8 ms·

> The customers who ordered this option will instead receive the 60 kWh pack, but range will be software limited to 40 kWh. can't see anybody hacking that.

by nikcub 14y ago

seanmcdirmid 14y ago
> ... and can be upgraded to the range of the 60 kWh upon request by the original or a future owner. Does anyone remember Intel's co-processor "upgrade" for turning a 486SX into a 486DX?
- kragen 14y ago
  Yes. It was called the "487SX", and was billed as a floating-point coprocessor, but it was actually a 486DX with a lower pricetag and a different pinout. When you installed it, the motherboard disabled your 486SX and just used the "coprocessor" for everything. What's the relationship to Tesla's crippleware strategy here? I'm not seeing it.
  - seanmcdirmid 14y ago
    Oh, it was worse than that: the 486SXs themselves were actually 486DXs with a disabled FPU. So not only were you buying a 486DX to "upgrade" your CPU, it was basically the same CPU!
    - jacquesm 14y ago
      'Disabled' sounds like there might have been some way to re-enable them, but iirc there was a whole bunch of traces that were zapped with a laser, making the disabling a rather permanent affair.
      - seanmcdirmid 14y ago
        Rumor has it there were a few people who were good enough with a soldering iron to re-enable them.
        brk 14y ago
        Huh? The traces were buried inside the ceramic chipset. And even at the scale of technology in that era the soldering was usually done with vibration, not heat. I find this statement rather implausible. It would take far too much time and equipment to make it worthwhile vs. just upgrading your CPU "over the counter".
        jacquesm 14y ago
        I think that was a joke and I think it went over your head :) I do know a guy that can solder scary small stuff (SMT, angel hair and other incredible feats of steady handedness) but I think that re-bonding zapped chips is even beyond his soldering capabilities. If only because you first have to get access. I did once re-solder a pin on a 386 which had broken off and I was pretty proud of that but for the most part once you get to the package boundary it is game over.
        brk 14y ago
        "I think that was a joke and I think it went over your head :)" As had I. My first computer was a PCjr., with a hand-soldered memory expansion card. I repaired broken pins on processors multiple times (maybe like 5 or 6), since in that era I was also building "custom" machines for people. But in regards to this being a "joke", it just seemed like really bad folklore to me. Having lived through that era, I just really didn't see even any remote humor in the comment.
        seanmcdirmid 14y ago
        You guys are young! To the downvoter, this was a very common claim as far as computer folklore goes: http://en.wikipedia.org/wiki/Talk%3AIntel_80486SX http://en.wikipedia.org/wiki/Talk%3AIntel_80486SX > The joke about the drill through the precisely correct location is likely a joke made by someone who knew what they were talking about. It was likely not meant to be taken literally, but sort of a tease without explanation. The floating point unit was in fact never removed with a laser, it was disabled by tying a "disable floating point" input pad to power or ground using a bond wire in the package. This joke was probably a reference to the fact that if you disconnected that bond wire, the default state of that input pad was to enable the FPU. So I think the idea behind the joke was that you could disconnect that bond wire by drilling through the bond wire and thereby disconnecting it. The practical way to do this would be to flip off the lid on the PGA package to expose the die and bond wires. People did have a steady enough hand to straighten bond wires on occasion when that were shorting to adjacent bond wires. With a knowledge of which bond wire it was, a person with a steady hand could probably remove that bond wire with a tweezer-like tool. This was possible on the early units in a PGA package.
      - rayiner 14y ago
        Could also be parts where the FPU failed QA but the rest was fine.
    - quux 14y ago
      IIRC the SX's were DX's with a manufacturing defect in the FPU. To increase overall yield of 486 family processors, intel would disable the inoperative FPU by cutting traces with a laser and sell the otherwise perfectly fine chip as an SX. I think, even if you had some kind of super powers that let you reconnect those traces, you still wouldn't have a fully functional DX, unless you also used those super powers to fix the defective FPU.
      - seanmcdirmid 14y ago
        This was never really verified. I'm sure after the 486s were fabbed for a while, intel's defect rate got low enough that they started turning good ones into SX's. By the time the P5 came along, they stopped playing these games all together and started rating by clock speed alone.
