LG Gram 2021 announced: 14, 16 and 17

I didn't play with any prochoot things. Just performance cooling and got 80% more on Cinebench 23 than my old Gram which is totally consistent with Notebookcheck's comparative results in R15.
The expert reviews are approximative, they don't mention ANYWHERE about LG control center, first they did the 16 and then the 17 and probably they discovered at the 2nd attempt the performance mode.

Try testing again when you get the new laptop.

 

Try cinebench r20, impossible to get those results. Same for the expert reviews gfx car chase bench. My CPU was running exactly at the thermal limits decreasing to 22 watts after 1 minute, to get the scores they had it would have needed to run around 28 watts for 2 minutes.

The 16 is currently not available at any Amazon store for decent price. In Spain it's listed cheap and listed as available from 30.04, but before that was 23.04 while the 17 is available right now with discount. If the 17 actually delivers better battery life, 50 nits brighter screen plus better contrast, it's putting the 16 inch to a hard spot.

 

As I already wrote do not trust the expertreview. Here is a nice comparative review where it shows similar processor power and slightly longer battery life for the 16.



And another review of the 17inch. Performance results are not different from Notebookcheck.



To sum up: The internals are very similar, any difference in performance in the tests you've read are probably discrepancies in the testing procedure as I already mentioned. The 17inch will get slightly worse battery life due to the bigger screen.
So the only criteria to make the choice are actually screen size and portability/size. And price I guess.
I have downloaded also the Cinebench R20, I will launch it later and post the results also.

 

For all of the faults notebookcheck has (sometimes incorrect info.. etc..), the one thing I really like is that they

1. let cinebench run in a continuous loop and post all the scores so you can see long term performance
2. Actually state the PL2 and PL1 power
3. Show screenshots of CPU-Z and HWiNFO so you can see the reported wattage to be sure it's 28W.

I'm also shocked from your previous results that the cooling can actually handle dissipating nearly 28W continuously. I think the vents on the bottom this time around are really helping a lot. I can blow air into the vent under the hinge on my 8565U unit and it does help the power level be able to be maintained at a little bit higher state. The capacity of the thermal solution was the Achilles heel of these previously.

 

yes I agree notebookcheck has the best reviews - but the 17" having higher perfomance than the 16" isn't the first review we have seen. Plus just sometimes (not in all) numbers for the 17" which are impossible to replicate on the 16". Hence I would guess different cooling.
Then the battery runtimes - I do guess lg changed the display, and is producing a higher quality at least for the 17" now - or there are several models that is even consuming less power. The 16" may or may not be updated too (better contrast, brighter, better calibrated).

About the cooling - well I'm not really convinced except it being super quiet. My T480s has no problems at all cooling 28W with a single cooler as well and a vent that is no bigger - and only reaching about 70-75 deegrees. Mind it blows out the air to the side instead of bottom - that is a little better IMHO - especially if you work with the laptop on your lap. Suck in air from below, blow it out to the side. it reduces space for hte ports however... (but then same thickness, about same amount of ports if not more). Heat going out to the side is great - because the heatpipe on the way there will of course dissipate some heat too - so it's much more ballanced heat wise vs the super hot spot around the middle to the fan exhaust on the Gram.

My gram 16 can after 2 minutes only cool about 21watts of CPU, or ~26 watts for CPU/GPU combined with the CPU throttling. Reaching the threshold (at 20° room temperature). And that is with performance mode. Mind that performance fan mode is still very quiet. My T480s gets maybe 10db louder (twice as loud) but never burns my lap... The lg gram cannot be used on the lap for longer periods than 20-30 minutes (at 25° room temperature) surfing with Chrome. With Edge it's okay - just my mileage.

 

My 2019 can't sustain more than 14-15.5 watts without blowing air into the back to get some flow, which is not something you're going to do unless you're just doing it to see if it can actually go over 15W sustained. But I will say one thing, it's super quiet and due to no vents on the bottom you can use it on your lap all day with no issue. Tradeoffs I guess. I never use it on my lap though, I used to once in a blue moon. So I don't really care about that part, I'd rather have the vents.

With Tiger Lake H45 with embeded TB4 afoot soon though, the gram line is essentially dead to me.

