Hi all,
My R2 has just arrived so as promised here is a thread comparing it with my R1. I plan to complete the following tests and will update this thread accordingly as I get the results:
1.Initial impressions/Looks/Noise/Heat
2.Power consumption
3.Battery Life
4.Benchmarks Synthetic (3dmark06, vantage, fritz chess)
5.Games (BFBC2, GTA4, COH, DOW2, COD WAW and COD4, Dirt2, SupComm 1& 2)
6.Video Encoding (AutoGk, Handbrake)
If anyone wants anything else let me know.
-
Initial impressions - The stealth black is fantastic! It looks so much better than the cosmic black of the R1. It has a nice feel when you touch it and doesnt attract fingerprints as easily. I will update further done the line to its durability.
The port layout has been slightly changed, I really miss the VGA as I take it to work and often plug into a monitor there. All our work monitors are VGA onlyOther than that the port layout is better than the R1 as the video outputs are now further back. The only thing I dislike is the ethernet is too close to the front for my liking (I am very fussy!)
Noise wise, so far it is much quieter than my R1, I have just put it through 3dmark06, the R1 kicks up the fan almost straight away and sounds like a mini jet engine, the R2 kicks up slightly but still remains quiet! I havent had a long gaming session yet so I cannot comment if it will kick up once the temps increase.
Idle noise seems good so far, I will update once I have finished installing everything
Turbo Boost - noboby so far seems to have done any testing of the turbo boost, so I have been testing it with prime to see what the multipliers use under load.
Without overclock and with Prime loaded on all 4 cores turbo boost fluctuates between 9x and 13x but mostly sits between 11x and 12x. This seems a shame to me as I was hoping it would sit at its 14x as listed by intels specs for both cores loaded. The temps were fine so I do not know why it would fluctuate so much (desktop and i7 mobiles do not fluctuate like this, they sit at the highest level unless temps or power thresholds get met) I do not know if this is a power issue within BIOS, and hope a new BIOS will improve.
OverClock. I managed to get 156 overclock on my i7, this is a bit of a shame but still a nice increase over the stock 133. I will try more at a later date.
With the 156 overclock I re-ran prime. now with 4 threads it sits between 9x and 11x but mostly at 10x.
This seems to me why the overclock does not seem to give big gains. As although it is higher base clock the multi is lowerthis again I hope Dell can improve on in a BIOS update
-
I would rep you but I don't know how. Thanks for this info, keep em coming.
Pictures would be nice too -
to rep you just click the scales symbol underneath the user name and post count
Will do, I am stuck at work but will take some photo's tonight. Sadly my photo skills are pretty poor so dont expect too much!
Quick update, still havent had the fan kick in high yet! the laptop is warmer on the left side than the R1, obviously due to the lower fan activity, it is not uncomfortable though (unlike a sxps 16).
I am currently trialling different overclocks with prime to see if I can get some better results. so far it seems to perform best on stock settings! Go figure. Also WEI is 5.9 without OC, and 5.8 with OC (have tried 156, 150 and 140) -
thanks for the info. Glad to see more first hand R1vsR2 comparisons. Even the little things like mentioning the port locations and heat. You said a little warmer on the left side. As in the palm rest? B/c that was a minor gripe of mine of how warm the memory got on the left palm rest.
You're not the first to mention that while OC'ing you notice lower gains. Have you tried turning Turbo off in bios and just going OC and see how wprime runs?
+repped thanks ninja
*edit* just a few random questions. what screen did you get? the samsung or tha auc or whatever. Just curious if they're still shipping two others. Build quality? same as R1? Does the power supply have a blue LED charging indicator - since it's a new wattage haha. -
in answer to your questions:
The plamrest seems to heat up quicker at idle but is actually no hotter than the r1, i guess its the lower fan speed and hotter i7. the touchpad seems warmer but again by no means hot (trust me if you have used tghe sxps or studio 1555 you will know what a hot touch pad is like!)
I got the AUO screen, so both my m11x's are the same.
Build quality seems slightly better on my R2, not much in it.
the new charger does have the blue light
Taking some wattage readings at the mo and initial findings are the idle is about 20w! so a lot higher than the R1's 11w. I havent mucked about with the settings yet but it seems the 335m is never fully off. I might have missed a basic setting but I am going through a lot of installs and benchmarks.
The max wattage is about 55w so very similar to the R1, I am sure they could have used the original PSU if they had wanted.
