You're right that his way of scoring it is wrong, but you're absolutely wrong that CPU overclock has minimal impact. You're thinking 3dMarkVantage.
http://www.tomshardware.com/reviews/cpu-gpu-upgrade,1928-13.html
Pay attention to overall scores for the same card... and this get's even more pronounced when we talk about newer generation cards
-
chewietobbacca Notebook Evangelist
-
Eggs Scrambled Notebook Evangelist
See my post above as well. -
chewietobbacca Notebook Evangelist
And you're going to be wrong
If you really want me to, i'll turn on my desktop with my 4870X2 and do some runs... easy differences -
Eggs Scrambled Notebook Evangelist
I didnt realize so many people thought I was talking out my butt on this board
(see my earlier post with it all spelled out)
You'd be surprised how much a system that actually has a major cpu bottleneck can improve when overclocked in 3dmark06 when a system with no major bottleneck like the intel quad system above gets such huge differences from 200mhz overclocks. -
why do l4d and l4d2 have such different results on tests. l4d2 is almost exactly the same, i dont see where all the extra power goes.
-
well, perhaps you are right, it goes against all I'm used to know about 3dmark, but I must admit i'm not "updated" about the 06 version, what was a sure thing to me is that a 20% increase in let's say 3dmark03 score would certainly means that the system is much faster in gamming performance... perhaps the 06 is stupid and put that much score dependence in the CPU... basically saying they have "detonated" a working and consagrated program and make it useless... very strange indeed...
anyways, probably with the "cpu fix" (overclocking) it would solve the bottleneck and make the GT335 shine... only when we have the Vaio new Z benchmarked we will know... but I will try to find some other GT330 notebook with more "real world" tests... -
Eggs Scrambled Notebook Evangelist
it looks the same but there a lot more crap under the hood in the second.
One of the things they worked intensely on (surprising in only a year huh?) was the differences in common infected so that you felt like you were killing a massive amount of different people rather than a clone army. -
Eggs Scrambled Notebook Evangelist
you didnt need to quote all of those posts
suffice it to say its a lot of speculation at this point, part of the appeal of the m11x is the brand new graphics card made just for it that has more cuda cores. The exciting thing about the overclock was that the inarguable deficit in the m11x was its cpu and now that it has a huge boost in the OC option (33% or whatever is a LOT of improvement) it's going to perform so much better than expected.
Let me stress that in the reviews that have surfaced they said or showed that the m11x can handle almost a guaranteed average of 30fps in medium-high settings when the cpu was clocked at 1.3ghz (and thus the ram was 800 instead of 1066mhz) so improving that and also having the potential for overclocking the gpu and even the cpu even farther with setfsb or bios mods make this one heck of a powerful 11 inch machine.
It's really quite unbelievable when you think about it. -
Oh no I know you're not dissing Alienware. I just wanted it to be known that some of Valves games play pretty well on the M11x.
I had to do some research because I want to get L4D2 and be able to play it on the train or plane.
But yeah the way Alienware was thinking was probably putting the best GPU in it as possible and giving it a decent CPU since a lot of gamers will need the GPU more. Luckily they gave it a CPU that can overclock extremely easily and we'll see how far up it goes.
-
Well, this proves the point I guess. He did a 46% overclock on his CPU and it netted a 22% increase. I imagine if he would have OC'd his GPU 46% the net would be much higher than 22%.
3dmark works for what its intended to do, give a broad guestimation of what your computer is capable of. Honestly, I would rather have a higher 3dmark score than a lower one
-
I agree with that very much! that's the major drawback to me in the M11x... if they put a 13'3 it would be perfect... but 11'' sux... and that's not all... the worse is that there's no docking station for it... if they released a docking station it would be easy to use a bigger display to play games, and use the notebook to work the other times... but without it you would always have to connect many cables... I certainly don't plan playing even a minute in a 11'' screen...
one competitor i'm looking at is the Gigabyte M1305... it has a desktop GT220 card in it's docking station (connect with a proprietary PCI-E 4x port), a 13'3 screen and weight 1.8kg (versus 2kg start weight of the M11x)... it has the same CPU as the M11x but a score of 5500 3dmark06 points... so I don't know what to think about it's performance now... -
chewietobbacca Notebook Evangelist
-
L4D2 is much more power hungry. My computer (core i7 920 at 3.8ghz with dual gtx 280's) can run L4d upwards of 200+ fps in most situations (sometimes in crazy battles it will be mid to low 100's). (this is maxxed out, highest AA highest AF) but last night playing l4d2 i saw 60-80fps when the shiz was hitting the fan (same settings, highest aa af). So they did something to make it way more graphics hungry.
