*OFFICIAL* Alienware 17 and M17x Benchmark Thread - Part 4

Yes, that is a P570WM with a desktop 4930K 6-core CPU and 880M SLI. It is not a direct competitor with the product this thread is intended for, but it's definitely fun to look at big numbers like that.

Nice job, j95.

OK everyone else, let's knock off the chit-chat and get down to business.

Spoiler

 

Since PassMark only acknowledges and tests one GPU, this is relevant here for comparison with the M17xR3/R4/17 product line.

PassMark Results | Baseline Validation: PassMark Software - Display Baseline ID# 241765

3920XM @ 4.5GHz | GTX 780M (No Overclock)	3920XM @ 4.5GHz | GTX 780M (1033/1500 @ 1.0500V)

With SSD memory-cached...

PassMark Results | Baseline Validation: PassMark Software - Display Baseline ID# 242044

 

Yeah, I think the extra voltage from 1.050 to 1.062 made all the difference for running at 1019/6000 for my alienware 17.

By the way Mr. Fox I was looking at your batch files and it goes over 1100mV though precision x doesn't go that far it's just maxed at 1100...is that an nvidia inspector thing?

Finally does overvolting increase GPU temps in general?

Sent from my One using Tapatalk

 

I might get back onto this soon. I downgraded my CPU to a 3840QM because my 3940XM's memory controller stopped working with my ram at the speeds in my sig, blue screens everywhere! Doesn't matter too much though, as long as I can achieve close to a 10k physics score in '11 without running into heat issues I will be satisfied. I'm going to turn my attention to pushing the 780m. We shall see...

 

On your first question, some are different than others. But, 1.100V is enough for some decent overclocking. Check NVIDIA Inspector for max voltage, but it should not be different in EVGA Precision X. If it is, use NVIDIA Inspector (or the batch files) if you want to use more than 1.100V. Whatever the slider limit is using NVIDIA Inspector is probably your max. Using the batch files probably will not let you set more than the max. You can check that with GPU-Z to confirm.

Second question, yes, definitely... voltage does increase load temps. It will on CPU as well as GPU. That's OK as long as they are managed, but you can expect them to rise and should watch to be sure they stay under control.

Man, it would be great to see you benching again. It seems like forever, bro.

 

Looks like my GPU overclock is drawing in an extra 43 watts. So it's using 165 watts over the spec'd 122 watts. I am not sure now much the other components are drawing from the 240 Watt power adapter.

Sent from my One using Tapatalk

 

Hi
I run a Overclock on my i7-4930mx and soft Overclock 79/200 on gpu, with a default VBIOS. During the test of 3DMark 11 and 3DMark use my machine now up to 225 watts.

 

So I got hungry and pulled the trigger on another 3940XM. A question now and I also will start a thread about this if no one here sees this. What is the verdict on M17x R4 systems and the 330w PSU? I have read conflicting reports about whether it charges the battery or not and whether I need a mod. Im going to be pushing my system fairly hard.

 

So first time benching again using 3DMark 11:
NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4

Ran without a hitch, no artifacts, glitches or anything. 10k GPU, 10k CPU and max 75C on both with fans on full speed.

 

Nice job! Your "obsolete" M17xR4 is already stomping the Alienware 18 with 780M (SLI disabled, of course) in 3DMark11. +1 Rep :thumbsup:

Alienware M17xR4 - 3940XM, GTX 780M (1) Benchmark (P9873) versus Alienware 18 - 4930MX, GTX 780M (1) Benchmark (P9601)

If we can get the TDP limit fixed, this will turn into a blood bath for the 18 with a single 780M, LOL.

 

Oh come on! So close. Wish that Physics score was a bit higher though:
NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4

 

Almost cracked the 10k's buddy! Cmon you can do it!

 

I cracked it last night:
NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4

 

P10507 NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4

Highest Graphics Score 10903...
Physics 10412 ruins everything.