        kbenson 14y ago
        > By the time the P5 came along, they stopped playing these games all together and started rating by clock speed alone. Weren't the original Celerons P-II chips with half the L2 cache disabled or something? I also seem to remember they were arbitrarily limited to uni-processor functionality, and some modders found you could drill a portion of the chip to make them work on dual and quad CPU mainboards. If that's correct, I think it's overly generous to say they stopped playing these games (although maybe they took a small break).
        bbatchelder 14y ago
        In the late nineties I had a "dual-celly" workstation. They were all the rage then (especially if you visited Hard OCP and/or arstechnica back then). IIRC, there was a special adapter that sat between the CPU and the socket which did something to enable SMP. I also recall some folks doing something with graphite to re-enable disabled traces. Can't remember if it was related to Celerons or not. Similar examples would be the Promise Ultra66 -> UltraRAID conversion with a simple resistor and flashing of firmware. Of course that game is still being played to this day with graphics cards.
  - klinquist 14y ago
    A better analogy is the early US Robotics Sportster 14.4kbps modems that could be turned into a 16.8 HST with an init string.
- goatforce5 14y ago
  How about IBM's "Golden Screwdriver"? http://www.sdsusa.com/dictionary/index.htm?g#golden http://www.sdsusa.com/dictionary/index.htm?g#golden
dvanduzer 14y ago
I'm not sure there's a big ROI for Tesla to develop elaborate countermeasures. 4,750 units of the 60kWh car at $62.4k yields $296.4m in revenue. If 4% of the orders are for the 40kWh car (190 units at $52.4k), there'd be at most $1.9m in lost revenue by "giving away" the slightly beefier battery. It's plausible they'd still come out ahead given the cost savings of not setting up that production line even if they gave away all 190 software upgrades for free.
- batgaijin 14y ago
  Uh... I don't really get this. Why not lower the price for everyone and get a better rep among customers? I mean if Tesla wants to have lifelong customers that's where you start. But then again they seem to be going into the marketing of their tech so maybe not.
  - eru 14y ago
    Wouldn't that be way more expensive?
  - qb45 14y ago
    Because this is the price required to make the profit they wanted to make on the 60kWh option. The fact that some cheap low performance version turned out to be infeasible because customers prefer expensive high performance version doesn't mean that you should (or could) reduce price of the expensive version. Furthermore, some owners of crippled 60kWh units will probably end up paying for "upgrade" to 60kWh. In the end these units will sell for full 60kWh price even though the buyers initially wanted 45kWh version.
    - makomk 14y ago
      To be honest, I've always suspected that the 40 kWh was more about marketing than about actually releasing a commercially viable option - they wanted to hit the price point Elon Musk had announced on his roadmap several years earlier, but they couldn't produce a product the market actually wanted at such a low price. It might even be a slight loss leader.
skriticos2 14y ago
You want to hack the software of the car that can potentially kill you and others when the controls don't work as expected. Really? Let's see why I think that's not an option: * you loose warranty (on $50k, that's big) * you get liable for any damage your car does because of software failure no matter if it's because of your modifications or not * you can get injured or killed because the car is not obeying your command This is not a computer game we are talking about. I wouldn't apply the same metrics.
- qb45 14y ago
  People already mod cars. Many of these mods are significantly more complicated than changing one number in some EEPROM and yet almost all accidents are still caused by driver error. There is a very good incentive to not screw up your car.
- lawnchair_larry 14y ago
  None of your points are really correct. Warranty is only void for specific repairs determined to be directly caused by the modification. You are never liable for anything that you didn't cause. It's not as if warranty and liability are a binary value that changes if any part of the car becomes tainted. We've been modifying cars long before computer games, and in far more intrusive ways.
- jotux 14y ago
  So should legislation be introduced to lock the hood of current conventional vehicles to keep people from making modifications to them? Why are software modifications on an EV any more dangerous than software and/or mechanical modifications on a conventional vehicle?
- jlgreco 14y ago
  Regardless of how smart of a thing it is to do, many people hack their cars today. There is even a very sizable industry in place to support it.