 

well I don't feel the H45 is an option for lg gram users. I'ld rather guess next generation U processors will also increase core count. A big.little concept however would be way more needed - as I've noticed so far tiger lake really uses too much energy for office stuff or browsing. It didn't improve over Kaby Lake Refresh in that respect, or likely even got worse. Then a main reason for me to exchange laptops has always been new video codecs not implemented, most serious work I offload to servers anyhow. So AV1 decode of tiger lake is essential to me for the future vs older or Ryzen 5000. I realized Ryzen 5000 likely doesn't support that yet also on the mobile processors (same for Apple M1).

 

The Gram 16 has the air vents in the back under the hinges like the previous generation and a small opening underneath, probably to draw in air. As you can see in the photo the opening is really small, the rest is decorative. I find it that it can get a little bit warmer than the previous generation, but nothing compared to the X1 Carbon. The silence of the Grams is unparalleled, especially compared to my old Vaios or even the Thinkpad. The boost in power while maintaining all the advantages of the previous generation (same long battery life despite the bigger screen/higher res, silent, etc) is impressive. And I consider LG having done an amazing job in actual user friendliness and efficiency that is beyond the numbers in the various tests. We can continue debating about actual battery life, consumption of the Tiger Lake etc etc, but truth is, it can easily last for more than 9 hours for productivity, it is super light and silent. Only the X1 Carbon or the Nano could be competitors but their screen is much smaller. The Dell XPS have horrible keyboards, are loud and very often have quality issues.

 

Here is the Cinebench R20 Benchmark. It is even better than what Notebookcheck got in the convertible. I had to run it a 2nd time to get this, the first time the laptop was hot after working for 10 hours.
This morning I put it in flight mode, disabled antivirus and of course performance mode. Hope it helps.
The old gram was throttling a lot sometimes down to 400Mhz, I never saw this one below 1200-1500Mhz
And I don't believe that there is any actual difference in the internals between the 16-17, they use the exact motherboard for economies of scale reasons and most tests show this. Still the difference with the previous generation is worth the upgrade if you require more power but still need a silent and light laptop.
.

 

Oh I never saw the vent below the hinges. So with the monitor all the way back (e.g. on your lap) the vents are blocked, while if you have the monitor angled more towards close - they are opened further. Maybe the 17" has 1mm more there - the 16" has the best screen to body ratio - it would make sense if the 17" has a bit bigger opening. And well, I would always consider benchmarks are started from idle, not from cold laptop...

Techadvisor also has lower scores for the 16", including lower battery life!
www.techadvisor.co.uk/review/lg-gram-17-2021-3803169/
vs
www.techadvisor.co.uk/review/lg-gram-16-2021-3803093/


If the openings are bigger on the 17", maybe the vents need to work harder on the 16" shortening the battery life? Or the 17" has a more efficient display panel on newer version. Or both..

"I measured various aspects with a SpyderX colorimeter and the gram 17 has a solid maximum brightness of 406 nit (marginally higher than its predecessor). It also offers excellent colour reproduction with 100% of both sRBG and P3, then an impressive 89% of the Adobe RBG colour space. "
vs
"In our benchmark tests, the Gram 16 scored 100% accuracy when it comes to sRGB and DCI-P3 colour spaces, but a lower AdobeRGB score of 77% means creative professionals may want to calibrate it before using it for work. Also, with a maximum brightness of 310cd/m2, it’s not the brightest screen I’ve ever seen, but that’s not too noticeable in everyday use."

that's again a difference not explainable except if different technology... 24% brighter for the 17 and much worse on the Adobe RGB colorspace.


But yeah - the way LG botched up some software things - reviewers have a hard time. E.g. as long as you have that DTS crap installed - often not usable whyever - you do not get full loudness. Uninstall it and the speakers get 15db louder or so. I guess they reserved that in order to not have distortion for the effects when people also slide the equalizer on the lows all the way up.

 

Yeah, I want something with more than 4 cores and TB4, just hanging around to see what H45 laptops come out this year as my next upgrade. I'll probably keep the Gram around but it probably won't get much use once replaced.