 

That's impressive. I wonder why my Physics score won't go higher. Perhaps it's because you have faster ram?

 

Número uno P10551 NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4
Graphics Score 10989
Physics Score 10445


We need to fix the CPU power limit (renaming...) 99W 128 99W Pri 1040 Flex 14. RAM +200 pts.

 

The RAM does affect the physics score, definitely, in 3DMark11. But, the power limit is the biggest barrier that needs to be broken. It's really inexcusable that Alienware did this to the M17xR4. And, I seriously doubt it happened by accident. I'm very agitated by this kind of shenanigan. They essentially ruined the Alienware 18 with artificially imposed limits crippling the hardware.

 

It all started...but unfortunately no one said a word about this (leading to wrong conclusions) It was all about USB3 dropping and Creative chip issues (sorry) It's fixable, the board is power capable. unlike the new AW..? BIOS updates 880M TDP power management issues...?

 

which drivers are you using for gtx 780m?...thank you

 

In case you haven't already done so, this is mandatory for good performance: NVIDIA Kepler VBIOS mods - Overclocking Editions, modified clocks, voltage tweaks. GTX 680M, 780M, 880M and all of the lesser Kepler GPUs will provide unacceptably poor performance with a stock vBIOS. This would be a prerequisite to fiddling with different drivers.

 

I will do it, but untill then I was trying to get the corect scores for stock bios....that I am far far away from them

 

"Stock vBIOS scores" could be misleading. What some people call stock may be using default value with no overclock, but modded vBIOS. There is no way to tell, other than the results will be consistently better scores with the svl7/johnksss vBIOS, even without any overclocking.

 

this scores NVIDIA GeForce GTX 780M - NotebookCheck.net Tech ...my last gtx 770m did like there

 

Hopefully svl7 can work his magic. I've offered him my support and volunteering my system for testing over at TI so there might be light at the end of the tunnel. But we shall see. I wonder if anyone has a m18x triple heatsink for me to borrow? I want to see if there is a possible mod to use it on our machines.

 

R4 users who want to unlock the potential 3920xm/3940xm, 20C thermal headroom...:hi2:

Yes it's possible, now it's worth it...AW 17 triple pipe heat sink (4930MX) never happened/ mistake.

 

Wow interesting. When did you start that?

 

A year ago (R3), I thought that "Dell Cares" would have provided a revised version. Triple pipe heat sink for AW17/MX, like they did with the R3 580M/overheating. I was wrong. Haswell/MX with a double pipe htsnk is simply ridiculous. Certainly not a Dell/AW Engineers' decision, the best designed cooling systems.

Back to business, hands dirty...

 

Actually, you don't.
NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M18xR2
I'll get around to running it correctly later.

 

Yep. Predictable, August 28 2013...

 

Still numero uno on a M17x though

 

FireStrike 7148 NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4

 

First Benchmark I posted on here. I know theres better ones out there for a 7970m cf'd M17xr2 but its my best one to date.

AMD Radeon HD 7970M video card benchmark result - Intel Core i7-920XM Processor,Alienware

and Firestrike

AMD Radeon HD 7970M video card benchmark result - Intel Core i7-920XM Processor,Alienware

 

3dmark11 10900 over all score.
NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-3940XM,Alienware M18xR2

 

Sweet Jesus! Dem some mean scores. Excellent work.

 

Finally got my Haswell Fire Hazard OC'ing decently so I was able to get some better physics scores:

AMD Radeon HD 8970M video card benchmark result - Intel Core i7-4700MQ,Alienware 0MPYM4

3DMark11 doesn't like my clock settings apparently, that's the first time I've seen the time measurement error. I'm going to run Firestrike later, but it's late and I'm going to bed...

 

This looks like unstable cpu;-( Also think the last sys ifo from futuremark is not completely stable

 

most likely because I messed with the reference clock trying to squeeze a little more speed out of it.

Sent from Tapatalk from an increasingly expensive smartphone

 

How high have you adjusted the reference clock?