- jacquesm 14y ago
  I've restored an old car from the ground up. That was lots of fun. One day someone will do it to a model 'S'. Working on old vehicles is interesting because you get to be creative because of a lack of manufacturer supplied parts. Welcome to the aftermarket, where nothing is too crazy. Want to slap a 160 HP engine into your classic mini? That's possible: http://jamesfawcett.co.uk/cms/index.php?option=com_content&view=category&layout=blog&id=7&Itemid=11 http://jamesfawcett.co.uk/cms/index.php?option=com_content&#... Then you go through road approval just like everybody else and you insure it. One day every car driven today will be out of warranty, you are always liable for your vehicle, that's why you insure it and you make sure you pass the road approval tests, you can always be injured or killed when you are driving, software or hardware doesn't matter much. Forgot the safety on one of your steering rods? Too bad! This is 'hacker news', modifying stuff and finding the edges of what you can do with ordinary household items (such as cars) is one of the things we do here. The fact that life is risky is fairly well known but as far as I know it hasn't stopped anybody from sticking their fingers in places where they don't belong.
michael_miller 14y ago
It will be interesting to see if anyone is successful hacking the Model S. I assume that the software is signed with Tesla's private key, and verified by the car. If someone finds a way around this without requiring local access to the car, an attacker could cause a lot of damage. I shudder to think what a BackOrfice for Tesla would look like.
SeanLuke 14y ago
Though perhaps it was not Tesla's intent, this might be a feature, not a bug. Some cars (notably the Volt) will not charge an Li-Ion battery beyond about 85% nor allow it to drop below 15%. As a result the battery has an effective capacity of about 70% its full capacity. As I understand it the reason for this is battery life: these battery chemistries last much longer when only in a partially charged state. I think that Tesla is pushing this: a quick Google suggests they may be charging up to 95% and maybe dropping to 5% in their standard packs. 40kWH usable charge on a 60kWh pack brings this down to maybe 65%, which should last a good long time.
- timdorr 14y ago
  The Model S has two charge modes: standard and max range. Standard is the default and charges to 90% of capacity (approximately 240 miles in rated range). Max Range goes to 100%, but is discouraged for battery health reasons and has to be set manually each time you charge. It looks like Tesla is doing things similar to the Volt, but gives you the option to override.
  - SeanLuke 14y ago
    It always seemed to me that the Model S's two settings might be better described as "pushing it" and "really pushing it". Chevy's maximum state of charge is 80%. The Model S's 90% is pretty surprisingly high, but of course the Model S battery chemistry (Lithium Nickel) is different from the lower-charge one in the Volt (Lithium ion Phosphate).
mncolinlee 14y ago
All of this complaining about the difference between 40 kWh and 60 kWh ignores the fact that most people rarely take trips of that length anyway. I've always had to explain this to people who do not understand why range anxiety is not very important. If I need to take a long road trip, I will rent a car. The 2013 Nissan Leaf is the second cheapest car in America after five years of ownership and it has a great feature set that similar price point cars lack. If a Tesla owner needs to rent a car for a road trip off the beaten path, she can likely afford it. If a Nissan Leaf owner needs to rent a car, he's saving enough money on his purchase to rent a car several times per year and still pay less than someone who bought a similarly-equipped, new car. Rising gas prices will only improve the buying thesis for electric cars. The Bakken reserves are gushing oil yet prices on the world market have only increased.
- johnward 14y ago
  Some people do travel that far though. I travel at least 120 miles daily. I'm probably not the target market for these vehicles yet.
  - mixmastamyk 14y ago
    "most people rarely..."
    - johnward 14y ago
      "Some people do " I guess it depends on the area. Pretty much everyone in my family drives at least 40 miles one way.
- mncolinlee 14y ago
  Okay, that was really a COMPLETELY ridiculous downvote. It's almost like people are assuming someone who owns an $80,000 car doesn't choose to fly anywhere. No number of fools moaning will change the fact that someone who has earned $80,000 can spend it however they choose, including purchasing an energy-efficient automobile that won't go over 200 miles in cold weather without stopping to charge. Is that really any worse than buying a $21,000 Linux rifle to shoot deer at twenty yards?

