*OFFICIAL* Alienware 15 R4 Owner's Lounge

Yep, that's true if just the PL1 is the limiting factor.
More runs would be useful only if the thermals or other limits kick in after some time and the score degrades even more, otherwise even 2 runs would be enough (since the time window is shorter if the power draw rises v fast and tend to surpass even the PL2 limit).

 

You have also a second problem to take in account with thinner notebooks, trash cooling, ****y thermal paste, uneven warped heatsink etc... The Power consumption rises with temperature (because of I2R losses). It will always use a bit more power if your problems is as showed above.

 

Look at the current voltage. -200mv Cache and -180... your current voltage is 20mv higher...

 

And the cpu package power is lower for -200/-200.

 

Did you read the numbers as the Devil on speed?

 

Yes in TS since it varies during the bench. Look at the max voltage numbers in hwmonitor

 

max voltage is idle -> unimportant
current voltage -> load -> important

 

No of the voltage values are from idle, i started hw monitor each time during the test.

 

We talk about LOAD VOLTAGE. As long you don't have a proper desktop board with better voltage facilities settings you will get voltage spikes or strange volt attitude.

See also...

 

Btw elo captn, Leto here, not suprised to see you here haha

 

@Papusan I cant say I can read the graph well. Can you explain, why max voltage under load during the bench in hw monitor is lower on the -400 core / -180 cache test then the one on thee -200 / -200 test?

 

Re-read numbers in your pict.

 

-0.200mv core -0.200mv cache Undervolt:


-0.400mv core -0.180 mv cache Undervolt:


And don't forget the Monitoring tools Polling period frequency

 

and i already told you in the german forum that your unequal uv is not working...

 

Maybe I am dumb but from what I read the voltages are lower on the -0.400 core one? Would be rly happy to understand it at this point lol

 

Read Current Volt and Cpu Package Power. We don't talk about voltage spikes or idle volt. Compare thegraphs on left side. Higher Voltage = Higher Power Consumption.

 

Hm k I see. Still as mentioned before I restarted hwmonitor each time during the bench, there are no idle values, the system was under load. Will do some more testing. Its still undeniable that its the cache undervolt that mainly drops the temps.

Unfortunatly I cant run at -0.200 for long, it gives me huge temp improvement but its not stable anymore :/

Btw would you advice to run the system with 92C spikes after hours of gaming? Not really looking forward to do a LM repaste on a 2 weeks old system, but seems like its the only option atm.

 

btw... -400 would result in 0.7v load voltage if it would work...

 

Remember the much lower temp you can get... Often means you can put a in a better undervolt and better stability. Double gain for locked down processors. I would hate see temp above 80C on nice oc'd processors. Above 90C I would puke And None locked down Cpu's should reach that sorts of temps.

 

@captn.ko from what i see it reflects some middle number between -0.180 and -0.400, in this case its approximate -0.250 on the max voltage as seen on the screen

@Papusan I do hate the numbers. Unfortunatly I am stuck with this machine for now, regrets and so on, but I have to use it, since not getting another exchange from dell in forseeable future. The thing is I wonder how safe it is to run those temps. Average around 75-80, spikes up to 92 with current undervolt and no LM repaste.

 

@CanIbiteYou

The voltage under max is just the max registered value (it can't be correlated with the other values since it could be registered at a different time/load) and it drops with load, so after it reached the full load, it stabilizes at a lower voltage .

It could be possible that the initial voltage (right at the start of the test) is lower due to the high core offset, but once it is in max load it stabilizes at the proper value (view again the graph that Papusan posted or wait a minute until I find the same graph that contains the offset).

 

"It could be possible that the initial voltage (right at the start of the test) is lower due to the high core offset, but once it is in max load it stabilizes at the proper value"

Ye that would explain a lot, guess this is the case. In the end its a very minor difference (maybe placebo one lol) between 0.400 core / -0.180 cache and -0.180 core & cache Undervolts.

 

Buy a notebook cooler. Average 75-80 is good. If you can live with the 90's C spikes who is a awful numbers, Oh'well.
Don't forget that bad temps in the beginning only means it will be worse with time. Good luck.

 

Yes, dont wanna know what temps will look like when the current toothpaste dries out. Lm repaste it is, some time in the future.

 

Thank you for the info and the link. I have followed your instruction and my results were consistent. 1208CB minimum and 1246CB Maximum all other results were in between these numbers.LM this week (again) and Fujipoly 17 (this time not artic pads). Really reinstalling this new AWCC clearly made a big difference in system stability and performance.

 

So I got my new Alienware exchanged system in today. Tested right out of the box using XTU. Didnt any thermal throttle but boy did I see power limit throttle. Sometimes it would just stay throttled. This is without undervolt. What should I do and what causes power limit throttling?

 

I'm due for my exchange too.
Following a heatsink replacement and undervolt - 200mv and cache - 120mv.

The temps still spike but the power limit is throttling.
Hopefully the new ones better.... Will be my 3rd in 3 months.

I also had gt fuse limit exceeded... Whatever that means.

