*OFFICIAL* Alienware 17 R4 Owner's Lounge

Yeah, and no, it's not sleeping.

 

Is the bios updated to latest version?

 

Yep. 1.2.0

 

Is Rapid storage driver updated to latest one which was posted couple of days ago?

 

Yeah, everything is up to date. I figured out what was causing it, it's the Tobii Eye Tracker shutting off when the lid is closed. It never used to make that noise before when I closed the lid, though. It must be the BIOS update that "improved" Tobii functionality or the recent update to the Tobii Eye Tracking Core software.

 

Rei Fukai said: ↑
cpu layout is:

[0][2]
[1][3]

I have no idea if you are being rude but you must have notice i have no idea about computers, its my first computer that i own. i would be very happy if somebody who has upgraded only one slot of ram with the other shipped ram made it work. thank you very much

 

sir i quite frankly have no idea what you're talking about, maybe you could clarify, cause i do not understand you. IXVIXXII asked a question about the cpu layout, i reacted with how the cpu is laid out, and you are clearly talking about RAM (?) I have never reacted or quoted you in a post, that's why i'm not getting why you're feeling like that. as for a answer to your question, you can pair up different ram speeds, and models as long as it uses the appropriate ram Double Data Rate (DDR) technology. in this case it would be ddr4 in the SODIMM format. just take note that if you have 2 different speeds of ram, the ram sticks wil run at the highest speed the slowest can run, also can the voltage vary that they pull to get to their appropriate clock speeds (only important when overclocking the ram) but overall, yes you can pair up different brand of ram with different speeds, the only caveat is that the faster ram will only go as high as the slowest ram can run at it's full speed

cheers

 

I think the sound you hear is "windows fall update" related, because I'm also getting It, and I rolled back to 1.0.8 after trying the 1.2.0.

It must be either the screen or the Tobii camera shutting down.

 

Help here! What are the best OC settings for 7820hk at 4.3 and 4.4 GHz?

 

What happened? It seems FCU borked the CPU OC settings, correct?

 

This has happened ever since I got my machine, but does anyone else experience the Windows notification sound only coming from the right side of the laptop for the first quarter of the sound and then regular sound through all speakers for the rest of it? Odd problem.

 

Core voltage offset is in the bios at 120 clocks are from cpu 1 to 4 4.4 or 4.3

At 4.3 I can leave the core voltage offset at 0, and undervolt it till -0.45 in throttle stop, but overall 4.3 should handle from the 1.2 to 1.25 ranges, and going to 4.4 should be around 1.30 to 1.4 Volt Wich is pretty high. The 7700k uses less V to get to 4.4 or up (at all 4 core's). Currently running a 4.4 all 4 core's myself with a Max of 83 on my hottest core's.

 

Ni @Vasudev I ask because since the last update of Win10 something changed, because with the same OC settings used, now the CPU works unstable ... For example, in 4.3ghz, when I execute a stress test, the frequency drops from 4.3 to 4.0, 3.9 or 3.5 and then go back up to 4.3 ... and so on all the core... (btw the power plan its ok)

I found these errors in the Event Viewer:

The speed of processor 3 of group 0 is limited by the system firmware. The processor has shown this reduced performance state for 71 seconds since the last report.

The speed of processor 6 of group 0 is limited by the system firmware. The processor has shown this reduced performance state for 71 seconds since the last report.

The speed of processor 5 of group 0 is limited by the system firmware. The processor has shown this reduced performance state for 71 seconds since the last report.

There are a lot of these messages ...

I use this OC settings...

For 4.3GHz
1-Core Ratio Limit Override: 43
2-Core Ratio Limit Override: 43
3-Core Ratio Limit Override: 43
4-Core Ratio Limit Override: 43

Core Extra Turbo Voltage: 0

Power Limit 1: 230000
Power Limit 2: 230000

 

I think the virtual HTT cores are downclocking themselves due to no load. Are you using latest TS 8.5?

 

TS 8.5???

 

ThrottleStop.

 

ahhh I'm sorry, I'm asleep ...

Yes, I have the latest version, but I'm not using it. I do OC from the BIOS ...

