn't stop it due to the lack of any abort system. Atypical to be sure, but they've had other launches that have killed people, and these are just the ones we know about.
Basically what I'm saying is that China really needs to put launch abort systems in their rockets.
What is a phone encrypted/locked with? A password.
And what is the U2F protected by? Nothing. Anybody who gets hold of the dongle can use it, at least getting into the system protected by a mobile app would require them to steal the device *AND* get the password. And not all phones are locked with a password. There are phones locked with biometrics, or patterns that couldn't quite be called a password.
As a company, I wouldn't rely my security on unlock passwords.
So you wouldn't rely on a system that requires a device be stolen and then its password cracked, but you WOULD rely on a system that only requires the device be stolen with no password required?
How often do you enter your unlock password when other people could, in theory, watch you?
Once every few weeks, maybe every few months. I'm not using a password as the primary means to unlock my device.
With remote-wipe you can never be sure whether the attacker didn't crack the phone
With a dongle, you can be 100% sure that the thief has instant and complete access, because there isn't even a password to protect it.
and now just sent a fake "I'm wiped" message.
Do you have any evidence that it is possible to spoof the success of remote wipe on a modern smartphone running the current operating systems? On top of this, there is also the possibility of de-authorizing the device on the server-side with the 2FA provider.
Do you want to copy supid strings from your phone to your computer?
Considering how much more secure the system is than a USB dongle that anybody can grab and plug in, yes. Very much so.
U2F protects from those too by also authenticating the server.
It authenticates that somebody plugged in my dongle. It doesn't authenticate that I was the one who did it.
The smartphone can be lost/forgotten, but at least smartphones tend to be encrypted/locked with the option to remote-wipe. A U2F dongle that is lost would seem to offer no such protection.
The apps for 2FA services tend to offer a rotating key, so it's not a fixed password that can be guessed.
Best on the market? Errm, it has a bunch of deal-killer restrictions. It requires that the device that you're trying to log in on have USB ports (sorry smartphone/tablet users) and you need to carry around a physical token for you to lose/forget instead of having an app on your smartphone. And while it doesn't require any software be pre-installed on the computer (since the device basically simulates a keyboard), it still requires that the system be configured to let random keyboards/USB devices be plugged in.