 

In the bios set CPU performance mode enabled. This should raise the power limits and avoid power throttling. Temps will rise more although.

 

Got mine a week ago. Runs in the mid 90s while gaming stock. I can tell in games that TT and PL throttling cause stutters in gameplay. I promptly installed TS and set max multipliers to around 32 and SpeedShift to 64. It basically parks at 3.2Ghz on the 6 cores while gaming in the low 80s. GPU runs perfect at same temps. Games run perfectly smooth after this. I'm also adoring the "quiet mode" as I can play a large chunk of my games with much quieter fans.

This thing was clearly not the best paste job. Once I get home from deployment I'll probably repaste. That is, if I don't return it for the freaking AlienFX lights acting stupid.

A few questions for other R4 owners.

1) Do your keyboard lights auto dim or shut off when you aren't using them for a while (I.e., the behavior of most laptops after a power saving time out?) - mine are on all the time, even on battery, and even when the display goes black after the set timeout.
2) Does anyone else have issues with a specific light theme getting "stuck"? I.e., it no longer responds to any commands from the AWCC software? It will say "theme applied successfully" but no lights change. I also cannot use the AlienFX key (F12) to turn the lights off when the lights are stuck which is even more of an issue to me.
3) Are the macro keys supposed to stay lit at all times, even when you turn off the AlienFX entirely? Mine are on all the time PERIOD - nothing turns the macro key light off

I used to have an Alienware 13r3. I remember the limited color options, and you could never set the keys to all white. I love the new millions of color options, but I'm having a hell of a time with the lights. I remember my 13r3 would dim on battery power, and then also timeout after a minute (a threshold you could adjust).


I have clean reinstalled the AWCC software per tech support instructions, and yesterday I reset the whole device to factory image using the built in restore partition. All my issues persist. The lights are sometimes responsive to AWCC, and the F12/AlienFX key works. Then all of a sudden for reasons beyond me, it stops working altogether.

 

Also - does anyone know if ThrottleStop causes issues with Alienware Command Center? I've noticed a correlation with turning TS on, and suddenly AWCC no longer works to control lights.

 

No throttle stop doesn't cause issue with AWCC. AWCC needs to be reinstalled also OCcontrol. To achieve good results you must uninstall completely AWCC and OC control in the regedit also under Dell (not alienware) follow Dell guide on how to clean unisntall commande center. Also i recommend going in to services and selet automatique startup instead of delayed startup (excuse my english my system is in french). Do the same for XTU services. That is what i did and since then AWCC is running perfectly. No more issues. Also performance have increased.

 

Throttlestop doesnt cause an issue, but AWCC is quite buggy. If i try and use the F12 key it wont turn the lights off. If i load AWCC then do it - it works.
Sometimes when i open AWCC the lights and overclock parts dont load, a restart fixes it. It is tempermental.
As for the lights dimming - they dont. I asked a tech if there was a way to set them to dim if not in use or after 10 seconds - like my XPS does but unfortunately not. Im looking for that white colour similar to the backlight to the XPS but doesnt work

 

As i said before, it needs to be cleaned installed. All the problem you mention i had before i reinstalled. FN-F12 works without alienware running as a program (but it is running as a service because i selected automatique). Even fan profil wouldn't work half the time before i reinstalled. Now it runs as it should. Also when you mention that lights and overclock don't load, i had this problem too before reinstalling. This has to to with OCcontrol and XTUservice. So uninstalled both: OCcontrol and AWcc, leave XTUservices alone, and don't forget regedit delete. After that test your CB15 and you will notice a big improvment. i went from 1115 avg to 1215 avg after that. Max 1246CB with HWinfo turned off.

 

Mine is more than buggy software I think. I can now recreate the problem. As stated, I've done two full "clean" reinstall per Alienware instructions, and I've fully reset my device to factory image. I also have an issue where my lid lights (head logo and side lights) stay illuminated, EVEN after a shut down! They were doing this when putting the device to sleep as well.

Basically, I shut down, reboot, launch AWCC, and everything looks fine. Then I launch a game (Civ VI in this case), and within a few seconds of loading a game, the macro keys no longer respond to the macro preset toggle button, AlienFX button does nothing, etc. I've also had a few instances of the lighting colors changing to a random color not related to any theme I have saved after loading the Civ VI game as well. Then when I quit to desktop, go back into AWCC, the theme and lighting controls do nothing. The only way to get it working is full shutdown, unplug, wait 60 seconds and reboot, etc.

If this is just "buggy" software - then I don't want an AW15 anymore. Which is sad because I was excited for the 16M RGB lighting color choices - but everything else is a big setback from previous gen machines with the older software. I could even live with the lighting if I could just set a basic white key background and have it stay that color, but the random color changes in game is just too much.