Both fun games though -
sprry for the noob question.. lol
but does anyone know if this thing has wi-fi??? -
very odd, i actually think that the common infected LOOK BETTER IN THE FIRST GAME
http://www.pcgames.de/aid,699853/Grafikvergleich-Left-4-Dead-2-gegen-Left-4-Dead-Wer-hat-optisch-die-Nase-vorn/PC/Vorschau/ skip ahead a couple slides.
i think its just poor optimization on valves part. -
Good links. One thing to remember, many graphics cards increase in performance as the CPU is increased (aka, for a graphics card at a lower resultion like a gtx 280, it needs more cpu power) So as you OC the cpu, the graphics card actually becomes faster even if you keep the same gpu settings.
It's one of the reasons that made the core i7's so awesome, if you were running dual graphics cards, the cpus opened up a world of performance for them. -
nope. not an option.
there's a 56k modem though. -
Could be, but there might be more pixels being rendered or something. I know the "heros" look SO much better in the second one.
-
chewietobbacca Notebook Evangelist
No, that doesn't prove your point at all.
Beyond the obvious fact that most GPU's rarely get above 20% overclocks, much less 46% - if your CPU is already holding you back, increasing GPU speed will not be able to increase your performance greatly
Increasing GPU speed only starts to see near linear gains when the resolution is extremely high - we're talking 1920x1080 and higher - and that's where the GPU becomes the bottleneck
An anecdotal one:
http://www.hardforum.com/showthread.php?t=1383884
19500 on Q9550 w/ GTX 216 -> 20200 w/ GTX 280 -
Yes, everything has wifi these days
-
This is by far the best question asked in the entire thread. +rep
-
We'll revisit the topic when our 11x's arrive. I promise you there is not a snowballs chance in hell of getting close to the q6600's numbers in this little lappy LOL your dealing with scores in the 15k's vs scores in the 6k's for the 11x. I see a net gain of around 200marks going from 1.3 to 1.73 maybe 300 if your using 06 instead of vantage.
-
http://www.tomshardware.com/charts/....1.0-3DMark-Score,Marque_fbrandx876,1697.html
Better link than some kid on a forum. His PC is messed if thats the only increase he got. GTX 280's are much greater gaming video cards than 260's.
And you are mostly right about increasing gpu speeds at higher resolutions. But if you are playing a game at a lower resolution and the GPU is the limiting factor (aka crysis) it doesnt matter how much OC on the cpu you get, you need to get a faster graphics card to notice a major difference. -
chewietobbacca Notebook Evangelist
Yes, but again, that's based on resolution involved and the game involved
These are run at 1680 x 1050, a good intermediate resolution:
http://www.pcgameshardware.com/&menu=browser&mode=article&image_id=839047&article_id=647744&page=1
http://www.pcgameshardware.com/&men...4104&image_id=839048&article_id=647744&page=1
http://www.pcgameshardware.com/&men...4104&image_id=839049&article_id=647744&page=1
http://www.pcgameshardware.com/&men...4104&image_id=839050&article_id=647744&page=1
http://www.pcgameshardware.com/&men...4104&image_id=839052&article_id=647744&page=1
For the most part, at a certain range, the CPU differences are negligible. If you're at the low extreme, sure, it starts to hurt (especially games that use more AI calculations/CPU calculations, such as WiC, UT3, Crysis), but by and large, the difference is in minimum FPS and the GPU is the still the bottleneck -
deleted. should have been with quotes.
-
Eggs Scrambled Notebook Evangelist
time for more math!
It looks like he increased his score about 740 points by increasing the clock from 3ghz to 3.2ghz, So with a 7% increase of his base clock (200mhz/3000mhz) he achieved a 5% increase in his score.
This is nowhere near the correct statistical way to do this, but if you directly translate that to what we'll get, for the 33% increase from 1.3ghz to 1.73ghz, which 33% is 4.7 times 7%, we would get a 4.7x5% = 23.6% improvement.
If that math works out anywhere near as intended in real results, the 6100 score at 1.3ghz would be assumed to go as high as 7539.6 at 1.73ghz.
Now don't forget that this is completely ignoring how much more a clock difference would make when both the cpu is a bottleneck and its lower clocks in general compared to a quad core's clock increases.
It would be interesting to see how close I was with this terrible math to real world improvements though. -
Folks, the whole point for the M11x is to become a new gaming segment: gamer's super netbook. A netbook-like size, but with ability to play newer games.
