UNREAL Standby Time! ARM Destroys Intel - Snapdragon X Elite - Asus Vivobook S 15
HTML-код
- Опубликовано: 1 окт 2024
- Link - shop.asus.com/...
Links to my Social Media
▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰
Patreon 🅿️ / phawx
Twitter:🐦⬛ / carygolomb
Instagram: 📸 / thephawx
TikTok: 🎵 / thephawx
▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰▰
Not 1 watt per hour
1 Watt hour per hour.
Which sounds awkward I know. You can instead use say "the laptop consumes one watt while in standby"
One watt is about what the Steam Deck also does by the way
While it's not technically correct to say "watts per hour," using the relationship between watts (power) and watt-hours (energy) will still lead you to the correct answer when calculating how long a battery will last.
When you calculate battery life, you're essentially using the correct relationship:
Battery Life (hours)
=
Battery Capacity (Wh)
Power Consumption (W)
Battery Life (hours)=
Power Consumption (W)
Battery Capacity (Wh)
@@mechanicalmonk2020 watt hours per hour.... Also known as watts... Lol
Phawx did misstate the units several times, which is just needlessly confusing. Also, one watt for standby power is stupidly high in the first place; pretty much all that should be active is memory self-refresh. There's no good reason for a laptop to have worse standby time than a cellphone; we're talking less tasks and far bigger batteries. Modern cell phones are doing rather badly too.
@@DJ_CRIZP one is a unit of power used/stored, the second one is of power consumption. dont mix them like this.
and yes, saying watthours per hour is completely correct if you want to say how much energy it consumed during a specified timeframe.
saying it consumed 1 Watt in an hour is incorrect as a Watt is not a measure of total power consumed.
you can say it was consuming 1 watt while it was sleeping, but you cannot specify a timeframe.
@@davidpodeszwa7010 power is not stored or consumed. Energy is. Power is energy per unit time. Expressing power draw as 'watt-hours per hour' is like expressing speed as 'miles per hour-hours per hour'. It makes no sense. During the multiplication of units, you have to start cancelling units somewhere, otherwise you would get ridiculous units. If you want to measure power draw (in metric and especially in electronics), you use watts. Anything else just means that you don't know the 101 of electrical engineering and probably shouldn't be commenting implying that you do. Lol
I have the Snapdragon X Plus Surface Laptop, and have charged it once since receiving it last week. The standby time on it is bonkers. Full disclosure I didn't buy it, it's a review unit sent to me by BestBuy for review via Tech Insider Program. I'm really impressed by the Snapdragon X SoCs this is the first laptop I've owned that I don't even think about shutting down when I am done using it, just drop the lid throw it back in its bag and carry on with life. Pull it out the next day and it basically hasn't lost any battery life. Now if we could get some better gaming performance out of one of these in a generation or two, we would have a contender for a new Handheld chip. Imagine a new steamdeck like running on something like this with absolutely rediculus standby time, and that instant wake back into your game? I'd be in love.
His last video was literally showing steamdeck being 3 times faster than this arm chip at doom eternal at same power usage.
Batteries degrade with charge cycles, if that's yours and doesn't need to go back, you should never run that off wall when you have a wall, never ever, that battery is a big ticket replacing
@@hamzashaikh9310 More performance at the same power draw
Almost nobody needs standby more than a day, this vid looks an advertisement rather than important info.
He needs to run Powercfg.exe /SleepStudy on each box, I'm betting his Qualcomm is not doing as much backwound during modern standby, or it's entering hibernate soon after lid is closed
@@MrPerrisC Far from the only thing that degrade batteries. Keeping a battery at 100% or empty for prolongued periods is actually worse for the battery than healthy cycles (= cycles that keep it away from maximum and minimum voltage)
@@hamzashaikh9310 and I literally said in a generation or two, I am damned well aware of how anemic the integrated graphics are on this chip. No one is claiming the current generation of snapdragon x chips elite or otherwise have any real gaming grunt. And during active use I don’t expect it to last much longer than anything else when gaming even with a theoretical new generation one being able to match or beat whatever high end integrated graphics AMD cooks up in the future. But if we can keep similar standby power draw as what these are showing now, then it would be a contender. As it stands now even with all the improvements, if I put my steam deck in standby, and forget to go back to it to continue playing or at least save and quit and shut down, I’m looking at a dead or near dead steam deck the next time I go to use it. And this is on my OLED deck that has otherwise fantastic battery life by any other metric. So I’m not sure what point you are trying to make?