 

if you're using the latest FCU, it could be due windows using speedstep to sense if your cpu needs to ramp up its speeds and how far, you see at first windows was telling the cpu to boost how it should boost, cause of a task windows would be performing, now with speedstep the cpu decides how far it boosts. it will run the application at full speeds first, and sense it can complete the task with a lower clock speeds, and clock itself down.

try to change your ppowerplan to high performance, you could do it by right clicking your battery in the task bar. go to mobility center, and then put the powerplan to high performance.

 

I am using the latest version of FCU and the high performance energy plan. Modify some things, such as "system cooling directive" to Passive.

Even with that power plan he continues to do the same, but a little less.

 

cooling i would let it be active, passive is when it doesn't have fans or anything, thus relying on the outside temperature to cool itself down. things like mobile phones, tablets etc have a passive cooling. our laptop has a heatsink that should be cooled (when pushing your machine) so when benching the cooling policy should be active, thus giving your fans the commando to keep spinning regardless of the load you have. but personally i do not think that's the problem, i think its windows acting up again. try resetting your bios, perform a battery drain (30 sec pressing the on/off button), and then see if it helped. it could be your bios not liking the current windows timing, cause it got used to the old timings (and the way it polls your cpu) or what also could help is, clearing your CMOS battery, by removing it.

 

yesterday remove the CMOS battery ... so that's not it either ...

 

This settings on hwinfo are ok?

 

dangit, that couldn't be it either. i do have noticed some differences between our settings, but i'm not sure it could impact the core stabl=ility priblem but you could give it a try.

 

this laptop was always used connected to the graphic amp ... in fact, never booted with the iGPU ... I will try to start with the iGPU to see what happens ...

 

I give up ... I'm sure it's a Windows/drivers or BIOS problem...

 

i'm still brainstorming what it could be, but i can't seem to come to a solution. last resort would be a clean install of windows. it could be a registry value that's causing the issue. i'll keep googling maybe i'll find something

 

@QuillP Reset HWINFO preferences.

 

Sounds like it’s best to skip the W10 FCU for now.

 

i think it's best to download it, and see for yourself if it would impact your experience. i personally don't have problems anymore (did a upgrade, then resetted my windows) and even before resetting i could feel it has become faster for some applications. but i think it depends on what programs there are on your Laptop,and how you use your laptop.

 

Im sorry i thought it was meant for me, and it stopped me in my tracks. thank you for your help and sorry for the inconvinience i have decided to upgrade it wiith crucial guarranted compability ram option same speed. it just so expensive that i cant afford to make a mistake so i have checked quite everywhere but nowhere i could find explicit information about getting ram to work perfectly with different brand. the one in my laptop with crucial for example as i cant afford to buy a pair. its impossible to find the exact brand and model in my laptop which i found frustrating. thank you very much and sorry for the misunderstanding.

 

Does somebody know if i should pair my ram which is single rank with another single rank or i should pair it up with dual rank.
thank you for any replay

 

If you can find or buy the exact same RAM stick then go ahead. Otherwise get pair of Kingston HyperX/G-Skill RAM sticks.

 

thank you for your replay but i have a budget problem i have the 100 euros that 1 stick of 8gb cost but i cant get the 190euros for the Kingston HyperX/G-Skill costs. it just too expensive for me to pay right now. So i decided with crucial but they have 2 version one is single and the other dual rank and i have already one single rank 8gb ram installed im just not sure which one to buy they are the same price maybe you can help me with this and if it is possible to do this. thank you

 

Yeah Crucial RAMs will do fine. Just match the speed and you're good.

 

So 1x 8gb single rank with 1 x8gb single rank is ok to be paired up?

 

So 1 x 8gb single rank with 1 x 8gb single rank is ok to be paired up?

 

Yes you can just make sure they're of the same make and timings to get dual channel mode. Can you post your RAM model no using HWINFO64?

 

Here is the info

 

Hi guys, thanks for responding @iunlock @Rei Fukai @alexnvidia . The stutters are gone after modding the HS, but that is not over. There is another big issue that I though was related, however even with good pad fit and CPU/GPU temps it is still present. I'm suspecting the motherboard having faulty power/voltage delivery components.