If you wait for the Asus 30JT, you will have a 13.3' i7 with a 310m gpu. Not sure if the cpu will be bottlenecked by the gpu or vice-versa, but the 30JT seems to fit with what you are wanting better than the M11x. Not sure how much more power in the M11x's 335 gpu, but it may be waiting on the SU7300.
None the less, the 30JT will probably cost more than a similar featured M11x. With no Dell discounts, probably the same price.
rushmore is online now Add to rushmore's Reputation Report Post Edit/Delete Message -
chewietobbacca Notebook Evangelist
No, that link i sent was just to point out to you how the same config with a much faster real world card is NOT reflected in 3dMark06
I could dig up TONS more links from forum users who have benched it themselves, that one was just the first one to pop up
That Tom's link is also pointless since it doesn't give the hardware platform it's being run on. For all you know, it's a Core i7 975XE at > 4.0GHz -
That is biggest selling point for me for the Sony Z you get a faster machine with bigger screen in a smaller package. Yes it is smaller not going to figure it out again but the volume length x width x height of the Z is less than that of the 11x and it also weighs less. To me it is better in every way but the price tag oh the price tag LOL If I was going to use it more often I would definitely go that route but I can't see spending that much on what essentially is just a toy for me to amuse myself.
-
One thing from the link you posted earlier, notice how his GPU clocks were going up with each CPU OC? That means his GPU was being starved for power. I think your M11X will be exactly the same. The GPU is much more powerful than the CPU that feeds it, so OC'ing that CPU should net us pretty awesome gains.
-
The toms link has a true baseline computer that it uses to run all the graphics cards. Who cares what the CPU is, its just showing the GPU power increased by switching cards. You honestly think a kids post on a forum is more indicitive of real world performance than Toms? Wow
Edit: In fact, to determine GPU power you NEED the have the CPU as little of a factor as possible, aka, have the most powerful CPU around to keep that from limiting the GPU's power. -
Here is a good link about 3dmark Vantage, and its cpu power.
http://www.tomshardware.com/charts/2009-desktop-cpu-charts-update-1/3DMark-Vantage-1.0.2,1396.html
Notice how the overall score doesn't change much even with the faster cpu's?
I think thats one of the major changes they did for Vantage. -
Eggs Scrambled Notebook Evangelist
you're right, we really need to just get the laptops and then have some hard numbers to go by.
The point I was trying to illustrate however was the percentage point impact of what that guy achieved with minimal clock changes to his quad. I understand it's a quad core and all but I guarantee ya also that 3dmark06 isnt 100% efficiency optimized for quad cores either, so we may see an even bigger improvement with a dual core, plus the % increase from 1.3 to 1.73 is a lot more than you are making it out to be.
Think of the extreme example, a computer with a 100mhz processor but a gtx 280. It would get like 300 3dmark points total or something else hilariously low. But then you give it 133mhz instead, and on a system like that with an extremely high cpu bottleneck, the 3dmark points will skyrocket because of just 33 more mhz.
I know it wasnt a perfect example of this whole bottleneck business, but suffice it to say, this guy got 5% increase of his total 3dmark points when he has like a super-rig. If our cpu is such a bottleneck, then even the seemingly trivial overclocks like 1.3ghz to 1.73ghz will raise the 3dmark06 scores considerably.
I at least guarantee it's a heck of a lot more than the 300 points you're talking about! -
chewietobbacca Notebook Evangelist
Remember, it's notebooks, you pay more for smaller than anything else
Vaio Z:
3.3 lbs (w/ standard battery)
12.4" width
8.3" depth
1.0-1.3" height
M11x:
11.25" width
9.19" depth
1.29" heigt
So yeah, roughly the same size, one has a larger screen and better CPU, but the other is half the price.. -
Eggs Scrambled Notebook Evangelist
Exactly! See that's what alot of people in this thread arent realizing when they think the samsung or sony z is so much better, if they are only 1000 points away from the m11x with their top of the line i5 laptop processors, imagine what will happen when we alleviate the cpu bottleneck by only a couple hundred mhz. It's going to be awesome benchmarking this thing
-
chewietobbacca Notebook Evangelist
I never said that about the kid. I said that it's an example of where 3dMark06 is NOT an indicator of real world performance - i.e. the scaling of the scores is not a reflection of real world performance!
