Support NeoGAF

[Wiped89]

Updated the OP with the proper stats now that I'm on my PC instead of a phone.

The point of course isn't to say Switch is better than PS4. The fact it uses less power is not surprising. The point is to say, even with the gulf in power between the consoles, the fact the Switch uses so little power is seriously impressive. And you should watch that PS4 sleep mode...

 

[Colossal Moo]

My guess is that Switch's low power consumption comes from the following things:

1) Improvements in semiconductor manufacturing technology. Chips built with smaller manufacturing processes tend to use less power.

2) The Switch's CPU and GPU probably runs at a lower clock rate than the PS4's or Xbox One's equivalent chips. Chips which run at a lower clock rate use much less power.

3) The Switch's chip is probably designed for a portable device. This means the designers engineered the chip to use less power by turning off unused sections, lowering the clock rate when the chip is not being used, etc. There are probably a lot of other tricks I do not know about (I AM FAR FROM BEING A CHIP DESIGN EXPERT).

4) Nintendo targeted a low power design because the Switch operates in portable mode and using less power means a longer battery life. I think this is the most important reason the Switch uses so little power. The X1 and PS4 are optimized for cost and performance. The Switch is optimized for cost, power usage and performance. You get what you optimize for and Nintendo optimized for power usage.

 

[whyreggielie]

An iPhone 7, to say nothing of the model coming out tomorrow.

False in every way. An iPhone 7 has half the ram first of all with a much heavier operating system to hog some of that ram than the Switch. The iPhone 7 graphically is capable of about what an Xbox 360 can output, however it can barely sustain that for more than one minute due to heating issues. Let's not even get into battery life if it could even sustain those graphics levels without overheating.

I'm actually posting from my iPhone 7 and I know damn well that it doesn't come anywhere close to the switch.

 

[JordanN]

There are some people that are vehemently against Nintendo, regardless of the quality of what they're bashing

Ironic, you say this. Because the last two weeks I was praising Nintendo and Iwata in what has probably been a long time.

Grow up, there is no conspiracy. Nintendo is not your dad/mother. They like Sony or MS are all corporations. Defending them because someone offered one very small dissenting criticism about them is just free PR for them.

 

[JP]

The lower the wattage is the better for me but I'm not at all surprised by the power the Switch uses, it's really expected.

 

[Narroo]

False in every way. An iPhone 7 has half the ram first of all with a much heavier operating system to hog some of that ram than the Switch. The iPhone 7 graphically is capable of about what an Xbox 360 can output, however it can barely sustain that for more than one minute due to heating issues. Let's not even get into battery life if it could even sustain those graphics levels without overheating.

I'm actually posting from my iPhone 7 and I know damn well that it doesn't come anywhere close to the switch.

This is the key difference between a portable game system and a phone that everyone likes to ignore!

Ever since the Switch was announced, people have been quick to compare it to smart phones, while ignoring that smart phones can't keep up high-performance indefinitely.

 

[Rellik]

It uses 70W while sleep downloading though according to this article. Which when you factor in the size of games these days, and the number of patches and updates, could be pretty often.

We're getting sidetracked though. The point is the Switch uses way way less in both usage and sleep.

It really isn't often unless you're on dial up.

This has been a bizarre thread watching you try really hard and moving goalposts to knock the sleep mode on the PS4.

 

[geordiemp]

It uses 70W while sleep downloading though according to this article. Which when you factor in the size of games these days, and the number of patches and updates, could be pretty often.

We're getting sidetracked though. The point is the Switch uses way way less in both usage and sleep.

Does the switch background download 20 GB updates to its hard drive.

Oh wait LOL

What a silly post.

 

[RoarOfTheEarth97]

Damn I didn't know the PS4 used that much power when on sleep mode...

 

[llien]

PS4 (average usage): 140W (standard game) up to 200W (high end Pro game)
PS4 (sleep): 10W, up to 70W while downloading
PS4 Pro (max rated): 300W

You have compared apples to oranges (claimed max consumption to actual one).
PS4 PRO (I stress it) in actual tests consumes:

Project cars without patch 104W
Infamous First Light 4K 155W vs 148W launch PS4 1080p
Source gamespot

Which is PRO not simple PS4 and it is like hal fof hte announced number.


Things naturally don't scale that well with power, so something smaller consuming vastly less is not big deal.

Disclaimer: touched switch in Saturn shop, it had the latest Zelda and was connected to a TV, graphics looked very underwhelming to me (not far from what my smartphone can dish out)

 

[Bucket-o-roadkill]

Hardware designed with portable use in mind is very light on power usage, who'd have thought!

 

[Wiped89]

It really isn't often unless you're on dial up.

This has been a bizarre thread watching you try really hard and moving goalposts to knock the sleep mode on the PS4.