 

Techadvisor gives a score of Geekbench of 4668 for the 16 vs 5187 for the 17. Expertreview it is 5516 vs 5079, everything is higher for the 16 except the GFXBench. Again I don't trust these reviews. Any conclusion on performance between the 16-17 cannot based on these mediocre reviews and I believe, since the internals are the same, that they cannot be significant. The different screen can be a valid issue, affecting the battery in favor of the 17.
Now I provided also the R20 score that you were asking and I got better scores than Notebookcheck that you said was impossible. So please when you get the new laptop, try again the benchmarks because I am afraid you were drawing conclusions on a defective one.

 

Actually I was thinking the same, I really need TB4 so AMD was out of the picture. In terms of performance I was at the limits in my workflow with my old Gram but it was still ok. But I was convinced here in the forum to give a try for the 16 and I don't regret it, the upgrade in speed is impressive. For my daily work plugged in the eGPU I don't need to use performance mode, already optimal is sufficient and I notice the differences, absolutely no lags any more. So I guess I can stick with the 16 for the next couple of years, since I don't expect my computing needs to increase that much in the near future.

 

Yeah I would like the 16 one again, but those screen differences have shown in every review. And according to the review 50-100 nits higher, better colors (though the 16 isn't bad and I don't need perfect accuracy, but wider color spectrum is always nice), coupled with less power draw plus right now in Germany virtually identical pricing for the 16" and the 17" have me a bit leaning towards the 17" now. Especially the brighter panel which will make it much more usable outdoors. Indoors the 16" is bright enough. I mainly use 80% during the day and 30-40% at night. If on battery and no charger for longer time I could instead do 10% and 50% (but anyhow as I tested the main difference in power draw comes from 80-100% - the lower values don't make a huge difference).

The 17" is about 10% more screen surface, if they managed to put a 25% more efficient panel inside (and in every review the 17" clearly is brighter, plus in general better contrast/accuracy) make me leaning towards the 17" now - even though I really prefer 16" form factor for portability. I would hope maybe the 17" keyboard is spaced a bit wider? Or is it identical because they crammed in the additional row on the side... Really could do so much more without that stupid numpad on the 16" and instead a proper enter, backspace and delete button. Those being so tiny often slows me down typing on the 16".

I still feel your way of testing from cold boot is a bit unfair. My CPU idles at 40-43 degrees (14-17 degrees above my room temperature). So from cold boot clearly you score higher. I will try now monitor fully open vs partly closed and look at the difference. Let's assess what this cooler actually does...

 

Starting from idle - screen closed close to it switching off:
1822 points. At the end of Cinebench R20 - 22 watts.

Screen vertical. At the end 21.5 watts.
1790 points

Screen fully open - Performance Mode.
1680 points. About 20.5 watts towards the end

Screen closed close to it switching off - silent mode
1660 points. About 18.5 watts towards the end of the test.
Screen vertical - silent mode
1620 points. About 18 watts towards the end of the test.
Screen fully open - silent mode
1550 points. About 17.2 watts towards the end of the test.
Edit: add while charging.
1460 points. About 16.3 watts towards the end of the test. Silent Mode fully open. So that's 190 points difference just based on whether we are charging or not - and how the screen position is set. While silent vs performance only gives 140 points difference with screen in best position.

The good thing is - the intel i7 is not the most efficient at higher wattage - so the losses are smaller. So while screen position and charging reduces the cooling by about 26%, the performance is only dropping by 20%.


In general the screen position is more important than the cooling mode! If you have it open quite wide - it will be close to fully open. So that makes those benchmarks even less reliable. But the cooling mode isn't a big deal! Normal is somewhere in the middle. I guess to save battery just use silent cooling mode (though silent is not fan off). I would be interested to have a fan off mode - I guess it would still get 13-14 watts at the end. And the 17" may just have that bit more volume to take up heat to give it some more headroom.

I guess from cold boot I can also score 1900. Then flight mode and antivirus off could be the missing points. I did not enter flight mode, nor switch windows defender off and I would too reach 1953 points...

 

No only the geekbench score is higher for the 16". The 4k media score is lower for the 16 - 118 vs 122.
the gfx bench 112 vs 188 a huge difference. I could also only score 112 or so.
And the disk transfer benchmark also higher for the 17" - slightly.
It would not make sense for the 17" to perform worse than the 16".