 

Got up to 102 before it reboot on me. Stable at 101.8...or seemed stable thus far. Not going any higher.

I was very surprised I got it as high as I did!

Sent from Tapatalk from an increasingly expensive smartphone

 

Average of physics score (i7-4700mq standard CPU) in 3dmark 11 is approximately 7200 p. Alienware has no good bios unfortunately. How is the temperature of the CPU during the tests ?. Use hwinfo 64 and Test your machine with Cinebench 11.5 and Wprime 1024 (8 threads). To recheck the stability of your overclock. Add some pictures with the temperature measurements and wattage if you can.:hi2:

 

Temps were fine, was monitoring with hwinfo as well as the afterburner app on my phone. I stress tested with xtu and plan on further testing as well! Thanks for the suggestions my friend!

Sent from Tapatalk from an increasingly expensive smartphone

 

¿Dónde están los "Sky Diver benchmarks" amigos? ¡Ándale! ¡Vamos!

Alienware 18 4.5GHz - 780M (1) max stable GPU OC < for my single-GPU Alien brothers

NVIDIA GeForce GTX 780M video card benchmark result - Intel Core i7-4930MX,Alienware 01W2J2

And, for an example of what it should perform like, there's the M18xR2... kicks its butt without even trying hard, LOL.

 

Well comrades, I seem to have an epic 7970M here.. I changed my TDP from 84W to 100W and increased max voltage to 1.1V and with clocks of 1000/1450MHz and my 3940XM at 4.1GHz, I got the following score of P8304 with a graphics score of 8.1K... I tried to push to 1050/1450 but for some reason halfway through the benchmark my systems shut down although everything was running cool.. Strangely my GPU also has stayed below 75C in these tests..

AMD Radeon HD 7970M video card benchmark result - Intel Core i7-3940XM,Alienware M17xR4

 

That is exceptional for a 7970M to be able to run those clock speeds. There are not many examples of them that do as well. :thumbsup:

It sounds like you may have exceeded your system's power handling limits. If you plan on doing much of that, probably need to make yourself a dongle mod like Nospheratu did and grab a 330W AC adapter.

3DMark11 and Vantage are more demanding than the new 3DMark suite. You might be able to complete all of the tests in 3DMark (2013 version) without the system shutting down.

 

Borther Fox, this power mod is not strictly required is it? I thought the 330W adapter was plug and play.. The motherboard would draw as much power as it needed from it^^^

 

It is plug and play on the new Alienware 17, but I don't know if it is on the M17xR3/R4. If it is, then that would be fantastic to just plug it in.

 

It is on the R4, confirmed by j95.

 

Awesome... makes things easy then.

 

Brother Fox, I have one question although its more of a worry. I will still run my 7970M at stock clocks but is it safe to keep the 1.1V or should I flash back the vBIOS with 1.05V? I'm just a bit worried that this might reduce the life of the GPU...

 

That is a legitimate concern, Tom. I don't know that 1.1V is necessarily unsafe, but it certainly is not going to enhance the life span of your GPU. As long as temps are OK and you are not seeing any kind of unusual behavior, you probably don't need to get too stressed out about it. After all, you can always buy another GPU if you need to. I don't know of anyone that likes the idea of having to buy replacement parts. The only good thing about that is that it may give you a good excuse to spend money on an upgrade where it might have been too difficult to justify if what you had was adequate.

For what it's worth, I seldom overclock or overvolt my video cards except when chasing numbers in benchmarks. This is always for short periods of time, with close attention to what is taking place. I watch to be sure my GPUs are not overheating and I watch for evidence of artifacts that might indicate I am pushing too hard. I back off immediately when I see signs that my GPUs are not handling what I am asking of them. Gaming typically does not require overclocking to have a good experience. Unless you are struggling with low FPS and need extra voltage to run an overclock that is too high for stock voltage, I would (and do) run stock voltage with no overclock, or a mild overclock that is stable and runs cool with stock voltage.