 

I feel your pain. It is a really buggy software. Mine was never as bad as yours but it was always buggy from day one. When you follow dell procedure to uninstall, don't forget that this procedure is for old AWCC so you also have to delete: Hkey Local Machine/ software/ DELL/awccservice and occontrol service. This is what worked for me. Now it is running perfect for the first time in 3 months since i did this. Maybe your issue is different. I never experience light stayed turn on after shut down. But FX not responding yes. My guess is that the FX seems to install whith OCcontrol and i believe that OCcotrol install xtuservices as well. You can always call Alienware tech support for procedure to reinstall, they will try over the phone but if he can't then is going to offer you a new MB onsite tech. If it doesn't solved the issue then is going for a second MB and it's going to fail again. After two attemp of failed repair they will offer you a new system exchange.

 

I did try searching the registry for other instances of AWCC and stuff, but I may not have gone to the DELL folder exactly as you point out, and I know for certain I didn't delete the OCControl stuff. I'm currently deployed, so I'm really loathe to return it and wait another month for a replacement order to arrive. Every other aspect about it was great. I got a pretty great IPS screen too, with no backlight bleed and pretty punchy colors. It runs pretty hot, but that seems par for the course with this gen design. TS seems pretty good at capping my freq where it's needed when gaming. Oddly I get higher CPU/GPU temps when playing a game like Assassins Creed Origins than I did running Valley and OCCT simultaneously for an hour.

 

I'm reading through the thread, I just wanted to chime in and say my 15 r4 has a 2tb 2.5" drive in it right now as I type this.

 

After applying Gelid GC Extreme and undervolting by -179mV my CPU never throttles, maxes out at around 85c, GPU definitely throttles at 90c, but only throttles down to around 1300 Mhz at the worst, so around as good as a 1070 Max-Q ... Which I'm pretty happy with. I'm going to be applying some Kryonaut tomorrow, so i'll report back with any improvements if I see any (When I repasted I definitely put to much on both, so that may be the reason for under performing Gelid, I have no idea what the stock paste job was like because I bought it as an open box from Microcenter and the previous owner apparently only put a small dot of AS5 onto both, which of course meant the GPU was immediately throttling within seconds of max load being applied). For reference I have the 15 r4 w/ 120 hz 1080p TN panel, GTX 1070 OC (Which does clock up to like 1900 Mhz before it throttles, so super impressed by that, if I can get the temps under control, this will be a beastly laptop) and it has the i7 8750H, which clocks up to 3.9Ghz on all cores. This is while gaming, because I don't really feel the need to put the system through unrealistic stress tests. I just run the types of games I'll be playing and monitor temps.

have installed a 1tb SATA SSD into NGFF slot 2, and a 2TB Firecuda into the 2.5" Bay. System came with 16GB of 2400 Mhz RAM, and a 256GB NVMe SSD & a 1TB 7200RPM HDD. the NVMe drive is very fast read, not so fast write, but to be expected I guess.

 

Kryonaut applied. Differences are minimal.

Rip.

They probably shouldn't advertise it as a full GTX 1070 if it throttles to max-Q levels within 15 minutes.

I am not a bad enough dude to put LM on it. So i guess I get to deal. Ah well. for 1300 it aint bad. (Open Box at Microcenter)

 

You need to make sure the heatsink contact is good. if the heatsink is raised too high because of too thick thermal pads, you will never fix it with any paste.

 

I started skimming around page 14, I'm guessing this was discussed already, but how exactly do you do that? What do you look for? I'll probably need to buy more thermal pads as well then? What's recommended? Thought I saw Gelid Thermal Pads mentioned...

Thanks for the tip.

 

You need to use cheap paste to check the spread and first start with the same thickness as stock pas. Check the paste spread and if it isnt good enough try it with thinner pads. And check the indentations on each pad while doing this. In short a lot of trial and error.

 

Does anyone not have problems with their machine? I'm considering buying one, but it seems like everyone complains about overheating. Do these problem only occur when playing games on max settings?

 

Presumably the ones who don't have issues don't complain. So it's hard to tell.

 

Exactly what I was thinking, most people who don't have problems are are not pedantic with their computers do not bother posting here or have never heard of the problems people here are dealing with.

 

Does anybody know exactly how thick the stock thermal pads on this unit are? (Alienware 15 R4)

I'm taking a look at pads to order to replace what I have and I guess I'd rather not open it up to find out.

 

I strongly recommend you to measure it yourself. Not all heatsinks are the same and not all tolerances are the same. When I used the measurements from iunclok on the 15R3 my unit was vastly different.

 

I was afraid of that. haha. Great.

So you what, put the heat sink on with no pads and measure the space?

 

Use same size pads, check the indentations. Use cheap thermal paste to check the spread on gpu, cpu and if ths spread is good keep that size. If the spread is bad, try a thinner size pads.

 

I read iUnlock's guide to LM on these machines (15r3/17r4) but one question i have is longevity. I'm slowly reading through the thread, but damn its' like 350 pages long.

Anybody here might help me be lazy and know?

How often do you need to upkeep this. Is it savvy to just leave it on there? The laptop travels a lot. Sees lots of vertical rustling inside backpack. last thing I want is some LM to drip out and fall some where else on the board and fry it. I'm not sure how much it moves, but it clearly moves enough that people want to put Electrical tape on it.

So what's the long LM game look like. Should I be redoing it/checking it ever few months? Once a year?

Thanks