However there is one more thing to really botch up your score - and that is charge it at the same time. The charging really increases the heat - so If you run any benchmark while the gram is charging - take a 10-20% hit! That and the screen fully open and you will approach real bad results. And I don't have a 80 or 100 watt charger. That would maybe get it to 30% decrease. Of my 3 - 65 watt chargers, the LG one is the heaviest, the biggest and the slowest. The Lenovo 65 watt travel charger is only slightly better. The compact, 120g heavy, noname china charger that I bought 3 years ago for it's light weight, small size charges the fastest according to batteryviewinfo. But the difference is around 39 vs 41 watts - so not big.

On my thinkpad there is no way to mess around that much with scores - the cooling simply has no problems with 28 watts. On a laptop that can only cool 21-23 watts long term - such things make a huge difference.

 

Just let us know when you get the new laptop how it performs

 

I am just disappointed that LG did not give a good screen to the Gram 16, mediocre brightness and pathetic contrast. Yes, the 17 inch is ok and people should go for it if they find the size manageable.

Notebookcheck says brightness of average 310 nits and contrast ratio of 929:1. This brightness and contrast is a regression from even the last year's Gram 15 which had average 341 nits and contrast ratio of 1307:1

https://www.notebookcheck.net/LG-Gr...t-15-6-inch-Laptops.530592.0.html#toc-display
https://www.notebookcheck.net/The-L...ith-a-great-display.481034.0.html#toc-display

Other premium laptops give 550-675 nits brightness, their South Korean competitor Samsung gives 675 nits which i have posted on the second page of this thread. LG Gram is a premium laptop and it is not as if we are talking about 500-700 dollar laptop.

I will still have to get the Gram 16 as it meets my needs and Xiaomi is not interested in selling premium laptops outside China with their wonderful HDR OLED Samsung display.

 

I'm pretty sure the contrast is way better on recent revisions of the 16 inch. But brightness definitely is lacking for a glossy screen. Maybe they just reduced it because they thought this way people are happier with battery life.

Haha they could add a brightness boost mode in their control panel that resets every day or reboot, like the performance mode...

But yes that's why I'm thinking back and forth about getting the 17"...

It seems my 16 from Amazon Spain that I ordered for a steal price isn't getting delivered anytime soon anyhow. And other Spanish retailers with cheap prices don't deliver outside Spain...

 

How are you sure that contrast is better on recent revisions of the Gram 16? What is this revision? Gram 16 was just launched. Give me some facts, figures and sources/links.

 

I'm just sure my gram 16 is way above 1000:1, but brightness is still only around 300.

 

How did you measure the contrast and brightness of your Gram 16? I will like to know what measuring devices you used and compare your figures with notebookcheck. I am looking for facts and figures behind your made-up claims.

 

Just Vs two other notebooks where I know the panel and contrast measurements from notebookcheck reviews. T480s with the awful qhd lg panel supposed around 800 and a Dell XPS 13 9370 supposed around 1100. Both have much brighter full black on full backlight. The XPS is also brighter on full brightness, but like twice as light on full black. The ANSI contrast with eizo test is also clearly better for my gram 16. No I don't have any professional tool for measuring it. But the differences were big enough to see in a darkish room

 

Ok, then i will stick to facts and figures in my post above sourced from notebookcheck measurements.

 

Well - I took some pictures right now (only have the T480s to compare). Full brightness black and white (the T480s is the small laptop).
The difference is huge... Black and white screen at full brightness. Clearly the gram 16 has better contrast. Both panels are LG... And sharpness is a big difference too - but yes well glossy is always sharper, isn't it. That's why I would love a semiglossy screen. Just enough to get rid of the worst reflections.

Well my T480s LG WQHD panel is awful, it hast a gamma of 1.4 instead of 2.2 (my gram is between 2.1 and 2.2) and the white is really not white at all. Colors are horrible. I think I lost out there on the panel lottery, that's also the reason for changing the laptop so quickly. I lately need to work a lot on laptop without my 4k 27" monitor that I'm usually using - so replacing the T480s with some other productivity laptop. First screenshot the T480s was in night mode however - the white isn't that bad.