I have just completed some Fritz chess benchmarks:
M11x R1 C2D @ 1.6ghz - relative speed: 4.69 KN/S 2250
M11x R2 i7 @ stock 4 core - 6.36 KN/S 3054
2 core - 4.56 KN/S 2190
1 core - 2.89 KN/S 1388
M11x R2 i7 @ 140BC 4 core - 6.26 KN/S 3003
2 core - 5.09 KN/S 2444
1 core - 2.70 KN/S 1294
M11x R2 i7 @ 156BC 4 core - 6.13 KN/S 2940
2 core - 4.74 KN/S 2274
1 core - 3.22 KN/S 1546
Looking at the results you can see the i7 is 35% faster than the C2D at stock speeds. OCing the i7 gives virtually no benefit and hurts it on all 4 virtual cores. This seems to me that either heat or power is holding it back. I dont think it is heat as the cores never go over 60c -
thanks again ninja for your responses. Nice on the charger, it's those little things hahah. That was just a random thought, surprised they did provide the blue led on the charger end =)
I had a Dell SXPS 1640 and yeah that touchpad/palm got really warm, combine that with the shiny black plastic and you could see sweat and oil marks out the . Kind of a pain. Sucks that area gets warmer faster, which already bothered me on the R1 a little bit. Thanks for further benchmarks -
+repped, loving your work!
-
Thanks for the rep everyone!
3dmark06 scores:
stock - 6962/3053/3099/1812
140BC - 6949/3054/3102/1780
156BC - 6977/3062/3108/1742
Again strange how the score lowers as the OC is increased!
I am running out of benchmarks I can run at work! when I get home I will get steam and BFBC2 installed and start the game benchmarks.
So far really disappointed in the OC and the higher heat on the unit but it is nice and quiet. dont know which bothers me more, heat or noise. -
hey ninja since you're keeping up with the thread rather nicely, I'll go ahead and ask some more.
Are you able to download at work? I'm curious if you can get the Final Fantasy XIV benchmark going? It's a smaller bench I think 129MB? Around that area.
The OC thing is very curious. I wonder why that is. I assume you're plugged in and high performance? Maybe it's a heat thing... fans active? Not sure what it could be except some sort of throttling. Anyway, I asked above, have you tried no-turbo and then OC again? -
Our firewall blocks any game related sites
if you have a link to it hosted on a non game related site I can give it a try.
Its plugged in and on high performance. everything seems to be working fine, its just the multipliers wont boost as high. Its got to be throttled by the BIOS for either heat or power but I cant see why. The CPU never gets over 60c on prime. I have tried raising the back which lowered the temp a couple of degrees but didnt help. I will put it on my zalman nc2000 when I get home. The fans are active but never hit the high rpm's of the R1. The R1 seems to have 3 active stages, the low idle rpms, then medium and high you get whilst gaming. So far this R2 has only had the idle low rpm and the medium(hope this makes sense!)
I did see your request, and it is next on my listI will turn off turbo boost and re-run 3dmark06, wei and fritz
-
I repped, well deserved.
Oh btw when you test battery life, how about trying with Hyper Threading AND Turbo turned off, compared to those two things turned on? I heard those things use power. -
I can turn Turbo off but I cant turn off HT in BIOS.
I will test battery with Turbo off and with Turbo and OC off to see the difference too. Before I do that I am still trying to tweak the system as it idles at 20w which is really high! it never fully shuts off the nvidia by the looks of it.
Does anyone know any good battery meters that show discharge rates? -
rock on ninja.. does this link work for FFXIV http://uk.download.nvidia.com/nzone/international/benchmarks/ffxiv/FFXIVBenchmark.zip
it's through nvidia.. -
-
^ not sure what you're asking if what shuts off. About the graphics? Yeah if you manually switch the R1 to integrated, the 335m should completely shut off, some issues are sometimes it doesn't shut off all the way and a stuck fan on becomes a problem while on integrated. Putting the system to sleep usually fixes this.
-
Well done aznguyen316! that link worked perfectly. Downloading now.
Couple more updates, no fan sticking issues so far. It has been so much quieter than the R1.
returns from sleep a lot faster than my R1 too.
Still cant get the power down below 18watts -
man it's simply great to hear from an R1 user that has some good experience with the laptop. I have wake from sleep issues with my R1, sometimes it'll take like 30seconds with me pressing keys and touchpad trying to get it to respond. Noise is good.. I have an SSD so I def only ever hear fan noise (or my damn hinges creak). Since I'm just at work doing some Photoshop stuff I'll be refreshing NBR a lot.. hence why I'm all over this thread
-
How can the machine perform worse when overclocked? Very strange, and reason to worry.
