How the Xbox’s default “instant on” feature could harm the environment

Get a Kill-A-Watt or similar device and go around your house, apartment, hole in the ground, van down by the river, whatever and measure your devices' power draw and be amazed / horrified. My burr coffee grinder consumes 0.3 watts if the grounds receptacle is seated, which lights up a bright blue LED. 0 if it's not seated, thus no blue LED.

Times countless scores of people and countless devices designed like this... it adds up real quick!

It really doesn't add up that quick compared to other sources of power draw, and every single Kill-A-Watt device that you buy has a carbon footprint of its own.

At one point I did the calculations on specifying combination USB outlets for a new university building, as I had to choose between the kind that would be having some minor phantom power draw the entire time, vs the kind with a spring loaded cover that switches them on. The entire phantom power draw of all several hundred outlets worked out to being less than a single LED potlight, which on its own is orders of magnitude less than the power draw of any of the heaters in the building (or you know the impact of someone leaving a window open when they shouldn't). When I talked it over with our environmental team, they all agreed that given the additional moving parts and complexity of the spring loaded cover plates, it almost certainly would be, at best, a net wash from an environmental perspective once you accounted for the additional manufacturing carbon footprint, as well as the impact of replacing the moving parts when they inevitably break.

Yes, we need to reduce our energy consumption but stuff like this feels like they're wasting a lot of effort to not make much of a difference. If we were just talking about waiting a couple extra seconds to boot your console up to game that one be one thing, but that point is kind of disingenuous since the real benefit of instant on is not having to wait several hours for the new several GB game update to download. That's the difference between being able to spend a night gaming and not.

A usb charger that's plugged in but not charging anything, typically consumes about 0.01-0.1 W. That's two to three orders of magnitude less than the Xbox instant-on feature. You can't really compare their impact one-on-one.

You won't save the planet by frantically pulling out your phone charger, but a kill-a-watt cán help you to identify stuff that does have a significant standby usage, like for example the Xbox, but settopboxes for cable or iptv are also notorious energy drains. And the fact millions of people use these devices often without having a clue, makes it add up.
 
Upvote
2 (8 / -6)

Ruefus

Ars Tribunus Militum
1,511
Bear with me a moment.

Given the rise in popularity of wireless charging and the inherent wastefulness of the technology, this article seems hypocritical.

Who here hates this feature with Xbox, but wirelessly charges their device(s)?

The reaction I’m guessing I’ll get, in addition to downvotes, is that the power loss is micro in comparison. It doesn’t count.

So where’s the line?

If you’re invested in conservation and environmental concerns, but waste energy purely for convenience, complaining about this is hypocritical.

I’m not suggesting you can’t use a technology or feature. I’m saying the concept of “do as I say, not as I do.....because mine is a small thing and so convenient. But yours is so *bad* ”........ undermines your complaint.
 
Upvote
6 (11 / -5)

SmileyBarry

Smack-Fu Master, in training
55
Devices like computers tend to use /the most/ power when booting up. Boot up with any frequency and you loose all benefits of being off in the first place.

Citation needed. I could be wrong, but I find that to be implausible with modern computers that can rapidly switch power states within milliseconds.
Most computers boot with their CPUs running at their fastest non-TurboBoost (or similar) speed. Once the OS assumes control, it will then enable downclocking to save power.

True for other components as well-- spinning rust drives need the most power during spinup; and most desktop computers spin up their fans to full speed at power-on or hard reboot, and then let them slow down once the BIOS/EFI is sure they are rotating or after the OS or a fan management utility takes over.

It's easily visible with a Kill-A-Watt, although there is a huge difference between the peak boot load and the desktop standby load seen on a gaming PC, an HTC/router PC, a Mac Mini, or a laptop.

Most of that isn't correct anymore. Fans don't spin up to max on boot. At most they start at the 50% mark, but usually at the minimum 30% mark (my own 2018 PC does this).

It's true that hard drive spin up takes the most energy but that's around 9W-10W peak on a performance 3.5" hard drive (WD Black specifically), less on average hard drives. On 2.5" hard drives you're looking at around 2-3W.