> The customers who ordered this option will instead receive the 60 kWh pack, but range will be software limited to 40 kWh. can't see anybody hacking that.

by nikcub 14y ago

> The customers who ordered this option will instead receive the 60 kWh pack, but range will be software limited to 40 kWh. can't see anybody hacking that.

seanmcdirmid 14y ago

> ... and can be upgraded to the range of the 60 kWh upon request by the original or a future owner. Does anyone remember Intel's co-processor "upgrade" for turning a 486SX into a 486DX?

kragen 14y ago
Yes. It was called the "487SX", and was billed as a floating-point coprocessor, but it was actually a 486DX with a lower pricetag and a different pinout. When you installed it, the motherboard disabled your 486SX and just used the "coprocessor" for everything. What's the relationship to Tesla's crippleware strategy here? I'm not seeing it.
- seanmcdirmid 14y ago
  Oh, it was worse than that: the 486SXs themselves were actually 486DXs with a disabled FPU. So not only were you buying a 486DX to "upgrade" your CPU, it was basically the same CPU!
  - jacquesm 14y ago
    'Disabled' sounds like there might have been some way to re-enable them, but iirc there was a whole bunch of traces that were zapped with a laser, making the disabling a rather permanent affair.
    - seanmcdirmid 14y ago
      Rumor has it there were a few people who were good enough with a soldering iron to re-enable them.
      - brk 14y ago
        Huh? The traces were buried inside the ceramic chipset. And even at the scale of technology in that era the soldering was usually done with vibration, not heat. I find this statement rather implausible. It would take far too much time and equipment to make it worthwhile vs. just upgrading your CPU "over the counter".
        jacquesm 14y ago
        I think that was a joke and I think it went over your head :) I do know a guy that can solder scary small stuff (SMT, angel hair and other incredible feats of steady handedness) but I think that re-bonding zapped chips is even beyond his soldering capabilities. If only because you first have to get access. I did once re-solder a pin on a 386 which had broken off and I was pretty proud of that but for the most part once you get to the package boundary it is game over.
        brk 14y ago
        "I think that was a joke and I think it went over your head :)" As had I. My first computer was a PCjr., with a hand-soldered memory expansion card. I repaired broken pins on processors multiple times (maybe like 5 or 6), since in that era I was also building "custom" machines for people. But in regards to this being a "joke", it just seemed like really bad folklore to me. Having lived through that era, I just really didn't see even any remote humor in the comment.
        seanmcdirmid 14y ago
        You guys are young! To the downvoter, this was a very common claim as far as computer folklore goes: http://en.wikipedia.org/wiki/Talk%3AIntel_80486SX http://en.wikipedia.org/wiki/Talk%3AIntel_80486SX > The joke about the drill through the precisely correct location is likely a joke made by someone who knew what they were talking about. It was likely not meant to be taken literally, but sort of a tease without explanation. The floating point unit was in fact never removed with a laser, it was disabled by tying a "disable floating point" input pad to power or ground using a bond wire in the package. This joke was probably a reference to the fact that if you disconnected that bond wire, the default state of that input pad was to enable the FPU. So I think the idea behind the joke was that you could disconnect that bond wire by drilling through the bond wire and thereby disconnecting it. The practical way to do this would be to flip off the lid on the PGA package to expose the die and bond wires. People did have a steady enough hand to straighten bond wires on occasion when that were shorting to adjacent bond wires. With a knowledge of which bond wire it was, a person with a steady hand could probably remove that bond wire with a tweezer-like tool. This was possible on the early units in a PGA package.
    - rayiner 14y ago
      Could also be parts where the FPU failed QA but the rest was fine.
  - quux 14y ago
    IIRC the SX's were DX's with a manufacturing defect in the FPU. To increase overall yield of 486 family processors, intel would disable the inoperative FPU by cutting traces with a laser and sell the otherwise perfectly fine chip as an SX. I think, even if you had some kind of super powers that let you reconnect those traces, you still wouldn't have a fully functional DX, unless you also used those super powers to fix the defective FPU.
    - seanmcdirmid 14y ago
      This was never really verified. I'm sure after the 486s were fabbed for a while, intel's defect rate got low enough that they started turning good ones into SX's. By the time the P5 came along, they stopped playing these games all together and started rating by clock speed alone.
      - kbenson 14y ago
        > By the time the P5 came along, they stopped playing these games all together and started rating by clock speed alone. Weren't the original Celerons P-II chips with half the L2 cache disabled or something? I also seem to remember they were arbitrarily limited to uni-processor functionality, and some modders found you could drill a portion of the chip to make them work on dual and quad CPU mainboards. If that's correct, I think it's overly generous to say they stopped playing these games (although maybe they took a small break).
        bbatchelder 14y ago
        In the late nineties I had a "dual-celly" workstation. They were all the rage then (especially if you visited Hard OCP and/or arstechnica back then). IIRC, there was a special adapter that sat between the CPU and the socket which did something to enable SMP. I also recall some folks doing something with graphite to re-enable disabled traces. Can't remember if it was related to Celerons or not. Similar examples would be the Promise Ultra66 -> UltraRAID conversion with a simple resistor and flashing of firmware. Of course that game is still being played to this day with graphics cards.
- klinquist 14y ago
  A better analogy is the early US Robotics Sportster 14.4kbps modems that could be turned into a 16.8 HST with an init string.
goatforce5 14y ago
How about IBM's "Golden Screwdriver"? http://www.sdsusa.com/dictionary/index.htm?g#golden http://www.sdsusa.com/dictionary/index.htm?g#golden

