Excited to see your updates! I’ve narrowed down my new laptop search between this (16” screen tho) or the brand new Asus Zenbook S16 coming out (hopefully in the next few weeks). I might end up going with the Asus just because there’s an option for 32gb ram. I also have the s9 ultra tablet so seamlessly connecting through the Samsung ecosystem is also a bonus.. I use the tablet connected to the PC as a second screen for drawing / editing in photoshop (I pay for the adobe suite - I don’t use all the apps but want to make sure this device can handle the basics). I like to play around with the latest AI LLMs/image/video too. Lastly, I’ll be going to school for cybersecurity so I need to make sure the device I get can handle the tasks involved with that & my hobbies mentioned above lol… ahhhh choicesss 😅😪
The screen on this laptop is the best quality screen I have ever used on any device, iPads, galaxy tabs , tvs, laptops, anything. This is an absolutely fantastic device. I love windows again. bye bye MacBook
Thanks for your in-depth review. As of today, Samsung has discounted the 14 inch Galaxy Book 4 Edge. It's heavily discounted. Instead of $1349.99 its reduced to $599.99. The 16 inch is discounted from $1449.99 to $799.99.
I got the Galaxy Book 4 Edge 16 inch and love seeing you do these tests. I've been using mine on High Performance mode and it gets a little warm, but not too hot and the fans running aren't very loud, but I've yet to do anything super intensive. I've been doing 4k RUclips playback mostly. Will be running my own tests, but there aren't really any other people with this laptop doing these tests. Glad that I found your channel. Keep it up! New subscriber!
I think they also settle down a bit after the first major round of Windows updates, assuming you haven't finished those yet. There's a Samsung firmware update somewhere in that mix that could have something to do with it. I thought about going with the 16" instead of the 14" considering how portable these things are. They're almost incomprehensibly lightweight... I love it.
I saw a price leak of the Asus vivobook AMD suggesting $1300 for the 32GB version with 2TB storage. I suspect AMD will be very price competitive with snapdragon to stop migration over to windows on arm.
I'm sure that's probably true. Especially since Qualcomm officially has lithographic superiority with this current chipset (at least, in terms of CPU efficiency metrics).
@@OtherSteve And Qualcomm has exclusivity rights for ARM on Windows for the REST of 2024. Plenty of time to refine, enchant and win over market share before exclusivity rights expire in 2025, which is when ARM, MEDIATEK, NVIDIA will try to enter the market with their ARM versions. It's a whole new world. By then though QUALCOMM will have a leg up, user review data, and better performance, PLUS the 4yr pc cycle beginning this time of year from 4 year Covid stay/work at home run ragged pc's... Qualcomm will grab that initial crest all through the rest of the year and back to school as well. And Windows Copilot is amazing. Longer battery life and low fan noise will be a hit with consumers.
I got the 16 in version, and I see a bigger difference changing the windows power modes, where Best Power Efficiency, cuts single core and multicore performance in half, balanced cuts multicore performance to 60% sometimes, where best performance is the reference in comparison; and energy saver cuts multicore performance by to 40% where single core is half
Well, I'm certain that Quiet + Best Power Efficiency will probably yield the longest runtimes, but of course, the question there is at what cost to performance? I'll try to check various combinations over the course of the next several days, but as of now, my money is on Quiet + Balanced for the most reasonable power savings without sacrificing substantial performance. It's also nice that performance appears to still be very good while plugged in in this mode. As such, it may turn out to be a sort of "set it and forget it" situation that most people could probably benefit from as soon as they unbox.
Very interesting video indeed. How much time did you manage to get out of the battery? I am planning to buy rhe 14inch version, and the two things that worry me the most are battery and coil whine.
So I should preface this by saying that I have discovered that one of my primary uses, taking calls via Phone Link using the Windows Studio Voice Focus NPU-powered AI voice filtration (which is AWESOME by the way), drains battery pretty quickly. I think it accounts for maybe 15 to 20% of the battery usage while it's happening... and I do it for hours every single workday. Having said that, on Quiet + Balanced, with 90 to 100% brightness, I'm getting between 8 and 10 hours on a full charge. I'd imagine without Phone Link, you could add an hour or two to that (probably 10 to 12 hours). On Optimized + Balanced, it's a bit less, but hard to say precisely how much less. Probably an hour? On Optimized + Best Power Efficiency, I don't notice much longer runtimes. I think Quiet + Balanced is probably the way to go for most people. Lower screen brightness settings really make a big difference here. The AMOLED panel is a battery hog at higher brightness settings. With normal indoor lighting, I find ~70% brightness to be the most comfortable setting for my own use. The luminance curve is not linear. I don't have the tools to measure it here, but 100% is probably around twice as bright (or more) as 70%.
@@OtherStevethanks, this is really helpful indeed. I watched some videos in which reviewers were complaining about battery, but 9 hours is a very respectable figure for an Oled panel.
I think it does somewhat depending on the application. It's hard to measure actual substantive performance differences on these newer chipset ARM models without just using the product and observing. However, I've done some video rendering recently and I do believe it's quicker on High Performance (Samsung) + Best Performance (Windows).
