Hacker News Reader

Comment by pizza234 | Hacker News Reader

Very interesting bit about power draw:

> The 100-Watt AC adapter included with the Surface Book 2 is not quite powerful enough to keep the system fully charged under a high load such as this, so you can see there is some power draw from the battery (the green line in the chart) which is leveraged to make up the difference. Over time, the battery drops to about 95% charge, at which time the system drops the GPU clock a bit to prevent draining the battery any further. Not all games are going to run into this issue, but certain games will require more power from the GPU and CPU than the power input can handle. However, the battery will never completely be drained, as the system doesn’t let it get out of control.

Dishonest designs like this should be punished by consumers, in particular, considering that the Surface Book tablet battery is already a critical part of the laptop (it's comparatively small, and it's hard to replace).

View on HN

Many, if not most, desktop-replacement laptops with 4+ cores are thermally limited on compute, and will throttle back under extended high load. I see it as a compromise of the form factor rather than dishonesty per se.

This isn't a case of thermal limiting - this is a case of the power supply being literally unable to supply enough power to run the system at its full capacity. If I can't use that 1060 at full power for more than 5% of the battery capacity, even while plugged into the wall, isn't that misleading?

The limiting factor might be different but the effect is the same. A rose by any other name would leave you just as disappointed.

(I'm not certain that the limiting factor is actually different, FWIW. Dissipating 100W from a laptop isn't easy. Macbook Pros limit at 80W AFAIK. If the power supply could keep up, I'd expect it to get too hot anyway.)

Dissipating lots of heat from a laptop isn't that hard, it's just that the tradeoffs take the form of gaming laptops because the laws of thermodynamics are not shaped by consumer preferences.

Yep, my 2016 15" MBP (2.9 GHz version) will lock the CPU clock down to 1.2 GHz after about half a minute of running some compilation on all cores. It makes for a staggerng difference in comparison to even old i5 2500K desktop CPU, which can easily outperform the laptop.

The main problem is not the thermal limiting - it's the tablet battery draining while being connected to the AC, which is counterintuitive and undesirable.

> Dishonest designs like this should be punished by consumers

This needs to go one step further because consumers don't generally care about the design of the device until something goes wrong, at which point they get burned and lose their investment. Retailers then misinform them of their repair or replacement rights intentionally. Either the retailer should warranty it for the reasonable life of the device which has to be quoted on the front of the packaging with no questions asked or they have to meet certain repair criteria with a full refund no questions asked if they can't repair it under a certain amount of time (critically a short amount of time like 2 weeks)

Consumer Rights Act in the UK is the nearest thing to this and gives you a position to argue for fitness for purpose over the span of 6 years from purchase. But it doesn't go far enough.

With the above, the Surface Book wouldn't even make it to market which is how it should be. It's disposable and nothing in that price range should be disposable.

>This needs to go one step further because consumers don't generally care about the design of the device until something goes wrong, at which point they get burned and lose their investment.

I literally got burned once when a badly designed laptop overheated (the plastic around the fan literally melted off), and I'm sure I'm not the only one to whom something like that has happened.

The whole "Apple slows down old iPhones" silliness now seems a bit different in retrospective :) :) We have a laptop that is not fully charged when... plugged in!

As far as I know, no battery, including iPhone’s, can achieve 100% charged. The concept of 100% charged is a software concept. It is hard to measure when exactly is 100%.

iPhone’s 100% is roughly around 97-98% of its measurement by the hardware. So why is that dishonest?

Am I confused here?

Well in this case, the 95% that the review mentions is what the laptop software (e.g. Windows) reports. So even if 100% is never achieved in all devices, here we're talking about not even the "perceived" 100% is achieved.

Macbooks used to do the same. Someone here remarked about this behavior.

I've been using MBPs daily for over ten years and haven't noticed anything like that </anecdote>

Is artificially lowering down the performance of iPhones to keep the battery for a day the same thing?

The technical mechanics are similar, but that seems to be the extent of it.

Apple (silently) changed the "failure mode" of batteries from reducing autonomy to lowering performances. Ars shows their power mode as "Recommended", I would guess the observation is a side-effect of that, with the laptop switching from "charged" to "charging" power mode (lowering components performances to charge faster). I saw no note about that, I'd like a re-do with a laptop configured for "best performances".

Of course Microsoft kinda shot themselves in the foot by limiting power input to 100W and designing a laptop with more peak draw than this.

No; in the SB case, tablet battery is being drained while connected to the AC. This is counterintuitive and undesirable (imagine you have a gaming session, then commute).

Reply on news.ycombinator.com