CPUs started regulating their own power consumption and automatically downclock based on use, but even at max clock they can sleep during cycles and consume drastically less watts. At high-perf power mode my 9900k consumes ~13W to idle at 0%-3%. This kicks in much sooner at the UEFI phase.
 
Upvote
5 (5 / 0)

SmileyBarry

Smack-Fu Master, in training
55
Somewhat off topic.

I bought a dishwasher 5 years ago but realized too late I had no use for it at the time, so I left it disconnected thinking I would save power. When I recently plugged it in the programs didn’t work correctly, as if the memory had been partially lost. I’ve read flash needs to be plugged in or this could happen.

I had to recycle it since it was cheaper to buy a new one than to repair it. A hefty carbon footprint considering it was new when I bought it and I never used it.

That's not related, rewriteable flash memory needs to be plugged in about once a year to reliably retain info but even then you might lose bits, not a whole chunk of memory. And appliances have read-only memory where this doesn't apply -- consider that it can just sit at a storage facility for years, unsold, and still work when it does sell.
 
Upvote
3 (3 / 0)

bluebonics

Smack-Fu Master, in training
54
Well, that’s a nice tangible way of tallying up consumption of electronics; in a very narrow and sensationalist headline.

Forrest for the trees. How about converting street lamps to LED first? Knocking off coal fired plants. And while we’re at it, let’s not sink server rads in the sea simply because it provides more energy “cost efficient” cooling for IT data centres.

https://home.uni-leipzig.de/energy/ener ... als/04.htm

58kWh is the human average. Per day!

There's no reason it has to be "this or that". Why do those things first when we can do them concurrently?

Sure there is. You're missing the proportionality issue here. Something like this is a negligible thing compared to major polluters. Stopping this has no actual effect on pollution and climate change because industry pollutes on a scale that is orders of magnitude worse.

It's like chastising someone who's whispering to stop making so much noise while you let someone else sounding multiple, deafening bullhorns continue on. And then you sit there and say "it's not this or that." You're wrong, it's precisely that. And things like this only serve to act as a false equivalence to our actual pollution problems.
 
Upvote
13 (15 / -2)

Eraserhead

Ars Praefectus
4,565
Subscriptor++
Well, that’s a nice tangible way of tallying up consumption of electronics; in a very narrow and sensationalist headline.

Forrest for the trees. How about converting street lamps to LED first? Knocking off coal fired plants. And while we’re at it, let’s not sink server rads in the sea simply because it provides more energy “cost efficient” cooling for IT data centres.

https://home.uni-leipzig.de/energy/ener ... als/04.htm

58kWh is the human average. Per day!

There's no reason it has to be "this or that". Why do those things first when we can do them concurrently?

Sure there is. You're missing the proportionality issue here. Something like this is a negligible thing compared to major polluters. Stopping this has no actual effect on pollution and climate change because industry pollutes on a scale that is orders of magnitude worse.

It's like chastising someone who's whispering to stop making so much noise while you let someone else sounding multiple, deafening bullhorns continue on. And then you sit there and say "it's not this or that." You're wrong, it's precisely that. And things like this only serve to act as a false equivalence to our actual pollution problems.

Incremental change is how we solve problems.

Besides the energy usage of “instant on” at 15W continually is more than half of the energy usage of a tall 60cm wide own brand freezer. Or more than half of a washing machine, tumble drier or dishwasher.
 
Upvote
-8 (4 / -12)
To put that 78 kWH into perspective, the average US household consumes a little over 10 MWH per year.

https://www.eia.gov/tools/faqs/faq.php?id=97&t=3

To put it into perspective, 500MW is the output power of half a nuclear reactor, or the full output of a mid-size coal power plant.

Oh, and multiply by three or four in the summer, when you use the AC to pump out the heat.
 
Upvote
-5 (0 / -5)

cerberusTI

Ars Tribunus Angusticlavius
7,199
Subscriptor++
Semi-related: I found out this week thanks to this video that my El Cheapo Vick's humidifier can draw 200W at full tilt, which was a giant surprise to me. I bought a whole-home model last week anyway, and it covers more area while using less power, and the energy savings will pay for itself in a couple months.