And you just hit the point we've been talking about - the CPU is the limiting factor in 3dMark06! It's heavily CPU dependent, and won't necessarily reflect real world in-game performance! -
Here are the two volumes (with math):
Sony Z (from website): 12.4" (W), 1.3" (H), 8.3" (D)
Sony Z volume: 133.796
Alienware M11x (from website): 11.25" (W), 1.29" (H), 9.19" (D)
Alienware M11x volume: 133.369875
Not only is the M11x the same thickness as the Z, the M11x has an overall smaller volume (and as Zlog pointed out earlier, no one measures by volume so it's really just moot). Please stop saying that the Z is smaller than the M11x; it's not true. -
The correct way of saying it would be "the CPU CAN be the limiting factor".
If you are running dual ATI 5870's on a single core processor, yes, the CPU is the limiting factor. If you are running a core i7 extreme at 4.0ghz (quad) with onboard graphics, your graphics are the limiting factor.
But in most situations, I would be willing to bet that the GPU is the limiting factor. Reason I say that is, cause most people are NOT running gtx 280's, or high end gpu's. Just among my friends alone, they are almost all running quad core processors but very old/cheap video cards. No matter how hard you push a CPU, if the graphics card is already fed all the power it needs, its not going to matter much. -
Odds are with a laptop of this size, MOST folks will be using it via Wireless. Alienware did a great job with the wireless networking, using a card that supports both 2.4 and 5ghz bands for super fast wireless N. Seems that to be able to do that, they only included a 100Mbps Wired connection.
Everything in a Laptop is a trade off, and personally based on how I intend to use the M11x I would much rather have high end wireless and Low end Wired connections.
If i'm sitting within cable range of my router, I wont be on the M11x anyway, thats what my XPS 730 with Dual ATI 5850's and 28" Monitors is for
-
Just called that phone # and it says my order is in the "Build Stage" also.
-
Unless your internet connection is 100Mbps, does it even matter?
-
Eggs Scrambled Notebook Evangelist
Suffice it to say I think everyone has become accustomed to laptops having great CPUs and the graphics cards are always lacking.
I know whenever I have looked at laptops in the past, like even for the m15x the default GPU is always crap or feels like it could be better. Then if there is even an upgrade option it costs 100 to 200 dollars more on top of the cost of at least 1000 for the most basic models with no frills. Not only that the market is saturated with laptops that have only had intel integrated graphics (crap no matter what you wanna say about them) or they have dedicated GPUs that are OK but considering the crappy battery life with ANY dedicated GPU you'd always want to go up to at least a 260m or whatever its equivalent was in the past. If you're going to shoot your battery life in the head and stab it with a stake in it's heart you may as well go the full monty and get some good gaming going on.
I think whats most special about the m11x is that it finally brings an option for a good gpu but a comparatively lower-end processor. All in a package that seems like it was transported here from the future, could you have ever guessed there would be a better gpu than a 210m with 11" anytime soon?!
Everyone being accustomed to the way laptops have performed in the past I think are completely unused to the way the m11x is going to be. As much as it seems obvious the 3dmark scores in particular wont see an improvement, I'm predicting it will be crazy drastic how much performance we'll see from overclocking. Then everyone will realize what makes the m11x so unique! -
Nope, Vaio Z is only 1.3" in the back by the hinge it is 1.0" at the front, it is smaller
And may I ask how you measure size without determining the volume?
-
Says im in production on the site, whats the number to check the build status?
-
Eggs Scrambled Notebook Evangelist
I think the gigabit networking is way ahead of its time as it is. Not only that, who gets a 11" extremely portable laptop and not only wants it to be connected to an ethernet socket the majority of the time but also can even take advantage of gigabit networking in regards to internet speed?
About the only thing it would have helped is multiple computer file transfers and even then it would be hilarious if this was the reason someone didnt get this laptop. -
1800 283 2210 then I think hit 3,2,1,1
-
I never really understood the gigabit ethernet not even sure what it's useful for other than in a business server environment with multiple computers all accessing at the same time. Not very many people can even tax a 100mbps connection let alone a 1000.
-
When you go to stick your Vaio in a 1" tall opening (a bag, box maybe) tell me if all of it fits in.
Consumers think of how big a laptop is by its biggest measurement, not its smallest.
And it really depends on your preference, but a laptop's size is measured by the overall footprint left on your desk (which, arguably, would be the width and depth). So the Z is 1.15" wider than the M11x while the M11x is .89" deeper. All this is really trying to argue is that the size is almost negligible so I would recommend concentrating on the hardware specs. -
If you hear music as soon as the phone picks up, it's option 3,2,1,1. If you don't, a different machine picked up and it's option 4,2,1,1.
-
is that a UK number?
*OFFICIAL* Alienware M11x Owners Lounge - Part 2
Discussion in 'Alienware' started by steveninspokane, Feb 2, 2010.