I'm not knocking the PS4. I hav owned every PS console from PS1 through PSP, Vita and PS4 Pro now. I'm a PS gamer. I'm merely impressed with the Switch energy efficiency, and tbh a little concerned at the PS4's consumption. I must admit I didn't realise PS4 rest was only 10W when not downloading, which is why I updated the OP.

 

[Handy Fake]

.

 

[John Wick]

It's hilarious how some people are trying to make points where are non, my iPhone also has low rest mode compared to my desktop pc

THIS

My LG mobile when connected upto a TV is far far more power efficient than my PS4 and my PC. That is amazing. It's Magic. A marvelous feat of engineering.

 

[Haano]

The Switch is running on mobile hardware, a glorified Nvidia Shield that was designed to use less power. They run games at lower resolutions that you'd expect. From playing the games, this is very clear. There's minimal objects, clever foggy draw distances, changes to 2D sprites etc. All the tech to keep the power usage normal enough for a good amount of play time.

Plus the switch was designed to run on batteries, so it has to have a low power usage.

Compare that to PS4 and Xbox which are designed to always draw power from a dedicated mains, use PC level technology and run at higher FPS and resolution than the switch.

It's not rocket science, its very straight forward.

 

[JoeNut]

I didn't know PS4 sleep mode used to so much power, damn...I never turn mine off

 

[Cygnus X-1]

Once more, it was designed by Genyo Takeda, right?

And it's an handheld as well, so it definitely was in the requirement specs.

 

[protomouse]

Power efficient hardware. The Nvidia tech inside the Switch delivers greater performance-per-watt than the AMD tech inside the other consoles.

 

[Beerman462]

As others have said it's not that impressive considering.

It's got half the RAM of the consoles and a lower power variant at that, no disk drive, no hard drive, and a tiny cooling fan.

It's also using a more efficient graphics core with NVIDIA.

As far as graphics go it's not great and it's comparable to 10 year old console hardware in that regard. Nintendo games look great, but as usual their cartoony style helps to hide it's short comings.

 

[cuate]

Mobile tech.

 

[ehead]

The new info is much appreciated. Those are actually good numbers. Now, I need to think of a way to make the battery last for a 10-hour flight. Even 6 hours would suffice.

 

[dampflokfreund]

As far as graphics go it's not great and it's comparable to 10 year old console hardware in that regard. Nintendo games look great, but as usual their cartoony style helps to hide it's short comings.

Uhm... no. Switch graphics regardless of the art styles are way superior to last gen consoles, because they are using state of the art techniques such as Physically Based Rendering, SSAO, Dynamic Shadows, high res textures, volumetric and dynamic lighting and so on.

Nintendo's cartoony games all have these modern graphical features, and other games as well, which make them look great besides the art style. Also, would you compare Doom on Switch to a last gen shooter?

 

[Panajev2001a]

Uhm... no. Switch graphics regardless of the art styles are way superior to last gen consoles, because they are using state of the art techniques such as Physically Based Rendering, SSAO, Dynamic Shadows, high res textures, volumetric and dynamic lighting and so on.

Nintendo's cartoony games all have these modern graphical features, and other games as well, which make them look great besides the art style. Also, would you compare Doom on Switch to a last gen shooter?

Do not know about DOOM, let’s wait to see its effective resolution when under load and remember it is still 30 FPS (more than double of the headroom compared to a 60 FPS one, think of bits like input processing that you cannot scale and thus take a larger percentage of frame time). Switch is lucky to have very rich feature set wise GPU: the core they built it from is quite a large jump and easier to use compared to previous generations.

 

[geordiemp]

Power efficient hardware. The Nvidia tech inside the Switch delivers inline performance-per-watt same as the AMD tech inside the other consoles.

Fixed that for you.

Comparing 2017 models (latest nm nodes)

Switch ~ 0.5 TF 20 Watts consumed
Pro ~ 4TF 150 watts consumed

Adding Hard drives and disks, or switch screen adds +/- so seems about right and inline with expectations.

Unless your suggesting Switch is doing 1 TF with 20 watts or 0.5 TF with 5 watts, then that would be above expectations.

Posters going back to older nodes (2013 Ps4) or quoting power rating is very funny.

 

[Panajev2001a]

Fixed that for you.

Comparing 2017 models (latest nm nodes)

Switch ~ 0.5 TF 20 Watts consumed
Pro ~ 4TF 150 watts consumed

Adding Hard drives and disks, or switch screen adds +/- so seems about right and inline with expectations.

Unless your suggesting Switch is doing 1 TF with 20 watts or 0.5 TF with 5 watts, then that would be above expectations.

Posters going back to older nodes (2013 Ps4) or quoting power rating is very funny.

Isn’t the Switch figure of 0.5+ TFLOPS assuming FP16 heavy shaders? In that case for the Pro you would put ~8 TFLOPS (kind of cheeky for both).