You need to loop the test for a bit to get useful numbers from the average column. Proper way to get averages for load is to start a looped test, then reset the counters. Let it run for 15m or so, then check the averages before you stop the load/benchmark. That laptop is showing ~58w during the geekbench run, not 13w.
Hey, thanks for watching! Yes, I realize that you need to run it for 15 minutes to get a realistic average. The purpose of this video isn't to get a long term average under load however. I'm really looking more at the momentary numbers in this video and the drop from the peak load down to the next level more than I am averages here; when I was resetting the counters and working with geekbench and furmark, it was mostly just for a rough reference. Likewise, realistically, I don't think geekbench represents a typical load from an office worker, so I was kind of more interested to see how the power profiles handled CPU loads from both applications (and the differences between them). It's kind of irrelevant anyway because the results I think speak for themselves regardless: The quiet profile is substantially less demanding in terms of power consumption at a similar level of realistic performance, unless we're talking GPU, in which case it's probably about 20% lower. For that reason, it's probably safe to expect substantially longer battery runtimes under typical work day conditions for the average office user when set to Quiet. I'm going to put this to the test in the real world and see how it turns out. So far, I've been getting somewhere between 9 and 10 hours on Balanced + Optimized, so if Balanced + Quiet provides comparable performance at 20% more battery life, I'd say that's a huge win. It's closer to the defaults Lenovo and Microsoft seem to embrace on their machines, which maybe helps explain the efficiency discrepancies between these devices after accounting for battery capacity differences.
I'm not sure I completely understand what you're saying, but if you are saying that the metrics are wrong, maybe they are. It's hard to tell, but there is no reliable way to measure power consumption on these SoCs yet directly (at least, not to my knowledge). Notebookcheck has run some recent benchmarks using external monitors in an attempt to get their arms around this, and I'm sure full battery drain tests are probably the only *truly* reliable way of discerning just how it affects runtimes... but even still, you have to admit, the delta between the different power settings *must* suggest some sort of adjustments to TDP/power provisions to the chipset. That's kind of the entire point here.
Thank you. Looking forward to your real world tests.
Excited to see your updates! I’ve narrowed down my new laptop search between this (16” screen tho) or the brand new Asus Zenbook S16 coming out (hopefully in the next few weeks). I might end up going with the Asus just because there’s an option for 32gb ram.
I also have the s9 ultra tablet so seamlessly connecting through the Samsung ecosystem is also a bonus.. I use the tablet connected to the PC as a second screen for drawing / editing in photoshop (I pay for the adobe suite - I don’t use all the apps but want to make sure this device can handle the basics). I like to play around with the latest AI LLMs/image/video too. Lastly, I’ll be going to school for cybersecurity so I need to make sure the device I get can handle the tasks involved with that & my hobbies mentioned above lol… ahhhh choicesss 😅😪
If you can afford the Zenbook, that's certainly a great choice
@Workaholic42 Going to wait a few more weeks until the Zenbook is out and watch for the reviews! I appreciate the reply!! 😁🙏🏼
The screen on this laptop is the best quality screen I have ever used on any device, iPads, galaxy tabs , tvs, laptops, anything. This is an absolutely fantastic device. I love windows again. bye bye MacBook
Thanks for your in-depth review. As of today, Samsung has discounted the 14 inch Galaxy Book 4 Edge. It's heavily discounted. Instead of $1349.99 its reduced to $599.99. The 16 inch is discounted from $1449.99 to $799.99.
I got the Galaxy Book 4 Edge 16 inch and love seeing you do these tests. I've been using mine on High Performance mode and it gets a little warm, but not too hot and the fans running aren't very loud, but I've yet to do anything super intensive. I've been doing 4k RUclips playback mostly. Will be running my own tests, but there aren't really any other people with this laptop doing these tests. Glad that I found your channel. Keep it up! New subscriber!
I think they also settle down a bit after the first major round of Windows updates, assuming you haven't finished those yet. There's a Samsung firmware update somewhere in that mix that could have something to do with it.
I thought about going with the 16" instead of the 14" considering how portable these things are. They're almost incomprehensibly lightweight... I love it.
I saw a price leak of the Asus vivobook AMD suggesting $1300 for the 32GB version with 2TB storage. I suspect AMD will be very price competitive with snapdragon to stop migration over to windows on arm.
I'm sure that's probably true. Especially since Qualcomm officially has lithographic superiority with this current chipset (at least, in terms of CPU efficiency metrics).
@@OtherSteve And Qualcomm has exclusivity rights for ARM on Windows for the REST of 2024. Plenty of time to refine, enchant and win over market share before exclusivity rights expire in 2025, which is when ARM, MEDIATEK, NVIDIA will try to enter the market with their ARM versions. It's a whole new world. By then though QUALCOMM will have a leg up, user review data, and better performance, PLUS the 4yr pc cycle beginning this time of year from 4 year Covid stay/work at home run ragged pc's... Qualcomm will grab that initial crest all through the rest of the year and back to school as well. And Windows Copilot is amazing. Longer battery life and low fan noise will be a hit with consumers.
@@OtherSteve are you talking about strix?