Here is how it shows:
Whenever I play a demanding game the performance is good unless I turn the view onto something more challenging for the GPU (a view of a larger portion of the map in BF1 or looking at a city when driving a car in PUBG). This instantly triggers VREL warning in GPU-Z even when the system is not really hot yet. The GPU load drops to 60-70% from 95-99% giving a large FPS drop that remains until I change the view or enter a different area in game. The same goes for VREL warning - not a short trigger like @alexnvidia described in a YouTube interview but it's basically showing VREL for most of the time when playing. The VDDC caps at about 1.06V but drops to 1.045 or 1.05V when I witness this performance issue and this always comes while VREL reason is on.

I have updated the BIOS to the newest possible, reset windonws and did a clean GPU driver installation with DDU, rolled back the BIOS to the original version but nothing seems to help. Interestingly this does not occur when stressing the GPU and CPU with synthetic burn tests PRIME95 + OCCT GPU test, only in demanding games - and very frequently.

Long story short - I only get full GTX 1080 performance when looking at walls with closed windows in games, in other cases permament VREL warning kicks in = GPU load + GPU voltage drops = plenty of performance lost.

 

VDDC dropping is an indication the GPU hit it's throttling limit so it steps down to the next available lower voltage and frequency boost. this can mean several things, your GPU is too hot (thermal throttling in pascal is quite conservative), your GPU hit power limit (180W) or the game simply can't utlitize the GPU properly. what was the GPU frequency when this happened? Vrel warning can happen even without any issues, it's indicating it's hitting some kind of throttling limit and it's reducing the voltage. i use it in conjuction with gpu load dropping to zero to identify the stuttering issue which i shared with steven from @OWNORDISOWN .

if i were you, i wouldn't worry too much about what you are experiencing now because it does not affect the game anyhow, maybe reduced fps but it will not stutter or lag or lock the game. I would suggest you to try to undervolt your GPU with msi afterburner and check the guide from @OWNORDISOWN youtube channel. this can help stabilize the GPU frequency allowing it to sustain its boost frequency and reduce power consumption considerably.

good job btw on resolving the gpu stuttering issue. i'm seeing more and more people are able to resolve this by ensuring proper heatsink contact and cooling. most recently @Pete Light was also successful at resolving gpu stuttering. you just need a lot of patience and top notch attention to detail but the end result is worth it.

 

Thanks! Wouldn't have made it if not for your, iunlock's and others huge input in this field. I had to open it multiple times, but I enjoy working around tech. I used ARCTIC thermal pads that are rated just 6 w/mk but still seem like they are much better quality than the stock ones (e.g. higher density, more sticky)

As for the GPU issue, the temps are really good, it hits about 66-68C after gaming sessions, the core max clock is at 1860-1880 MHz but it drops to about 1760 - I don't have the exact measurement because it prolongs over a greater period of time and the clock values always fluctuate a bit so these are average values. I also checked the power limit warning and it kicks in sometimes but it is clearly not connected to the performance drops and I as you said I probably shouldn't bother but these are significant drops, I'm talking 20 to 50 fps below what it should be, my friends GTX 1070 does better and mine has a 1080 inside :< btw the GPU voltage caps at 1.050V permanently, I never saw it go higher even for a second, it only drops even lower to around 1.040V.
Unfortunately I don't have screenshots on me right now but I always run GPU-Z or other utility programs. The VREL blue mark in GPU-Z covers almost the whole PerfCap Reason graph constantly for very long periods of time, PWR warning just shows up rarely in very short tiny stripes - 0.5 to 1 second duration.

On the side note, I used ThrottleStop a few times to check other possible Limit Reasons and it shows BD PROCHOT warning flashing red for CPU and GPU as well as EDP warning for Ring. Heatsink has good contact with all elements and the PCH max temp is 66-68C (dropped from 84C max just with proper use of a cooling pad!) so why BD PROCHOT? And Electrical Design Point? This does not occur exactly alongside the performance drops but fairly close. I run the 7820HK at 4.2 GHz, temps are decent as well (78-82C on gaming load - Grizzly Kryo), however even after returning to stock CPU settings in BIOS and XTU nothing changes.