 

[geordiemp]

Isn't the Switch figure of 0.5+ TFLOPS assuming FP16 heavy shaders? In that case for the Pro you would put ~8 TFLOPS (kind of cheeky for both).

Do we really know the Switch TF numbers ? Anyway, Switch is also on 20 nm is it not so not latest 16 nm but lower than original ps4 and xb1 ?

Anyway, still power draw numbers are what you would expect for the consoles power level.

If the Nvidia chip had been on 10 nm or something cutting edge for 2017 it would of been 10 W or more TF and would of been a good discussion.

 

[jts]

Why are we comparing TF when it comes to the consoles’ sleep modes? They’re not doing any graphical work, and very little, very efficient processing (ideally).

It’s just that ARM runs around x86 for those kinds of things, and even more so when paired to flash memory instead of a mechanical hard drive.

 

[richman555]

Partnership with NVidia has alot to do with the success of the Switch. Nintendo needed a technology overhaul from its aged Power/AMD architecture.

 

[test_account]

To be fair, i dont think its surprising that the Switch uses less power than a PC and a PS4/Xbox One, but i do think its impressive that techology in general have gotten to the point where you can get very good performance and use relatively little power consumption.

That includes having a game suspended yeah. Sorry I was on my phone when I posted so I'll update today with source links and more figures. When charging, it increases to 5W.

I understand. 1W seemed really low i think, but thinking about it, i think the Vita can be in sleep mode with a game suspended for maybe two weeks or so, so it would make sense that the Switch also uses very little power while being in sleep mode.

I didn't know PS4 sleep mode used to so much power, damn...I never turn mine off

Its not really that much. The "up to 70W" is if you download, then the machine have to write to the HDD and use WiFi/Ethernet as well. If you have your PS4 in standby 24/7 for a whole year, it might add $15 - $20 to your yearly usage. About $1.5 extra a month.

 

[Ganondolf]

If I r remember correctly the ps4 uses a lot of watts whilst downloading because the whole soc is either on or off so even a small job would require all cores active.

The switch uses less watts all round as it uses mobile based hardware which nvidia made very efficient for the performance.

Not sure why people are comparing TF to show power draw as power draw does not scale evenly with TF increase.

 

[jobrro]

If I r remember correctly the ps4 uses a lot of watts whilst downloading because the whole soc is either on or off so even a small job would require all cores active.

The switch uses less watts all round as it uses mobile based hardware which nvidia made very efficient for the performance.

Not sure why people are comparing TF to show power draw as power draw does not scale evenly with TF increase.

Yeah they had an ARM chip on board for the job but for some reason it didn't meet their requirements after the fact so they use the main CPU in Rest Mode.

That disappointing in the base model PS4 from 2013, not sure why they didn't upgrade the ARM chip for the Slim/Pro, or at the very least make an "Energy Efficient" rest mode box that has slower downloads but doesn't require as much power. Unless I am preloading I don't care too much about the Rest Mode download speed most of the time.

 

[Screamapillar]

Switch is basically a low-end tablet, when speaking of the specs. Nintendo also prides itself on acheiving low power draws, even on home consoles.

 

[Elfstar]

The one thing i don't like about Nintendo Switch™ is that it's almost too perfect.

 

[ecosse_011172]

This thread is embarrassing, wtf is wrong with you people?
Switch is a great, power efficient console (and games look great on a good Tv no matter what some Clown says).
The Ms consoles are great machines with great controllers, only lacking in exclusives.
The Sony consoles are great, only problem being the noise levels.

I.e. all the current machines are great.
If you are here to shit on the Switch power levels and attack the OP, you need to have a word with yourself.
Go and play some games you dicks.

 

[ecosse_011172]

And Switch is not a fucking tablet you morons.
Go and do all the things you do on your tablet on it, Netflix, web browsing, email, Facebook etc.
Then go play games of the Switch 1st party calibre on your tablet.
We know what happens.

 

[Nikana]

Switch is basically a low-end tablet, when speaking of the specs. Nintendo also prides itself on acheiving low power draws, even on home consoles.

Dude what lol. The x1 chip is now "low end?"

 

[ZOONAMI]

don't tell him there are phones with lower wattage that are more powerful than the switch either

(he might get upset)

Phones throttle much worse than a switch, imo there actually is not a phone more powerful than switch.

 

[llien]

morons.

You can do better than that.

 

[kristianity77]

The switch for a console that uses only 20 watts is impressive but only when you compare it to older technology. This thing pushes out graphics on a par and slightly beyond last gen systems. Using probably 25 percent of the power. Plus it's driving a screen and speakers. It is impressive no matter how you look at it.

Having said that though, I was more impressed with the vita at the time. I couldn't get that to draw more that 4 watts during gameplay when I tested it. 4 watts!!