The Intel system just seems like an outlier. Could you maybe try more devices? I know SD X Elite may come out on top, but 20% over 12hr standby it quite a lot.
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
Not at all. My laptop has same issue, its always dead after 1 or 2 days of sleep
@@lasue7244 it could be issue for your laptop. But not for everyone
@@HDRPC I've seen some articles complaining about an intermittent bug with windows modern standby could be that
@@MarvoloRiddle no bug. Mine is working fine.
Nooooooooo not 1W an hour
its 1Wh/h so just 1W
I know you really know your stuff and you know how it works... but please, PLEASE, stop saying watts per hour :X it makes no sense at all... Just say watt. The unit already means a certain amount of energy over a certain amount of time (literally it's joules per second). Watt per hour would be x joules per second per hour... what is this, an acceleration? And if you actually meant watthours per hour but just lazy on saying watthours, well that would be correct, but why would you multiply watt by hours to then only divide by hours again? Just say Watt :O ugh, my brain hurts... :P
Seriously though, it may seem like nitpicking, but a lot of people watch you and using the terminology this badly teaches other people to do it wrong as well.
Great tests though :)
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC Are you a bot? Every reply it's the same, with most comments.
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
@@KyleRuggles because youtuber is doing shenanigans
You definitely need to test with more Windows laptops, Intel and AMD. And maybe throw in some Linux too?
All batteries also feature a natural leakage, so even in hibernation mode, you would lose a bit of power overtime.
It would be interesting to compare this battery-only leak to standby depletion, and look at the difference.
It may appear that the SnapDragon is consuming even less power than this already very successful test implies.
He needs to run Powercfg.exe /SleepStudy on each laptop, I'll bet his x86 are either doing more, or entering hibernate at a reasonable depletion rather than a few percent
"Modern standby" must have hibernate enabled to guard against critical battery level
Well okay you're basically begging for this so I'll do it: Watt is power. It is an instantaneous measurement. Watt hour is energy. It is the cumulative value. You don't consume "watts per hour", you consume watts. After some time, the energy consumed is "watt hours", the average power (watts) multiplied by the time, or the integration of watts over time, if you like maths.
If you want to talk about energy in terms of a simple unit, that unit is Joule. Power can then be measured in e.g. J/s. People in physics and biology tend to use units like this. But people dealing with electricity don't generally do this. Instead, we use watts for power and watt hours for energy, because power in watts is a simple multiplication of voltage by current.
You're both wrong
@@farstranger69 neither of them are wrong.
But Wh is a more useful unit than Joule for electronics
It's all a matter of definition. We all know what he means, and no one is wrong here. Watts per hour makes sense and is no different to watt hours (unless you are a machine incapable of extrapolation , then you might get confused).
"Uhm actually" reddit ahh comment
The Snapdragon devices are using device tree instead of ACPI. This is making it difficult for Linux to get ported onto them.
Generative AI, including image generation, only exists because companies like OpenAI steal from artists without attribution or compensation. Just because they've improved at discouraging the models from creating images with shutterstock watermarks and artists' signatures does not make their output ethical. Gen AI certainly does not fall under fair use: they are not using copyrighted works for critique or parody, but rather they are used in their entirety to train statistical models that they sell access to in a way that fundamentally threatens the ability of real artists, journalists, writers, and others to profit from their work. Reportedly, OpenAI has even used transcriptions of youtube videos to train GPT-4. Your work has probably been stolen to feed this machine, so why are you supporting it?
So, there is some benefit to that Snapdragon X Elite - even if it is very, very small. Of course you can just use Hibernate option on your normal Intel/AMD laptop if you are not satisfied with standby
Interesting how youtubers are spinning the wheel on these laptops ... stand by time ... lol ... I have an m3 pro macbook pro and it eats the x elite for breakfast and it can do it all day long. Once you start loading x elite with real productive workloads (those few that right now work properly) and you see intel level or amd level battery consumption .. nice try though
How does this compare to AMD?
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC thank you for your input, but I asked about AMD because everyone keeps saying "AMD is great for efficiency" yet there's not a lot of data on specifically the standby state. Phawx here compared best case vs worst case. I'd love to know what else sits in the middle
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
@@EduardLevesque
Snapdragon junk. You people are loving that youtuber shenanigans.