If you watch his connextras follow up video, he addresses the fact that if one is using resistive electric heating and only using the humidifier in the winter, using the el cheapo Vick's one is the most efficient since the heat generated by it is heat that doesn't have to be generated by electric heating.

If you use an electric heat pump, then the vick's heat based humidifier will definitely result in more power use though since heat pumps are so much more efficient than resistive electrical heating.

The calculation may change if you have gas heating since it tends to be cheaper than resistive electric heating in most places.

https://www.youtube.com/watch?v=HfFAiCMLJ14
Gas is far cheaper for heat in most places, something like an order of magnitude for the fuel itself where I am (much of the cost is a fixed fee). Energy use is not the only consideration for a humidifier though.

All humidifiers have an issue in that there is stuff in the water which you do not want in the air, and this problem can be substantially worse if there are things growing in the tank. The primary method by which a humidifier fails is by contaminating the air and creating a problem which is worse than the one it solves.

Ultrasonic humidifiers are efficient at getting water into the air, but they also readily put everything else in that tank into the air, and dispersion leaves something to be desired so you need several with spacing for a larger area (or a lot of airflow). You cannot treat the water as it will dump whatever you treat it with into the air along with the water, so these require timely and detailed cleaning or they become a hazard. Unless you have very good water, they also create dust, which is more or less of a problem depending upon how much dust, and what is in that dust. These are better with distilled water, but the efficiency is gone if you need to distill water first. At least some filtering is advisable, preferably something that removes dissolved solids, which tends to mean nontrivial cost.

There are humidifiers which evaporate water with a fan and surface area, but that is an ideal environment for growth in some ways, so it should really be treated (which will no doubt get into the air a bit), and it will need filters as it will build up mineral crust. I have never owned one.

The steam variety solves this problem with energy use, as in boiling water off anything heavier than water either becomes scale or more concentrated liquid leavings you wash out, rather than getting into the air. It also disperses well, so you can put it anywhere.

What is best will depend upon your use, but I just switched from an ultrasonic to a steam humidifier. We have fairly hard water, which softening by ion exchange would not help for this use case. Last year I did not need one at all, but sometimes we get dry cold air coming in from Canada which lingers, and suddenly the humidity craters. For the realistically zero through two months we have cold dry air, and maybe one or two more with enough to even optionally warrant running one, steam seems to work out better.

The efficient method from my youth is to boil a pot of water all day until the humidity is acceptable, which if done on a gas stove is very cheap. The more expensive one without the open flame and with a humidity sensor involves converting enough energy to electrical resistance heating from gas in order to steam distill then boil off the water (deciding I need a better humidifier in the middle of January meant limited options, so this one ideally wants distilled water as input, which I suppose I am willing to do), and is about $50 extra in electricity per month for the expected duty cycle if it runs all the time and puts a few gallons into the air per day. I expect this not to be necessary very often and therefore much less in real use, so that works. If I end up needing to use it substantially I may buy a better integrated device with a suitable boiling chamber to directly feed it hard water in the summer when they exist again, which would halve that energy use. If it were a super cold area, I might consider a gas water boiler.

Even boiling water once is likely to be quite a bit more expensive than filters and treatment for evaporation if you need a lot of water constantly, but needing it only occasionally in a mostly wet area it is pretty competitive and has several convenient aspects. I mostly like it as it is very fast when you decide you need it, and contains some integrated protection against the most common way these fail in practice.

In the summer I would not run one as there is no need, but steam would likely be a less than ideal way to humidify air in a hot dry area.

The xbox thing is annoying as it seems unnecessarily high, but in the context of a humidifier that energy is going towards relatively high quality electronic water filtering.
 
Upvote
1 (1 / 0)

Chuckstar

Ars Legatus Legionis
37,477
Subscriptor
To put that 78 kWH into perspective, the average US household consumes a little over 10 MWH per year.

https://www.eia.gov/tools/faqs/faq.php?id=97&t=3
To put that 10MWH inperspective, except for a few countries in Europe the average EU household consumes around 4MWH per year.
https://www.odyssee-mure.eu/publication ... lling.html
It’s kind of hard to compare on just gross numbers. Nowhere in the U.S. is as poor as Romania or Poland, for instance. Climate makes a huge difference (Californians aren’t particularly low energy users in other ways, but the climate requires relatively little heating/cooling compared to many others). Also depends on whether electric is used for heating (as opposed to natgas or fuel oil).
 