 

Here is my Cinebench R20 starting from full powered off instead of idle - screen in optimum position, performance mode for cooling. That's the highest I can score without colder room temperature (26 degrees, but Prochoot at 97). As this is about 80 points higher than from idle, I do not feel this is really a fair result.

 

The scores are very decent, so I guess you had a lemon with the previous one. I have 22C here vs 26C, this must make a big difference. Anyway except if you render video 24/7 5-10% difference in the scores will not make a big change in real life usage. And please avoid using a laptop on the bed, it is the ultimate laptop killer, in dust and heat dissipation

 

No that's still the same one. I'm waiting to send back until I hopefully have received my new one.
Still don't understand how notebookcheck scored that high, except by increasing prochoot or also going from cold boot...

With screen vertical and from idle it's 1780 points not 1899. Those 1899 are kinda cheating on my gram 16.

 

Notebookcheck got 1948, just around the average for this processor, I don't see where is the problem. Cinebench obviously should run without any other processes running and the laptop not having overheated before, it is already pretty demanding.
Anyway if you were looking for a serious performance upgrade from the T480s the Gram is not the right candidate. The T480s has better cooling than the X1 Carbon which is a closer competitor in terms or relative weight. But compared to the previous Grams the upgrade is worthwhile. For me it is the perfect laptop. I need the numpad, I love the screen real estate, things that I cannot have with the X1 Carbon of my wife. And it is lighter and quieter.

 

This video sealed the i5 for me:
I am sure it's the better choice on the Gram. As fast or faster and cooler.

 

I'd get the i7 myself, just for the extra L3 cache and a little more iGPU oomph. Aside from that they are nearly identical if you can hold the same PL1. Quite honestly the variance in heat sink contact area and TIM application quality is going to make a bigger difference. Probably why it's a good reason to put some quality TIM on there like Gelid GC Extreme. But from what I've seen that only really makes a difference in idle temps, I think with equal quality heat sink assembly flatness and contact pressure + quality TIM we're going to saturate that heat pipe/sink assembly well before we get to any real differences. I suppose if you're hitting the iGPU + CPU hard and it's a CPU bound load (I dunno what load could possibly be CPU bound with that weak iGPU) could throw off a little bit of balance between the CPU and GPU and cause the i5 to sneak ahead just a little.

 

Well that cache and igpu is gonna need electricity all the time. And the gram should be configured to not use any single cire turbo over 4000mhz anyhow as it instantly overheats.
For videos it's the codec not the power of the igpu that matters. Video encode again the cooling as there is no av1 encoder. I don't ever play games.

I feel any cooling that can do over 27-28 watts will then mean that the i7 really gets faster. The additional turbo definitely only makes sense with good cooling and not caring for battery life. The power increase vs computing power increase is not worth otherwise. I5-1145g7 maybe a good compromise. Too bad undervolting isn't possible anymore. I'm sure it would save 5-10 percent power on demanding stuff and accordingly heat.

I do really hope that Microsoft together with Intel can still improve in tiger lake efficiency for doing light tasks. It's way less efficient than kaby lake refresh for them instantly increasing clock too much. Only video decode it does really great. Just quickly moving the mouse uses 4-5 Watts per hour additional and then it waits 10 seconds or so before going into deeper battery saving.

Running cooler is a big plus for me so it's i5

 

These pictures do not show the brightness and contrast numbers but one important thing i can see in the first photo(White full brightness) is that the LG Gram 16's colour temperature is very cool as mentioned in the notebookcheck display analysis:

"X-Rite colorimeter measurements show an overly cool color temperature out of the box and color accuracy subsequently suffers at lower saturation levels. Our calibration attempt brings down the color temperature to improve grayscale significantly."
https://www.notebookcheck.net/LG-Gr...t-15-6-inch-Laptops.530592.0.html#toc-display

We can make the colour temperature warmer via LG Control Center, can you play around with that slider and post some photos(White full brightness) at different warm settings?

 

yes - it's definitetly on the cooler side. But my T480s is too warm, and the whites are yellowish. But that screenshot was with night mode activated on the t480s - night mode is always way warmer as it cuts out blues.

 

If that's what's really important to you to that degree of picking it apart, you can just turn it down to 15W and never have to worry.