dvanduzer 14y ago

I'm not sure there's a big ROI for Tesla to develop elaborate countermeasures. 4,750 units of the 60kWh car at $62.4k yields $296.4m in revenue. If 4% of the orders are for the 40kWh car (190 units at $52.4k), there'd be at most $1.9m in lost revenue by "giving away" the slightly beefier battery. It's plausible they'd still come out ahead given the cost savings of not setting up that production line even if they gave away all 190 software upgrades for free.

batgaijin 14y ago
Uh... I don't really get this. Why not lower the price for everyone and get a better rep among customers? I mean if Tesla wants to have lifelong customers that's where you start. But then again they seem to be going into the marketing of their tech so maybe not.
- eru 14y ago
  Wouldn't that be way more expensive?
- qb45 14y ago
  Because this is the price required to make the profit they wanted to make on the 60kWh option. The fact that some cheap low performance version turned out to be infeasible because customers prefer expensive high performance version doesn't mean that you should (or could) reduce price of the expensive version. Furthermore, some owners of crippled 60kWh units will probably end up paying for "upgrade" to 60kWh. In the end these units will sell for full 60kWh price even though the buyers initially wanted 45kWh version.
  - makomk 14y ago
    To be honest, I've always suspected that the 40 kWh was more about marketing than about actually releasing a commercially viable option - they wanted to hit the price point Elon Musk had announced on his roadmap several years earlier, but they couldn't produce a product the market actually wanted at such a low price. It might even be a slight loss leader.

skriticos2 14y ago

You want to hack the software of the car that can potentially kill you and others when the controls don't work as expected. Really? Let's see why I think that's not an option: * you loose warranty (on $50k, that's big) * you get liable for any damage your car does because of software failure no matter if it's because of your modifications or not * you can get injured or killed because the car is not obeying your command This is not a computer game we are talking about. I wouldn't apply the same metrics.