Upvote
2 (4 / -2)

SplatMan_DK

Ars Tribunus Angusticlavius
8,295
Subscriptor++
Did a PlayStation write this?
Why do you ask?

Do you believe it's a bad thing to criticize a product for certain poor attributes, merely because said product has competition from other vendors?

Or, in simpler terms: Do you think Microsoft and the Xbox should be immune to criticism merely because other console vendors exist?
 
Upvote
-5 (3 / -8)

SplatMan_DK

Ars Tribunus Angusticlavius
8,295
Subscriptor++
The fact that the system takes 45 seconds to startup is the real atrocity here. With the SSD, the Xbox should be able to cold boot in a couple seconds.

I'm surprised it doesn't hibernate like Windows. With a fast SSD, resuming from hibernation takes a few seconds only.
64 GB of RAM says otherwise. :-(

But you're right for the console off course.
 
Upvote
0 (0 / 0)

cwsars

Ars Tribunus Militum
1,509
Subscriptor++
It's easily visible with a Kill-A-Watt, although there is a huge difference between the peak boot load and the desktop standby load seen on a gaming PC, an HTC/router PC, a Mac Mini, or a laptop.
Most of that isn't correct anymore.
Opinions don't supersede data. You can observe the difference in boot power consumption versus desktop idle easily yourself if you use a Kill-A-Watt meter.

Fans don't spin up to max on boot. At most they start at the 50% mark, but usually at the minimum 30% mark (my own 2018 PC does this).
I've got a slightly older Asus motherboard with "advanced" fan control (meaning 4-pin fan headers with PWM support) and zero-RPM support.

Yes, the UEFI will slow down the fans itself before the OS runs. However, the fans still go to 100% briefly after a hard reboot or power cycle. Most fans won't start rotating from idle at 30% power, and some won't start from idle at 50% either.

I suspect your motherboard does exactly the same thing, although it may well lower fan power very quickly once it confirms that the fans have spun up such that it is not as obvious.

It's true that hard drive spin up takes the most energy but that's around 9W-10W peak on a performance 3.5" hard drive (WD Black specifically), less on average hard drives. On 2.5" hard drives you're looking at around 2-3W.
I agree with these numbers. They mean a bigger difference for desktops and a smaller difference for laptops or streaming devices like an AppleTV.

CPUs started regulating their own power consumption and automatically downclock based on use, but even at max clock they can sleep during cycles and consume drastically less watts. At high-perf power mode my 9900k consumes ~13W to idle at 0%-3%. This kicks in much sooner at the UEFI phase.
Your UEFI motherboard almost certainly defaults to booting in max non-TurboBoost CPU speed, just like mine does. The CPU isn't allowed to reduce the clock frequency until OS starts up APM and scribbles to the MSR enabling that capability.

Yes, I could choose to boot my desktop at 800Mhz instead rather than 4GHz, but that just makes the OS bootstrap take longer than the CPU power savings justify.
 
Upvote
-2 (2 / -4)
I primarily heat my home with resistive heating elements, and I need some form of heating 75% of the year, so I think I'll just not worry too much about whatever my devices use in standby mode. It all adds to my total electric heating anyway, at about the same efficiency.

I can see the issue in areas where you have to use AC for a big portion of the year, though. Not only do the devices draw power, but the AC will have to spend extra energy removing that extra energy.
 
Upvote
3 (3 / 0)
there's only maybe two comments that I have identified as being from actual owners

i have the series X myself, and I can tell you there are more than a few reasons for the standby mode. it would be nice if the author would have researched them instead of assuming everything was done for a quick boot and update check. i'm new to the ecosystem myself, only since christmas, but can point out a couple things that wake the system during normal usage

the xbox isn't just standalone with the tv, it's also linked to my android devices, for at least a couple different useful functions. there's probably more that i'm not aware of yet but that's not the point

1)remote play. game streaming to your other devices (beta)
2)remote control of the system. you can actually fully power it on and off remotely, and add/remove apps/games, by using an app on another device, located anywhere. helpful when you have kids and need to do something from work

just saying. there's reasons for this that owners are aware of and appreciate
 
Upvote
9 (9 / 0)
To put that 78 kWH into perspective, the average US household consumes a little over 10 MWH per year.

https://www.eia.gov/tools/faqs/faq.php?id=97&t=3

What's up with the strange units in the title of the article? Why is it referencing billions of kWH instead of MWH or GWH? It's a strange obfuscation.