EDIT: I have Afterburner since day 1 but I can't change the power or voltage of the GPU, does he show how to do it? I will take a look, thanks a lot for letting me know.

This really stays on my mind day and night ;D I hate unsolved mysteries, especially in this field.

 

More correct:
Quality bins:
1) 7700K.
2) 7700 (fail bins for 7700K).
3) 7820HK (fail bins for 7700)
4) (not sure, I think a UV chip here).
5) 7700HQ (lowest tier).

I do not know where the 7820HQ or 7920HQ fit into this.

 

18xx sounds about right at stock voltage and power limit. even at 1760, you are boosting at near max, which is 1771MHz. 68-70C is honestly on the warmer side for GTX1080 in alienware 17 R4. the cooling system is capable of performing miracles, so much so i would say this is by far the best cooling system for gtx 1080 in a laptop, miles ahead of msi and their overrated cooling heatpipes. the only downfall is the heatsink manufacturing tolerance which caused poor contacts but this mostly applies to CPU side.

i have tested my laptop with witcher 3, maxed out settings, temperature has a hard time hitting 60C over 1 hour gameplay, that's with undervolt and overclock to 1.9+GHz. without undervolt, getting past 65C is going to be a challenge for this cooling system. it's really that good. MSI with their heatpipeS party and really good contact net u 80+ on the GPU, close to what desktop FE offering.

try testing your system with another graphics intense game like witcher 3 and see if similar behavior occur. for GPU undervolt, like i said earlier, refer to @OWNORDISOWN youtube channel for the guide. i use TS to undervolt my CPU only and i dont mess with BD PROCHOT and other settings. use MSI afterburner to monitor the gpu stats in game.

also dont go overboard with the thermal pads as it will raise the heatsink significantly causing poor contacts with the GPU or CPU instead.

 

Is the backspace deleting characters really slow or is that just me? I end up just Ctrl+Aing instead of holding down backspace. Here's how long it takes in realtime
https://i.gyazo.com/c906c03b7dd0d7c6ff534e39b58ad085.mp4

 

Intel Core i7-7920HQ is the BGA KING and 7820hq is the 7820hk's sister

 

On the German AW Website they offer a 4k, 400nits Display with Gsync as an €150 upgrade to the 400nits, QHD display most people are using. So I had high hopes, that AW finally started using the famous 4k IPS panel from AUO, which received great reviews in many other notebooks, such as the 17inch Predators or Aspires VN7.

Checking this on the US website seemed to confirm my thinking.

So went back to the German configuration, however it was saying that this UHD/4k display is TN. Hmm, I was not aware of any OEM using 4k TN panels with Gsync on gaming laptops, so I contacted the sales guys from AW. And after a lot of discussion they confirmed it is TN. I do not trust them to be honest, as a) I do not believe AW is building different machines for Europe and b) I am not aware of such a panel.

Has anybody some more information on this ?

 

What the hell. 4.1ghz 45watts

 

Ive ordered the a17 r4 with gysnc 4k ips panel and will confirm what panel it is in the next 2 weeks. I suspect it is the AUO panel as that is rated at 400nits - very similar panel to the Alienware 17 r3 4k panel and some people received it on the earlier Alienware 17 r4s.

 

Piece of crap processor. 45W TDP 7920HQ will instantly throttle at full clocks in any game that uses all 8 threads. And with the higher 4 core active clocks, undervolting may not work like it did on the 7700HQ. You could keep the 7700HQ, at its maximum 3.4 ghz, under 45W at all times under the most extreme loads (AVX small FFT prime 95, 8 threads) with a -100mv or more undervolt, preventing all power limit throttling. 7920HQ would reach TDP rather easily in Battlefield 1, GTA5, etc.

 

Yeah, the big BGA brother is better binned.

Undervolt may not work like it did on Intel's worst Core mobile core i7? Better binned, means possibility for high undervolt. Regarding P95. It's a reason Kaby come with AVX Instruction Core Ratio Negative Offset Intel doesn't state in their data sheet, that their processors can hold max Turbo in fully AVX instructions load.