 

But you can't disable the cache and the additional graphic units. So far the few reviews point out that the i5 saves about 10 percent of cpu/GPU power Vs i7. So should mean about 5-8 percent longer runtimes on battery.
I really hope Intel and Microsoft can still improve the power handling. It's crazy you can kinda watch a movie twice as long as editing in text editor, surfing and remote desktop on the of gram. I get about 9 hours of work time on that 80wh battery (max brightness, but at min brightness that just improves to about 12 hours. That's no better than my first laptop ever in 2006 - with a 90wh battery back then. Watching movies however that would be a huge difference...
Can easily watch youtube for 14-20 hours on the Gram depending on display brightness and not going for more than 4k resolution.

This i7 is sadly efficient as crap for office work. Comet lake maybe was best in that regard, ice lake worst since ages for intel. I do think it's an firmware issue tailored for max responsiveness even with power save in Windows activated...it's fine if the CPU goes to 2-3 ghz in ballanced mode, but in power saver for such things it should never go above 1.5 GHz, if at all. 1ghz one core active should be plenty. For video playback it indeed is really efficient.

Besides the i7 costing at least 100 Euros more, more like 200 euros more. And yes - I think on battery it makes sense to disable turbo mode. On AC I would limit single core to 4000mhz, all core turbo to 3200. Any more overheats too quickly, and doesn't give noticeable improvements. Actually this increases benchmark results for any benchmark that is topping out on thermals. As any faster really uses loads of power and will be over after a couple of seconds anyhow. That way you keep it faster for longer.
Well repasting and adding another heatpipe would likely lead to big improvments. Looked back at the thread about 2019 lg gram 17 where someone did it. If you get an 20% improvement on heat transfer here that would make a big change (and also increase battery life with the fan spinning less/slower).


Oh yeah - and you cannot set the CPU to 15W with tiger lake anymore. You can only limit PL2 and PL1 as you like. But at 2800 Mhz all core it uses about 23 watts.
The only other usable limit is speedshift value - the higher you go, the lower that CPU clocks. Actually the windows power profiles do not much else than adjusting the speedshift value.
Intel doesn't allow much tweaking for tiger lake. It's all relative now and many many settings are just disregarded. Speedshift is ruling everything kinda - the exception being PL1 and PL2 and overall turbo duration or disabling it. Plus you can set Prochoot with Throttlestop. You cannot limit the speed at all otherwise. So the i7 without turbo is actually a 23 watt CPU not a 15 watt one.... The 15w is pure marketing. The U series should be sold as 28w - because that is what they really are, the turbo will go into 40 watts if the cooling can handle it... Over are the times when it was 15watt and turbo up to 28...
AMD doesn't stick too much to their wattage either, but much better than intel however with Ryzen being way more efficient at high loads (not so sure about low workload)

The i5 consequently will be lower power than i7.

 

Actually increasing the Prochoot doesn't even closely offset room temperature. Today I ran cinebench r20 from an air conditioned coffe place. I guess it's 20-21 degrees here.
Prochoot 98 (the max you can set without the laptop powering down) - 1981 points
Prochoot 90 - 1919 points.

With room temperature 25-26 vs 20-21 increasing Prochoot by 8 degrees is only getting you a little advantage seemingly, but can only offset room temperature maybe 30%. So a 8 degree prochoot change is about as much as 3 degree room temperature change. So my results are actually in line all around - I did not expect this behaviour. I would have thought that increasing Prochoot by 8 degrees should offset 5 degrees of warmer room temperature (most reviews are done at 20-21 degrees). Even though with increased Prochoot I constantly run into the Prochoot value.. (it will overshoot by up to 2 degrees sometimes - so you cannot go higher than 2 degrees offset. better use 3 degree offset - 98 instead of 90 will give you about 3% higher scores on thermal bound benchmarks - even though you kinda increase the thermal headroom of the CPU cores by 10%).
Another heatpipe would really be the best solution. LG really should have invested 10 more grams or so in a more efficient heatpipe on the 16" and 17" models. Not for actual performance, but lower temps and increasing battery life a bit.