Well, kWh are potentially more relatable on a personal basis. At the very least, you likely read it in your monthly utility bill and may have more exposure to it if you have an EV or solar installation. Even with a basic scientific understanding you could imagine a 1kW device using power for one hour.

Switch it out for 4 TWh and it becomes much more abstract.
 
Upvote
3 (3 / 0)
"Given those numbers, our hope is that most users would be willing to wait an extra 5 to 10 seconds for their console to restart if they knew the impact," Horowitz writes.

Yeah, that's fine, but it's not just about the extra 5 - 10 seconds. It's also about whatever maintenance the system has to do when it boots up. Game updates couldn't complete because the system was shut down.

It feels like we need a middle ground option here. A choice that allows the system to boot up periodically, check for updates and shut back down if there aren't any.

Im honestly surprised that MS didnt add an ARM SOC just for this scenario.

Or perhaps, a couple of Jaguar cores somewhere for the same reason.

I know, easier said than done, but they knew how “important “ and useful this power mode was going to be and how the world is looking more and more to power consumption.
 
Upvote
1 (1 / 0)

wk_

Ars Tribunus Militum
2,383
The trouble with using power when off is that houses have a lot of idling devices. My now broken audio system (from 2004.) draws 25W when "off" (the official figure from the manual). The system that barely has 50W RMS power. It also has an "eco" mode where it uses 3W. The only effective difference when it is in its default power mode is that it shows the time. From both of the modes it gets running almost instantly, because it's not that it has an entire OS to load or something. That is an obvious example of the lazy engineering. I guess that all circuits are running and that it simply disconnects the speakers.

In my house there is a STB, two TVs, two PCs with monitors, plus ACs that have their remotes. Then the microwave, the bread maker - which I unplug from the socket because it does not have anything resembling a power button... So you can easily count to 10 devices that have to work (at least partially) even when turned off to be able to notice the remote or the "soft" power button. If each of them would use 20W, that would be 4.8kWh per day, 150 per month. I think that it was reasonable that EU stepped up and forces the power limit on idling devices.
 
Upvote
2 (3 / -1)
It's generally agreed now (but go find your own links, seriously they're there) that letting people think they're green is one of the worst things we can do as a society.

It's supposedly (plausible, but what else, how would you confirm?) the same bias as "I've dieted for a bit, reward time" and such

This is separate from just not knowing, like buying a glass bottle to replace disposable plastic bottles and cups (but thinking that does good is the danger I'm alluding to)

We'll all have countless examples of this, I'm (quasi-)aware of this, and I still probably suck, but I try.

There was also this dipshit I knew in university, maths student so fellow in the pursuit of hardcore logic blindsided me with crazy levels of religion.

All my years of turning off lights has been negated by his this:

"WTF, you leave the oven on while you go to lectures?"

->

"Yeah that way it's pre-heated when I come home"

->

"Do you have any idea how bad that is? The university has a diesel plant!"

->

"Don't worry, God will provide for man for as long as he is upon the Earth!"

That's when I found out. We have no fucking hope do we?
 
Upvote
1 (3 / -2)

SraCet

Ars Legatus Legionis
17,220
Bear with me a moment.

Given the rise in popularity of wireless charging and the inherent wastefulness of the technology, this article seems hypocritical.

Who here hates this feature with Xbox, but wirelessly charges their device(s)?

The reaction I’m guessing I’ll get, in addition to downvotes, is that the power loss is micro in comparison. It doesn’t count.

So where’s the line?

If you’re invested in conservation and environmental concerns, but waste energy purely for convenience, complaining about this is hypocritical.