qb45 14y ago
People already mod cars. Many of these mods are significantly more complicated than changing one number in some EEPROM and yet almost all accidents are still caused by driver error. There is a very good incentive to not screw up your car.
lawnchair_larry 14y ago
None of your points are really correct. Warranty is only void for specific repairs determined to be directly caused by the modification. You are never liable for anything that you didn't cause. It's not as if warranty and liability are a binary value that changes if any part of the car becomes tainted. We've been modifying cars long before computer games, and in far more intrusive ways.
jotux 14y ago
So should legislation be introduced to lock the hood of current conventional vehicles to keep people from making modifications to them? Why are software modifications on an EV any more dangerous than software and/or mechanical modifications on a conventional vehicle?
jlgreco 14y ago
Regardless of how smart of a thing it is to do, many people hack their cars today. There is even a very sizable industry in place to support it.
jacquesm 14y ago
I've restored an old car from the ground up. That was lots of fun. One day someone will do it to a model 'S'. Working on old vehicles is interesting because you get to be creative because of a lack of manufacturer supplied parts. Welcome to the aftermarket, where nothing is too crazy. Want to slap a 160 HP engine into your classic mini? That's possible: http://jamesfawcett.co.uk/cms/index.php?option=com_content&view=category&layout=blog&id=7&Itemid=11 http://jamesfawcett.co.uk/cms/index.php?option=com_content&#... Then you go through road approval just like everybody else and you insure it. One day every car driven today will be out of warranty, you are always liable for your vehicle, that's why you insure it and you make sure you pass the road approval tests, you can always be injured or killed when you are driving, software or hardware doesn't matter much. Forgot the safety on one of your steering rods? Too bad! This is 'hacker news', modifying stuff and finding the edges of what you can do with ordinary household items (such as cars) is one of the things we do here. The fact that life is risky is fairly well known but as far as I know it hasn't stopped anybody from sticking their fingers in places where they don't belong.

michael_miller 14y ago

It will be interesting to see if anyone is successful hacking the Model S. I assume that the software is signed with Tesla's private key, and verified by the car. If someone finds a way around this without requiring local access to the car, an attacker could cause a lot of damage. I shudder to think what a BackOrfice for Tesla would look like.

SeanLuke 14y ago

Though perhaps it was not Tesla's intent, this might be a feature, not a bug. Some cars (notably the Volt) will not charge an Li-Ion battery beyond about 85% nor allow it to drop below 15%. As a result the battery has an effective capacity of about 70% its full capacity. As I understand it the reason for this is battery life: these battery chemistries last much longer when only in a partially charged state. I think that Tesla is pushing this: a quick Google suggests they may be charging up to 95% and maybe dropping to 5% in their standard packs. 40kWH usable charge on a 60kWh pack brings this down to maybe 65%, which should last a good long time.

timdorr 14y ago
The Model S has two charge modes: standard and max range. Standard is the default and charges to 90% of capacity (approximately 240 miles in rated range). Max Range goes to 100%, but is discouraged for battery health reasons and has to be set manually each time you charge. It looks like Tesla is doing things similar to the Volt, but gives you the option to override.
- SeanLuke 14y ago
  It always seemed to me that the Model S's two settings might be better described as "pushing it" and "really pushing it". Chevy's maximum state of charge is 80%. The Model S's 90% is pretty surprisingly high, but of course the Model S battery chemistry (Lithium Nickel) is different from the lower-charge one in the Volt (Lithium ion Phosphate).

mncolinlee 14y ago

All of this complaining about the difference between 40 kWh and 60 kWh ignores the fact that most people rarely take trips of that length anyway. I've always had to explain this to people who do not understand why range anxiety is not very important. If I need to take a long road trip, I will rent a car. The 2013 Nissan Leaf is the second cheapest car in America after five years of ownership and it has a great feature set that similar price point cars lack. If a Tesla owner needs to rent a car for a road trip off the beaten path, she can likely afford it. If a Nissan Leaf owner needs to rent a car, he's saving enough money on his purchase to rent a car several times per year and still pay less than someone who bought a similarly-equipped, new car. Rising gas prices will only improve the buying thesis for electric cars. The Bakken reserves are gushing oil yet prices on the world market have only increased.

johnward 14y ago
Some people do travel that far though. I travel at least 120 miles daily. I'm probably not the target market for these vehicles yet.
- mixmastamyk 14y ago
  "most people rarely..."
  - johnward 14y ago
    "Some people do " I guess it depends on the area. Pretty much everyone in my family drives at least 40 miles one way.
mncolinlee 14y ago
Okay, that was really a COMPLETELY ridiculous downvote. It's almost like people are assuming someone who owns an $80,000 car doesn't choose to fly anywhere. No number of fools moaning will change the fact that someone who has earned $80,000 can spend it however they choose, including purchasing an energy-efficient automobile that won't go over 200 miles in cold weather without stopping to charge. Is that really any worse than buying a $21,000 Linux rifle to shoot deer at twenty yards?