-
Mackan - I dont think it is reason to worry. Dell have obviously limited the laptop either thermally or via power use. This will hopefully be resolved in a BIOS release.
It is fully stable at either stock or 156BC for me, its just we would want it to turbo higher whilst OCed -
Also just for you, first results with turbo off! Firstly I can still only get to 156 even with turbo off, I was expecting to get the full 166
Secondly, the power consumption is 2 Watts less at idle and 10watts less at full load!
Fritz cess benchmark for 4 cores: 5.24 KN/S 2515 (about 20% less than stock clocks with turbo boost)
3Dmark06: 6284/2900/2864/1458 (this is very similar to a R1 stock bench, the CPU score is almost exactly what I get with an overclocked 7300 cpu) -
did you extract the .zip file before tyring to run the application? I read there may be optimus issue with the benchmark so that could also be the D3D error. Bummer
and so it looks like it's the turbo boost that gives the i7's the extra power since even an OC'd 156 is only as good as an SU7300.. That's very interesting.
so basically do turbo only and no OC -
yep extracted the zip. I also went through the folders to see if there was a DX setup. I will have a play when I get home from work and try and get it working for you.
The Turbo definately gives the i7 its performance, clock for clock the i7 is probably only about 5-10% better due to the new architecture (and HT where utilised) now if only we could get turboboost to work better with the overclock we would be rocking!
The only issue that is p1ssing me off at the moment is the idle power consumption. I think this is the reason why the battery life is only about 5 hours. If we could get the nvidia to fully switch off I reckon we could gain another hour of life and also have less heat (the R1 takes ages to heat up whilst on intel, but heats up quite fast on the nvidia. the R2 heats up just like the R1 on nvidia. By this I mean the palm rest warm up, the bottom panel also heats up and the fans kick in more foten) -
-
did your FF benchmark come with an installer?
Also for anyone interested, the F2 battery option still does not allow you to stop charging like on my m15x -
Just ran the exe... "FFXIVBenchmark.exe"
-
hmmm, strange, I tried that. must be an optimus problem. Will go and have a play
-
What are you using to measure the power? And is it possible that the higher power is due to the integrated card being on the same die?
-
I have a kilawatt meter that measures it at the mains.
Its not due to the integrated card as that is no way 10watts at idle. to give you an example of idle figures:
R1 with intel - 11w
R1 with nvidia - 21w
R2 - 20w
It seems too unlikely its anything but the nvidia raising the power -
It is an optimus issue, you're going to start experiencing these things. I suggest you see the problems first, then post it in the thread. Steam has most of these issues,, battlefield or i mean punkbuster will kick you out for detecting multiple direct3d's. I hope you don't experience it though.
Good work here. -
Here's one free Apple IRep for Ninja.
I guess I'm stuck with my (by mere months) outdated, "Iowa-class" R1. Oh well. She's old, she's great, and I love her. lol -
wow! I have just played BFBC2 for half an hour and WOW
32 man server, native res shadows low textures high and rest medium with 2xAA 4xAF and it was playing 30+ fps with the odd dip! I am now officially sold.
One thing to mention is this unit runs a lot hotter than the R1 but the fan never gets remotely load. I think there will be a bios release with a louder fan profile at some point. Internal temps ramain good (CPU max 68 and gpu 65) whilst gaming. Its just the palm rest that gets hot. -
odd dip?? what do you mean by that?
-
-
amodar - dont worry, I am getting round to that. I will update this thread to show comparative temps, benchmarks and battery. I am still installing everything and currently struggling with DOW2 on steam, it wont let me change settings
Any ideas arklight? I know you struggled with some steam games -
-
I think he meant wow as in
-
lol sorry for the confusion, aznguyen316 was right
Just been testing TF2 and LFD2 and have seen large gains with the R2. I am trying to put a fair bench together at the moment to show the difference -
Actually you don't need to add anything to the nvidia control panel. It just automatically runs the 335m if you're going to play and it is by default.
Also it will say intel HD since people are saying that it's the main GPU for the display and the 335 with the optimus support driver uses the 335 to optimize the intel + use the 335 to play games.(correct me if im wrong) This only works for games or 3D applications, if you're running windows based testing like 3dmark or whatever that's run in a window by default, it will detect and use the intel gpu. Some games doesnt have support for optimus like counterstrike, but the 335 will kick in when you play.
1. Try running the game first, if it doesnt work and it crashes, try number 2.
2. Just make a new user account and run the game from there. that should work. I made another user account for games. And all the games that aren't working on my main account worked from there. except for assassins' creed 1, which crashes on the loading screen before the first mission.
hope that helps. -
First post ^^
Really nice work ninja, keep it up,
BTW how did u solve the VGA problem u mentioned? an adaptor? -
some of my R2 benchmarks for you ninja, I can OC to 162 so I tried a few variations. With OC, OC with Turbo, and stock i.e. no OC and Turbo on.