Edit: playing with the values a bit more I managed 2050 points. The important change is to make kinda limit the processor to 28 watts by disabling pl2.
Disabling turbo all together will run at 2800 MHz and around 22 watts. If you make sure at the begining it only runs at 28 watts instead of the hard limit of 30 watts that of set, you will get higher overall score due to operating a longer time without thermal throttle.

It's still kinda crazy for intel to sell a 15w U processor with guaranteed base frequency of 2800 MHz, which will need around 22 watts to keep that speed. It should be called a 22 watt processor plus turbo. And of kinda designed the cooling to exactly manage those 22 watts at 20-21 degrees.

The max with good cooling would be 4100 MHz all core, but yeah the of is quite far away from that. Single core 4700. Other notebooks hit 40-45 watts on pl2. LG already took the smart move to set 30w due to the limited cooling. Decreasing pl2 to 0.2 seconds or so likely will give you best performance. But disable all together to get better scores for anything long term. Not sure how much it actually helps to have 1-2 seconds or pl2 to directly run into thermal throttle instead of better limiting power a bit more conservative and having 20-30 seconds before thermal throttle kicks in

 

We can make the colour temperature warmer via LG Control Center, can you play around with that slider and see if it makes any difference?

 

I have to see again, but I think there are just 3 settings, normal, cold and warm. If you want different profile you need to use .icg profile in windows or use the intel graphic center

As I said, I tried the .icg profile from notebookcheck, but without it on my gram 16 the colors are more accurate and better. There are definitely different revisions or versions of the panel, but they name them all the same. The panel name for the 17" is the same since they switched to 16:10

 

Okay - I just looked at it using a white paper for comparison - there is 10 steps overall. And yes moving the slider 1-2 steps towards warmer seems to give the most realistic white. But the colors don't pop as much anymore - and I feel it lowers the overall contrast a bit. One step towards warmer actually may be the best compromise for contrast and color accuracy on my lg gram 16. But yeah - as all tests have shown, if you want color accuracy you better get the 17". That one has aced all tests so far with 0.6-0.7 deltaE from factory plus 50% higher contrast, and 33% higher brightness (plus longer battery life). Now of course you can just hope they don't have a panel lottery which cannot be identified without measurement device or having to lg gram side to side. Or hope that they do in future put the same quality panel also in the 16".

Order:
full warm,
full cold,
middle,
1 step towards warm
2 steps towards warm

Problem - the white ballance of the camera changed - so you can only compare to the white of the paper, not each picture. I think 1 step towards warm gives me the best results in real life. I'm not even sure it makes any sense posting those pics - without setting a manual white ballance on my phone.


What LG actually should do is offer an RGB, DCI-P3 and Adobe RGB profile. Especially on the 17 the Adobe RGB would be fine enough - the 17" panel can show most of Adobe RGB too.
Right now the default seems to be DCI-P3 - meaning popping colors and good contrast/very good color separation.

 

Probably the best solution would just be to buy an X-Rite or Datacolor colorimeter and calibrate it.

 

Can some people post their idle cpu power package use?
After the newest windows cumulative update I don't get black screen with panel self refresh enabled anymore, but package power at idle is at 1-2w. Just moving the mouse and it goes up to 3w.

Before it would be 0.5w with panel self refresh enabled, which was impossible for me due to screen flashes.

Did the update just enable something to detect broken hardware or does everyone have so high idle power???

I now need around 9 percent battery for 1 hour of surfing in edge with screen at 80percent brightness. Or 7 percent with minimum brightness. That's way too much Imho. Video playback uses lot less power..

 

https://www.trustedreviews.com/reviews/lg-gram-16-2021

According to this Gram 16 review, contrast ratio is 1415:1 and 333 nits brightness. So, these massive variations is due to panel lottery and it all depends on our luck.

Another important information which we have been reading in different reviews is very cool colour temperature and this review proved that by measuring 7054 K white colour temperature. So, moving the colour temperature slider 1 step towards warm is the best choice, we just need a few days for our eyes to adjust to it after watching vivid mode 7054 K whites.

