Hacker News new | ask | show | jobs
by DanielStraight 3667 days ago
Maybe this is just pointless pedantry but...

This isn't a software upgrade. It's an on/off switch.

There's a difference. A software upgrade implies some new code which cost real money to develop is being made available. That's not the case here.

As others mentioned, manufacturers do this already for cost reasons (reducing variation) on other software, but in that case it's legitimately a software upgrade because the new software being turned on cost actual money to develop which they recoup by putting it behind a paywall.

In this case, it probably actually cost Tesla money to implement the software to cap the battery.

What we have here is software being used to implement a paywall on hardware to recoup hardware development costs (presumably developing the 75 battery cost more than developing a 60 would have). It's a software-controlled hardware upgrade which happens to not actually be implemented as an upgrade because of cost control reasons.

I don't know that there is precedent for this precise business model.

4 comments

Lots of people do this, it's often more expensive to make different versions of things then is it just to software cap something. So they make the same version for multiple price points and have variations of software cap in place.

CPUs for example are clocked at a set speed when they reach the consumer, they can always go higher without any negative impact (infact most are underclocked). It's basically a software switch that turns your 3.2 Ghz CPU into a 3.4 Ghz CPU.

You buy a SAAS product they have multiple levels, it's just a software switch that says you can X package or Y package. This is slightly different because the features may be more resource intensive or something else, but basically it's the same.

Sure there is, nVidia and ATI/AMD have done pretty much exactly this sort of segmentation of many consumer->pro video cards in drivers practically forever.
I think that with graphic cards it is a little different: the manufacturing process is simply not perfect, so attempt to produce card with 56 Streaming Processors (Pascal) can result in 40 SP units working fine and 16 being broken. So you disable broken units and sell the piece as lower end card instead.
They also do it on cards that pass the higher-end binning process fine and sell them downmarket anyway when there is more demand there.
And IBM has done this for a long time with its mainframes & zSeries. You can pay a fee to switch additional processors on.
You can even do it temporarily, which is neat (year-end reports or data crunching, for instance).
Many SSDs are built with extra, inaccessible, capacity, for wear leveling purposes (of course, I don't know of any SSD manufacturers selling unlock codes for the extra storage). Lithium ion batteries similarly degrade with use. Perhaps an unlocked S60's battery will age better, which would be better for buyers in the long run. Tesla also sells preowned cars, so if extra capacity helps with wear, this would also help on that front (and introduce the possibility of selling the unlock twice?).
Carrier-grade microwave radios are usually sold at a base price for a specific speed (e.g. 100 Mbps). Software licenses can then be purchased to upgrade the capacity (e.g. to 200 Mbps, 500 Mbps, or 1 Gbps). Licenses are also sold to enable features such as encryption and hitless backup radio failover.
That's a little different because those extra features required additional work to implement in software and just happen to run on the same base hardware. This is just hardware being artificially constrained. It's like buying a 6 pack of Coke and actually getting 8 cans but you can only drink 6 of them unless pay an extra $3 to "unlock" the two cans already in the box.