Dell Driver 189.69
Standard/ OC CPU 162, no boost/ OC CPU 162, Boost
3dMark06
6934 3D Marks/ 6500/ 6929
3093 SM 2/ 2953/ 3091
3108 SM 3/ 2958/ 3110
1736 CPU/ 1547/ 1731
L4D2 - Fraps results from single player - dead centre
22 min/ 33/ 34
64 max/ 65/ 65
51 avg/ 51/ 52
L4D2 Settings
Resolution 1366x768
Display Full Screen
Anti Aliasing 2x msaa
Anisotropic 2x
Vert Sync triple buffered
Multicore render enabled
Shader very high
Effect Medium
Texture High
Page Mem High
This may be useful to know for some folks, I had installed the verde mod drivers, however this afternoon they stopped working? cant fathom out why so I reinstalled the drivers from dells website and they seem to be working much better than i remember first time round. This may be something to do with the daily updates that the nvidia optimus driver seems to apply for game profiles as a guess? can anyone else clarify this? thanks
Hope this is of some use. -
I noticed that too. I was one of the first who was eager to install the modded inf for I thought it will resolve some issues. I guess I or we paid more attention to the .exe adding on the control panel rather than playing the game. Mine would not add anything at all except .exe of updaters. but when I play, it runs ok after reinstalling the stock dell driver. -
It is also true that the latest drivers from Nvidia can sometimes be very buggy. This latest set (the 256 series) seems to have some varying issues for some users. For example, 258.49 (Quadro) were the only drivers of the latest set I was able to get working on Hybrid Graphics notebooks. The others just didn't really function properly at all. Bleeding edge does not always mean best.
I have a gift for you guys that will hopefully be up pretty soon too. Should help make figuring out some of this Optimus stuff a little bit easier. -
i can't wait.. Thanks Nautis!
-
-
Im totally poisonned to get one -
-
the reason I sound optimistic is because so far, the stock performance has been so good for me I dont really care about the overclock (and I really thought I would).
I am able to play BFBC2/TF2/LFD2 with solid fram rates (which I was unable to do with my m11x r1, it would drop below 20fps when a lot of action was on screen in all the above games). If the overclock is improved then I will happily use it but in the meantime I am enjoying finally being able to play BFBC2 fluidly.
The thing that is bugging me the most at the moment is the idle power consumption, I am 99% sure it is keeping the nvidia card on whilst on intel gfx. I am going to try a few things today to fix it.
Also I am going to update this thread today and move a lot of benchmarks to the first thread. Thanks Spoonface, good to see some comparable results. I also started benchmarking LFD2 last night, comparing the hotel lift scene which is a system killer with all the flames and hordes of zombies. the R1 would dip to about 12fps for me whereas the R2 didnt ever drop below 25fps! quite a big improvement in a cpubound game -
You can PM Erawneila about it and ask about the observed overclock behavior. Personally I think it is very worrysome that the overclock doesn't seem to provide any performance increase at all so far. Makes it kind of pointless to have included it in the BIOS in the first place, and might misslead people when buying the R2.
But the AW engineers must have done performance benchmarks, as well, right?
By the way, it is possible that the overclock makes the CPU reach it's limit when it comes to power consumption. I think the Turbo Boost on the i5/i7 only kicks in if voltage/power consumption/temperature, etc are within limits. It's possible there isn't much room for that on these ultra low voltage CPUs. -
From what I can see so far, overclocking the i cores may be a misleading statement. All that really seems to happen is that when you OC, it just prevents the CPU from stepping down when it doesnt need the power, this has the effect of keeping the framerates more stable and prevents the dip in performance. It doesnt actually boost anything at the top end over and above the standard settings of no OC with turbo on. Basically OC the core just means that you are locking it to a certain speed and not letting it drop any lower, even if it doesnt need the extra ooomph.
The processor can only produce so much power, what the i cores seem to do is let yoiu decide how you want that power distributed, all be it in a PITA fashion. If that is the case (im guessing here lol) it would be nice if the option to switch between these modes was possible at the OS level rather than the bios. That way, tweaking turbo mode and OC mode on/off for certain games and apps would be much more useful and less hassle to acheive.
M11x R2 vs R1 comparisons/Benchmarks
Discussion in 'Alienware M11x' started by ninja2000, Jun 28, 2010.