I’m not suggesting you can’t use a technology or feature. I’m saying the concept of “do as I say, not as I do.....because mine is a small thing and so convenient. But yours is so *bad* ”........ undermines your complaint.

This is a stupid line of reasoning. The idea that wasting a small amount of electricity is the same as wasting a much larger amount... why not just waste an infinite amount of power, if that's your thinking?

I don't know why wireless charging keeps getting mentioned as a waste of electricity, by many people on many threads. It's 80% efficient. If it takes you 3 hours to charge your phone via a 5W charger, switching to wireless will waste about 3 watts per day.

This XBox feature we're talking about in the article uses ~9.5W basically all the time, i.e., 228 watts per day, i.e., it's about 76 times more wasteful than the inefficiency of wireless charging. Saying that it's equally valid to complain about one but not the other... really??

To put the supposed inefficiency of wireless charging into more perspective, my TV uses ~1.5 watts when it's turned off just to check to see if it should turn on via the remote control. So that's "wasting" 12 times as much power as my wireless phone charger. How much does your TV use? And considering that, do you still think that wireless charging is a big problem?
 
Upvote
5 (6 / -1)

Eraserhead

Ars Praefectus
4,565
Subscriptor++
To put that 78 kWH into perspective, the average US household consumes a little over 10 MWH per year.

https://www.eia.gov/tools/faqs/faq.php?id=97&t=3
To put that 10MWH inperspective, except for a few countries in Europe the average EU household consumes around 4MWH per year.
https://www.odyssee-mure.eu/publication ... lling.html
It’s kind of hard to compare on just gross numbers. Nowhere in the U.S. is as poor as Romania or Poland, for instance. Climate makes a huge difference (Californians aren’t particularly low energy users in other ways, but the climate requires relatively little heating/cooling compared to many others). Also depends on whether electric is used for heating (as opposed to natgas or fuel oil).

Even rich people in Britain only use 4-5 MW a year.
 
Upvote
0 (2 / -2)
"Given those numbers, our hope is that most users would be willing to wait an extra 5 to 10 seconds for their console to restart if they knew the impact," Horowitz writes.

Yeah, that's fine, but it's not just about the extra 5 - 10 seconds. It's also about whatever maintenance the system has to do when it boots up. Game updates couldn't complete because the system was shut down.

It feels like we need a middle ground option here. A choice that allows the system to boot up periodically, check for updates and shut back down if there aren't any.

Im honestly surprised that MS didnt add an ARM SOC just for this scenario.

Or perhaps, a couple of Jaguar cores somewhere for the same reason.

I know, easier said than done, but they knew how “important “ and useful this power mode was going to be and how the world is looking more and more to power consumption.

Pussycores are not that great, they're just better than the heavy machinery ones which /were/ complete trash.

Many believe (I used to work on games, like Destiny, proper stuff where we got the good APIs that the peasants didn't get) it was the GPUs that tipped it towards AMD. AMD had been trying for a while to push "hetrogenous computing" and APUs (but it wasn't until the last generation they could actually zero copy and share memory properly) - but there was half assed support (GPU-open) - publishing the register configs isn't that big of a deal when there are revisions for each verson - they just kinda abandoned it

Anyway with the now last gen it actually worked. The L3 and memory system was completely done for the consoles.

AMD also has a great GPU architecture, basically there's a normal scalar CPU and instructions that fall into two classes, vector, or scalar - the scalar one is free to branch jump and do whatever and controls the control flow, any vector instructions it encounters it feeds to the vector units (which obviously only apply it to the live lanes)

This is a great system, very very general purpose, because that scalar core is basically a "normal" processor it can loop, branch back, branch forwards, compare whatever all it wants the vector units just do whatever it throws at them

But yeah the pussycore series was not good, they just didn't suck so bad with more than 15-25 watts between 4. They were only good if you didn't know about Sandy Bridge or even Nehalem - and were like "oh yeah, Bulldozer!" then "oh this is crap, oh yeah Bobcat/Jaguar/Puma"
 
Upvote
-6 (1 / -7)

SraCet

Ars Legatus Legionis
17,220
For comparison the Switch draws about 10 watts when it's active and playing a game.

The switch is also literally a handheld with an included docking station.