I got the 16 in version, and I see a bigger difference changing the windows power modes, where Best Power Efficiency, cuts single core and multicore performance in half, balanced cuts multicore performance to 60% sometimes, where best performance is the reference in comparison; and energy saver cuts multicore performance by to 40% where single core is half
Cant wait to see that competing with book5 with lunar lake in a few months, finally some good options for windows users
Thanks for these tests. So, what setting combination will give the maximum battery ? And how much will it translate in hours ?
Well, I'm certain that Quiet + Best Power Efficiency will probably yield the longest runtimes, but of course, the question there is at what cost to performance? I'll try to check various combinations over the course of the next several days, but as of now, my money is on Quiet + Balanced for the most reasonable power savings without sacrificing substantial performance. It's also nice that performance appears to still be very good while plugged in in this mode. As such, it may turn out to be a sort of "set it and forget it" situation that most people could probably benefit from as soon as they unbox.
@OtherSteve thanks. If you could test in quiet + best power in real usage, that would be great.
@@sham8996 that's a good idea. I'll check into that as soon as I get the time and let you know.
@@OtherSteve awesome, thanks.
@@OtherSteve any update?
Very interesting video indeed. How much time did you manage to get out of the battery? I am planning to buy rhe 14inch version, and the two things that worry me the most are battery and coil whine.
So I should preface this by saying that I have discovered that one of my primary uses, taking calls via Phone Link using the Windows Studio Voice Focus NPU-powered AI voice filtration (which is AWESOME by the way), drains battery pretty quickly. I think it accounts for maybe 15 to 20% of the battery usage while it's happening... and I do it for hours every single workday.
Having said that, on Quiet + Balanced, with 90 to 100% brightness, I'm getting between 8 and 10 hours on a full charge. I'd imagine without Phone Link, you could add an hour or two to that (probably 10 to 12 hours). On Optimized + Balanced, it's a bit less, but hard to say precisely how much less. Probably an hour? On Optimized + Best Power Efficiency, I don't notice much longer runtimes. I think Quiet + Balanced is probably the way to go for most people.
Lower screen brightness settings really make a big difference here. The AMOLED panel is a battery hog at higher brightness settings. With normal indoor lighting, I find ~70% brightness to be the most comfortable setting for my own use. The luminance curve is not linear. I don't have the tools to measure it here, but 100% is probably around twice as bright (or more) as 70%.
@@OtherStevethanks, this is really helpful indeed. I watched some videos in which reviewers were complaining about battery, but 9 hours is a very respectable figure for an Oled panel.
does the high performance mode make any difference
I think it does somewhat depending on the application. It's hard to measure actual substantive performance differences on these newer chipset ARM models without just using the product and observing. However, I've done some video rendering recently and I do believe it's quicker on High Performance (Samsung) + Best Performance (Windows).
You need to loop the test for a bit to get useful numbers from the average column.
Proper way to get averages for load is to start a looped test, then reset the counters. Let it run for 15m or so, then check the averages before you stop the load/benchmark.
That laptop is showing ~58w during the geekbench run, not 13w.
Hey, thanks for watching!
Yes, I realize that you need to run it for 15 minutes to get a realistic average. The purpose of this video isn't to get a long term average under load however. I'm really looking more at the momentary numbers in this video and the drop from the peak load down to the next level more than I am averages here; when I was resetting the counters and working with geekbench and furmark, it was mostly just for a rough reference. Likewise, realistically, I don't think geekbench represents a typical load from an office worker, so I was kind of more interested to see how the power profiles handled CPU loads from both applications (and the differences between them).
It's kind of irrelevant anyway because the results I think speak for themselves regardless: The quiet profile is substantially less demanding in terms of power consumption at a similar level of realistic performance, unless we're talking GPU, in which case it's probably about 20% lower. For that reason, it's probably safe to expect substantially longer battery runtimes under typical work day conditions for the average office user when set to Quiet.
I'm going to put this to the test in the real world and see how it turns out. So far, I've been getting somewhere between 9 and 10 hours on Balanced + Optimized, so if Balanced + Quiet provides comparable performance at 20% more battery life, I'd say that's a huge win. It's closer to the defaults Lenovo and Microsoft seem to embrace on their machines, which maybe helps explain the efficiency discrepancies between these devices after accounting for battery capacity differences.
Also, the consumption peaks around 55 W. It does not run at 55 W during the majority of the test; it's much closer to 17 W sustained.
if android phone is benchmark, mediatek will loose around 20-50% in performance with half of the price, so I will wait 😂
HwInfo parameters so wrong,Hw Monitor Arm64 cpu gpu clock speed and all power metrics so thrutly
I'm not sure I completely understand what you're saying, but if you are saying that the metrics are wrong, maybe they are. It's hard to tell, but there is no reliable way to measure power consumption on these SoCs yet directly (at least, not to my knowledge). Notebookcheck has run some recent benchmarks using external monitors in an attempt to get their arms around this, and I'm sure full battery drain tests are probably the only *truly* reliable way of discerning just how it affects runtimes... but even still, you have to admit, the delta between the different power settings *must* suggest some sort of adjustments to TDP/power provisions to the chipset. That's kind of the entire point here.