 

well 1400 contrast and 330 nits is what I would guess for my Gram 16 too (being clearly better contrast than my around 800 contrast / and a tiny bit brighter than the 280-290 nits T480s)

 

Samsung has launched Galaxy Book Pro 15 with OLED screen.120% DCI-P3, 1,000,000 to 1 contrast ratio, 500 nits, 0.2ms response time, and VESA DisplayHDR 500.

Samsung is also providing different viewing modes: natural (sRGB), movie (DCI-P3), and vivid (AMOLED native). LG should have also given such screen options as well as different modes.

Any forum members here willing to give Samsung Galaxy Book Pro 15 a try?

https://www.windowscentral.com/samsung-announce-galaxy-book-pro-laptops

https://www.gsmarena.com/samsung_un..._one_with_rtx_3050_ti_graphics-news-48862.php

 

well they are disqualified for me because only FHD resolution. Otherwise they would be very interesting. Especially that they have AMOLED. I could overlook the 16:9, but not the pixxely screen resolution. I mean even the lg gram 17 should have a bit higher resolution...

keyboard uses default driver (Standard PS/2 Keyboard). Trackpad microsoft precision default drivers (HID Compliant Mouse).

 

I would but I'm still holding out for at least a 6-core i5-11400H.

Using 16:10 panels as a reference, feel free to use 1080/1440/2160 in place for 16:9, IMO the 1600p resolution on the Gram 17 is perfect. It's enough to completely remove the ability to see the pixels at 1200p and better for performance and battery life at native resolution than 2400p. Obviously if my only choices were 1200p or 2400p, I'd go for 2400p, but I'm hoping we see more 1440/1600 options crop up. My #1 priority for my next is going to be something with more than 4 cores though, I'd gladly make a sacrifice and go 1200p to get that. I don't think I'm going to be forced into that though when all the H45 units start dropping June through August or so.

 

Well yes battery life for 2400p sucks because Intel cannot get it sorted out. Not because the display itself needs that much power. For the 17 I would like 3000x2100 or similar as long as Intel uses too much power for higher resolution. Not sure about AMD. I guess and 5600u would be better anyhow

Edit: my black screen issue is back today. Don't understand why it wasn't there yesterday. So I have to activate panel self refresh again. Well I hope my new 17z90p gets shipped soon. i5 option with 16GB RAM. I think that is the best option. Go for 17" because of hoping to get a better (mainly as in brighter) screen vs 16"

 

The maybe best comparison review of 14,16,17 inch. They did run quite a few benchmarks. Seems that super low contrast panel is gone from the 16". Recently all reviews were around 1300-1500 contrast. Still every review lately puts the 16" around 330-350 nits while the 17" gets 400 nits. I am pretty sure this is not panel lottery.

https://tweakers.net/reviews/8948/lg-gram-14-16-en-17-lichtgewicht-laptops-met-lange-accuduur.html

As for the benchmarks, a bit unfair to the only 1135G7 that is had the slowest NVME SSD, and only 8GB Ram instead of 16GB of the 17". I"m pretty sure the gain from the i7 is minimal, it will only be faster for singlecore because multicore it is power and thermal bound and will actually rather be slower, and the 96 IU GPU only burns battery vs 80 execution units. i5 and 16GB Ram is clearly the best setup (and you don't pay the i7 surcharge).


In that test however the 17" did have the lowest battery life at 180 nits. Not sure how much the i5 vs i7 played part in that however. The 50 nits more brightness will be good outdoors. I tried working on an overcast day outside on my 16" and gave up after 30 minutes. Way too many reflections and too little brightness. Without any anti reflective coating outdoors the gram will be a poor choice any model however. Even with 500 nits it would still be problematic. Their 17" screen had bad colors however (they noted that one step to warm gave best colors - in line with my feeling) while maybe newer bathes come calibrated? Maybe 14" and 16" in future will also come calibrated out of factory..

 

I'd still take the i7 as at the points where I need the most CPU power, during gaming, I am using an eGPU and disabling the internal panel, so the iGPU won't even be doing anything to consume any power. Of course if you aren't in that niche and don't care about the extra iGPU oomph take the i5.

And being able to quickly switch between the three photos of the internals makes me even more disappointed in all that wasted space that could have been used for better heat management in the larger units. I guess they need to save $$$ and get those mainboards and heatsinks in large bulk as well as cut down on design cycles.