What does that matter?

Why does the XBox require more power to do nothing than the Switch requires to play games, regardless of how big the devices are?
 
Upvote
0 (1 / -1)

Ruefus

Ars Tribunus Militum
1,511
For comparison the Switch draws about 10 watts when it's active and playing a game.

The switch is also literally a handheld with an included docking station.

What does that matter?

Why does the XBox require more power to do nothing than the Switch requires to play games, regardless of how big the devices are?

Perhaps because at-idle, the Xbox is far more powerful than a Switch at full-song.
 
Upvote
2 (2 / 0)
"Given those numbers, our hope is that most users would be willing to wait an extra 5 to 10 seconds for their console to restart if they knew the impact," Horowitz writes.

Yeah, that's fine, but it's not just about the extra 5 - 10 seconds. It's also about whatever maintenance the system has to do when it boots up. Game updates couldn't complete because the system was shut down.

It feels like we need a middle ground option here. A choice that allows the system to boot up periodically, check for updates and shut back down if there aren't any.

Im honestly surprised that MS didnt add an ARM SOC just for this scenario.

Or perhaps, a couple of Jaguar cores somewhere for the same reason.

I know, easier said than done, but they knew how “important “ and useful this power mode was going to be and how the world is looking more and more to power consumption.

Forgive the double response I just wanna add:

That's easier said than done (adding an Arm chip in there) - because they have totally different coherency protocols and shit like that, if you want to do that, the Arm chip typically has its own memory and OS and is operating happy and independent of the "main" chip - except that it can interrupt it and poke it about events, eg "wake up now" (a very dumb IC did this for the 360's pads and probably does for them all)

If they both access the non-volatile storage (I did want to say drive but these speeds now are nearing RAM...) then you need to add write barriers and all kinds of nasty shit (and sometimes drive write caches re-order, and ignore being told not to - durable writes are fucking hard trust me) and a special filesystem

It gets difficult fast.

It's easy if you just want one to handle file downloads, you can say "back off while I copy this" or have it feed the main chip the data, but it's not an easy thing to write code for generally, so you'd be writing specific code like receiving a stream from one to the other

But how much storage do you give it? 128mib + minimum RAM to act as a buffer, when it has downloaded 64mib tell the host chip, the rest is for uploads?

That assumes the other chip doesn't want to use the network stuff too. Now you have to make sure they don't pick overlapping ports

It's really non-trivial to be fair.



QUICK ADDENDUM:
------------------------------------

Easiest way is to pinch a core, but this brings in cache pollution and may use precious bandwidth, this is way more of a problem than on older systems. For example RAM sucks at being RAM (I've written about this before, I'll find it if interested) - but is pretty good sequentially, it likes reading from large opened rows, another request comes along and it's for the same bank, gotta close that, get the new one ect

It also may fuck up the branch history, anything the prefetchers were copping onto. This is usually why the frame-rate stutters in some games when you get a message, it's not the overlay - that's relatively easy - it's getting the RAM bandwidth - even if you take and reserve an entire core (which is unpopular) you still get some of these problems (but you wont be context switching and fucking up random thread's caches)
 
Upvote
0 (2 / -2)

cwsars

Ars Tribunus Militum
1,509
Subscriptor++
I don't know why wireless charging keeps getting mentioned as a waste of electricity, by many people on many threads. It's 80% efficient. If it takes you 3 hours to charge your phone via a 5W charger, switching to wireless will waste about 3 watts per day.
Please use correct units-- this is Ars, after all-- power * time should result in units of energy like watt-hours or Joules.

Some folks over at iFixit and medium.com recently did measurements: "Charging the phone from completely dead to 100% using a cable took an average of 14.26 watt-hours (Wh). Using a wireless charger took, on average, 21.01 Wh."

This XBox feature we're talking about in the article uses ~9.5W basically all the time, i.e., 228 watts per day, i.e., it's about 76 times more wasteful than the inefficiency of wireless charging. Saying that it's equally valid to complain about one but not the other... really??
The instant-on XBox feature uses 228 Wh per day, which makes it consume about 30 times as much as the inefficiency of wireless charging. I would agree that the power wasted by wireless charging is not significant in comparison.