@@ultraviolet2497 IntelCoreUltra is more efficient.
is this a problem with Windows' modern standby or is it hardware!?
this should've been tested with other OSes
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
@@EduardLevesque Keep loving that youtuber shenanigans
Mr. Phawx, I'm even curious if you could test how many watts it takes to wake up the device. It could be that a sizeable portion of that 0.7W is attributed to just the wake up/boot up part of the device and not necessarily just on the drain when it's on standby. 🤔
How long could it possibly take to "wake up"? That power draw should be pretty negligible compared to the extended period when it's asleep
at this point pretty much anything destroys intel
I don't think my laptop's standby power drain is that bad as on the x86 laptop you used for the comparison, but this is very very impressive.
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
The average laptop is up there tho. The thging about these arm chips is that they are definitely cheaper to produce. And so we will see cheaper options that actually make sense, of course windows is to blame... If we need arm to have usable laptops something is wrong.
@@ibelieveinjesusinmyspareti2861Why do you think so?
@@ibelieveinjesusinmyspareti2861 If they are cheaper then why would qualcomm charge us higher prices compared to relative performant x86 laptops?
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
Please Qualcomm figure out your GPU problems. We need ARM chips in handhelds
Drivers will improve
have u seen how worse it is in gaming, i dont think we need it as of now..... for work its ok
isnt the problem is more likely difference in architecture and translation layer? games dev need to create native apps
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@LuminousSpaceit is dead on arrival?
@@LuminousSpace it out performs the steam deck even running under emulation. It’s driver support and consistency that’s the issue. Spider-Man miles morales can run at 50-70fps 768p without fsr. And it comes with a built in ai upscaler using the npu. Could be good in a windows handheld once drivers get sorted out
Our car UI also lists rate of consumption (power) as kWh/h. Which is silly.
Let’s simplify h/h = 1. And we get kW as our unit. Kilowatts. Watt is the unit of rate of consumption (power). Watt*hour (Wh) is energy. You can’t do W/h. That would be rate of consumption dividided by hour. Which makes no sense.
How do you say “69” without saying “Nice!”?? Can’t trust reviewers unless they participate. (Lol, JK.)
Well this is actually identical to my Fujitsu using the 1335u and 60wh battery (some other tweaks) but most importantly, Linux.
Remind me again: what's the best standby mode for x86 CPU? Just got a new laptop and I almost forgot this also needed to be adjusted.
If some one needs a point for comparison my steam deck had 90 something % battery life last night and is now at 85% after 15 hours of stand by off the charger. That Intel standby battery life looks to be slightly better than what I used to get with my Haswell U based laptop which is now 10 years old. Wow, much stagnation
@Phawx the mad scientist of gaming 🔭 🔬 ⚗️ 💻 🧪🥼🥽 🎮
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
I love the intro where the thumbnail comes to life!
yes,yes,yes, but its still sucs if you want to run anything other than arm optimization apps😅
A lot of people are getting on the Phawx about the statement "watts per hour" And while it's not technically correct, he is using the correct relationship so it is a moot point.
While it's not technically correct to say "watts per hour," using the relationship between watts (power) and watt-hours (energy) will still lead you to the correct answer when calculating how long a battery will last.
When you calculate battery life, you're essentially using the correct relationship:
Battery Life (hours) = Battery Capacity (Wh) / Power Consumption (W)
Naming nearly every unit wrong is just making a mess of things even if what you're calculating had a meaningful value.
That's 50 days dunno why he didnt say that aswell.
Snapdragon X Elite 🔋🔋💪🔥
Might just be my phone. But incase it's not - the sound is out of sync.
for me its's slightly
Yeah! It's great! The best thing for now for snapdragon laptops is put them to sleep! I'm sorry but it's sad true.
😂😂😂😂
I own one (Pro 11) and it's super nice to use. Gaming is meh, but it's not for gaming.
how much power does hibernation use? Its the setting that i mostly use. Takes like 1 second longer to get back to windows vs sleep, big deal. I know its less, but i never benchmarked it
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC lunar lake will be great. But my laptop has a 10300H, and it really does not lose a lot of power during hibernation. Luckily windows modern standby isnt available for it because that sounds horrible tbh
it was a big deal with hard drives thanks to reactiin time, power consumption and speed in general, now with nvme is just nothing basically
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
Watt per hour is redundant. That's like miles per hour per hour.
dropping 20% after 12 hours sleep is completely not normal for x86 laptops. i usually get 0.4 - 0.5% / hours on a properly setup machine. Sure, ARM and Mac are still better, but not that earth shattering better
How do the last Intel and AMD generations behave? Intel 13th gen isn't the freshest generation to compare with...