The correct solution is to strive for less than 1W "standby/off" power usage. For the Xbox, can't they set a wake-up timer and switch into this ~9.5 W mode long enough to check for updates as necessary, and then go back to lower-power standby most of the time?
 
Upvote
1 (3 / -2)

wk_

Ars Tribunus Militum
2,383
To put that 78 kWH into perspective, the average US household consumes a little over 10 MWH per year.

https://www.eia.gov/tools/faqs/faq.php?id=97&t=3
To put that 10MWH inperspective, except for a few countries in Europe the average EU household consumes around 4MWH per year.
https://www.odyssee-mure.eu/publication ... lling.html
It’s kind of hard to compare on just gross numbers. Nowhere in the U.S. is as poor as Romania or Poland, for instance. Climate makes a huge difference (Californians aren’t particularly low energy users in other ways, but the climate requires relatively little heating/cooling compared to many others). Also depends on whether electric is used for heating (as opposed to natgas or fuel oil).

Even rich people in Britain only use 4-5 MW a year.
US electricity consumption per household are quite high taking into account they typically cook on gas and also heat the water using gas. Not sure what they use that power for.
 
Upvote
4 (4 / 0)

raxx7

Ars Legatus Legionis
17,132
Subscriptor++
"Given those numbers, our hope is that most users would be willing to wait an extra 5 to 10 seconds for their console to restart if they knew the impact," Horowitz writes.

Yeah, that's fine, but it's not just about the extra 5 - 10 seconds. It's also about whatever maintenance the system has to do when it boots up. Game updates couldn't complete because the system was shut down.

It feels like we need a middle ground option here. A choice that allows the system to boot up periodically, check for updates and shut back down if there aren't any.

Im honestly surprised that MS didnt add an ARM SOC just for this scenario.

According to some updates added to the article, if you select "power saving" mode the Xbox still wakes itself up to check for updates once a day and then goes back to using ~1 watt.

Or put in reverse, the benefit of "instant on" is just shaving 5-15 seconds of boot time.
 
Upvote
0 (3 / -3)

raxx7

Ars Legatus Legionis
17,132
Subscriptor++
The correct solution is to strive for less than 1W "standby/off" power usage. For the Xbox, can't they set a wake-up timer and switch into this ~9.5 W mode long enough to check for updates as necessary, and then go back to lower-power standby most of the time?

The article has been updated.
That's exactly what the Xbox does when you select the power saving mode.
 
Upvote
1 (1 / 0)

SraCet

Ars Legatus Legionis
17,220
Please use correct units-- this is Ars, after all-- power * time should result in units of energy like watt-hours or Joules.

I would argue that "watts per day" is the correct unit because people tend to charge their phone about once per day.

Some folks over at iFixit and medium.com recently did measurements: "Charging the phone from completely dead to 100% using a cable took an average of 14.26 watt-hours (Wh). Using a wireless charger took, on average, 21.01 Wh."

I've read that there are huge variations in the charging efficiencies of various chargers. I have no doubt somebody did a test with one that was relatively inefficient. I've read that recent, efficient Qi chargers are up to 75-80% efficient.
 
Upvote
-7 (0 / -7)

cwsars

Ars Tribunus Militum
1,509
Subscriptor++
Please use correct units-- this is Ars, after all-- power * time should result in units of energy like watt-hours or Joules.
I would argue that "watts per day" is the correct unit because people tend to charge their phone about once per day.
You would be wrong about the units. People who recharge their phones once per day consume about 15 watt-hours of energy per day.

(Google understands "power vs energy" if you want to seek independent confirmation.)
 
Upvote
4 (4 / 0)

SraCet

Ars Legatus Legionis
17,220
Please use correct units-- this is Ars, after all-- power * time should result in units of energy like watt-hours or Joules.
I would argue that "watts per day" is the correct unit because people tend to charge their phone about once per day.
You would be wrong about the units. People who recharge their phones once per day consume about 15 watt-hours of energy per day.

(Google understands "power vs energy" if you want to seek independent confirmation.)

Oh, right, sorry, thought you were talking about something else.
 
Upvote
-3 (0 / -3)