Meteor Lake is terrible
First
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
Thats because you are using 24H2 on ARM
You intel device has issues.
this guy is commenting on a bunch of comments. coping hard with his intel junk
intel is cooked
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
since ryzen, and it istill on the oven, and the heat is increasing, no one to open the oven, it is burning basically now
Whilst the demonstration shows information that illustrates differences, as a function I find it rather sad that people would actually leave a machine for long periods, where it is doing absolutely nothing of value, instead using energy, regardless of how much or little, in a standby state, rather then doing the sensible thing and switching it off to use zero power.
If people really are that impatient, then it is more an indictment of society than anything else.
What is way more practically important are features that can lower power draw for applications and processes that don't need it, allowing only what is needed. That's an energy feature that carries actual weight and benefit.
Atleast chinese phones do this very well. Modes for benchmarks, modes for specific games, modes for youtube, calls etc.
It's what you expect for an architecture made for mobile. x86 has a long way to catch up...
that's not entirely true, there are plenty cases where on Android' Google Play services drain the battery like a motha'fucker
@@zMeul My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC you are aware the H series Intels are the performance segment and not the power efficient segment ? That would be a U series chip. Anyway the main problem with Intel laptops is the quality of implementation in a lot of them. Just recently I have seen a Linux kernel patch for Intel laptops that do not advertise their power capabilities properly (meaning they do not report boost states etc to the OS in the standard way they should). These early ARM things are hand picked models and probably heavily monitored by Qualcom ... let's wait for the cheap devices and how those will perform ...
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
Love it
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
That's CRAZY.
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
@@HDRPC good to know that with Meteor Lake the situation is better. I know LNL will be much better then what he showed in this video, and i hope Strix will be better too, because i need x86 to work and it seems that some program crash, others runs slowly because they can't use hardware acceleration, while others don't run at all trough emulation with X Elite. I know there are other people in the same situatio, and that's why i hope Strix and LNL will perform much better then the laptop tested in the video.
@@matthewsykes2646 fact is X Elite is dead or arrival
@@HDRPC yes, it's kinda dead on arrival but it still showed that,this time, Microsoft and developers are serious about Windows on Arm. The good part of this story is that if you need a very energy efficient laptop for office work or some light task you have more option with X Elite designs laptops, but if you want, or need, the best performance for video editing, gaming, 3D or other things you have a HUGE amount of choice with Meteor Lake and Hawk Point and, starting from the next month with Strix Point first and then in Spetember with LNL, there will be even more choices
@@HDRPC this guy is commenting on a bunch of comments. coping hard with his intel junk
You already showed x86 is more performant at the same power draw, "modern standby mode" runs background services differing on different hardware, your x86 laptops are probably doing more or your snapdragon is going into hibernate sooner
RUN Microsoft's SleepStudy from CMD do it on all laptops
Lol my i5 Tiger Lake X1 Carbon lasts for weeks in standby, but that might be bc of Linux. Would be nice if you also compared with Linux to see if it's Windows doing it's thing in the background
Shocking. A tablet with keyboard having good standby time
One question I do have is how the AMD laptops stand up against the Snapdragon.
Snapdragon X Elite 💪🔋🔋🔥
The thumbnail to video transition is phenomenal!
Second
Actually watching this on my lunch break on my lenovo yoga snapdragon laptop. Really enjoying the laptop so far!
Legion Snapdragon???
@@Xevos701 🤣🤣
@@Xevos701 actually the Lenovo yoga, my bad (legion was my last laptop). Still a snap dragon x model tho. Really nice so far
@@danielcable842 have you tried gaming? How they work without Nvidia or AMD drivers.
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
My Samsung Galaxy Book 4 Pro 62.1Wh dropped 2% in 8 hours of sleep.
Windows Balanced mode.
It has IntelCoreUltra 5 125H, it is a Chiplet architecture with 6nm/5nm nodes.
Lunar Lake is coming with Enhanced sleep states. MoP and Memory Side Cache.
Ok bot
I dont even know of he Is a bot. He Just spamming tho.
this guy is commenting on a bunch of comments. coping hard with his intel junk
@@ibelieveinjesusinmyspareti2861
This youtuber is doing shenanigans.