NOW! read full review about D900C/NP9260

This is my first post.... nice to meet you all

I am not sure how good PC review are...

but here is the link

http://www.pcmag.com/article2/0,1895,2138763,00.asp

they are reviewing Eurocom D900C Phantom

 

That's a fair review, although it didn't give me the impression that they used it much other then benchmarks, and I honestly don't care about synthetic scores.

Thanks for sharing the link though.

 

The review did not seem too extensive or in-depth.

I feel that Ziff Davis companies and Toms Hardware to be biased and/or inexperienced in some areas.

I will wait reviews from sites/places like HardOCP, MaximumPC, etc...

 

40 some odd minutes of battery life while watching a DVD ehh? I hope the extreme processor uses more juice the the one I ordered.

 

I thought the review was good myself; the writer gave it a thumbs up from what I could tell.

I'm wondering how loud those four fans really are.

 

Confused about their benchmarks... why'd they do 1024x768 for the 3dMark? If the "native" is the 1280x1024, that's sad that it only got 5795... mine hits 5485 and doesn't have sli (gpu not overclocked either)..

 

SLI really isn't that big of an improvement most of the time. I think the last couple reviews of the technology I read estimated only ~7% average improvement was common (Your numbers above show about 6%). Where I've heard it shines is that it doesn't bog down like a single GPU...

Think the same as a dual core.. without multi threaded software designed to take advantage, it's not really much improvement. It does however take more to bog it down since the two CPU's can run separate processes concurrently.

Multiplying technology generally only multiplies the amount of work that can get done at once, not how fast the job is finished.

Applying the above statement to SLI... If you send a frame to render, it can only go to one GPU for crunching.. it's in, then it's out, and you move on with your life to the next frame. Good use of SLI would be to alternate the render-ahead buffers so that odd frames are handled on one card, and even on the other (think RAID 0), but instead they just split the work up somewhat randomly between physics, light and pixel shader and whatnot, technically creating more of a bottleneck since now the two cards have to synchronize part's of a single job. (To render a frame, you have to draw the wire frame, apply texture, then ambient lighting, then physics, then directional lights, then AA and AF, this doesn't included the Disk/RAM/CPU I/O to shuffle things about)

Applying the above statement to a Tuna Sandwich ... If you send an order to have a tuna sandwich made, in an SLI environment it would be like taking two people and performing the following:
Person 1: Get bread from storage
Person 2: Get Tuna can from storage
Person 1: Get stuff to put into tuna mix
Person 2: Open Tuna and dump into bowl
Person 1: Dump other stuff into bowl
Person 2: Mix Tuna
Person 1: Apply Tuna mix to their slice of bread
Person 2: Apply Tuna mix to their slice of bread
Person 1: Give half of tuna sandwich to customer
Person 2: Give half of tuna sandwich to customer
Waiter: Assemble sandwich and provide to customer

Person 1 = GPU 1, Person 2 = GPU 2, Customer = Hi-Res Display, Waiter = CPU, Storage = HD/RAM I/O, bowl = reserved memory space where crunching occurs.

(The point is that a single GPU has less steps involved, which is where SLI get's hurt: e.g. no Person 2, however, as you give things to person 1 to get done, they will fall behind and your FPS will drop, this is where person 2 can pick up the slack and keep the ball rolling)

Fixing this will take a collaboration in the industry though, because the biggest culprit is the DirectX team. I guess the effort to branch the single card project to a multi-card project isn't something MS is willing to invest in.

 

Umm by what I can tell that's with 4xAA on? I don't think that is native? I believe the first score is supposed to be native and he just typed it wrong.

 

Good info, thanks... but it's still disappointing. The machine still looks "sick". It should be benching at least 6800 like other 7950GTX & Quadro 2500M sli rigs. With the inclusion of the 2.93GHz Extreme6800 it should have it spanking our older machines if everything is running right. It barely beat my T7400 rig... Something had to have been wrong.



Any other reviews out there?

 

If you were talking to me, I think you don't understand what I meant.

The first score of 9282 is actually with the default settings of 3dmark06. The second score I believe is the at the LCD's native resolution with 4xAA on. However I didn't write the review so I can only speculate.

I have not see another review, just one other guy that has one here... says his got 10,000 something in 3dmark06 default. No screenshots though so who knows.

 

We are working on the 9260 currently and will have some more detailed info posted in the coming days. With a single video card in Vista on default settings in 3D Mark 06 we achieved a score of about 5400. With Dual SLI Video Cards in XP with the same settings we were getting about 9200.

We have been overclocking it and testing some new processors with it and hope to post all of this new information in the coming days.

 

Some of the performance results don't really make sense to me, how is it possible that toshiba satellite P100 (single 7950GTX) was able to beat the D900C at company of heroes with fps of 42.3 vs 14.9?!?

I'm also surprised that the HDX-9000 with the new santa rose T7700 was able to beat the D900C with X6800 at a couple of benchmarks... pretty darn impressive really... or is that difference attribute to some other factor (dx gpus)?

A score of 9282 at low resolution would place the D900C just right below desktop systems with GeForce 7900 GT SLI, but that's not on default settings as suggested by Justin.

 

The results for Company of Heroes seem odd... although it must be noted that it was tested at the Native Resolution for each laptop, which is only 1680x1050 for the Toshiba and the HP, but 1920x1200 for the D900C. Even so, 14.9 FPS still seems too low.

Ah well, nice review; now if they'd just hurry up with the SLI 8700M GT's so I can order my beast.

 

me thinks that someone did not get optimized drivers, or did the XP Multicore hotfix.

 

Hmm, with the Clevo M590K 19" SLI for comparison...
I'm going to do another 3dMark06 score, but at default settings in 3dMark06 I get a bit about 6100 in WinXP / slightly oc'd processor / SLI 7900 GTX 512MB OC'd @ 7950 clock/memory speeds.

A score of less than 6000 in 3dMark06 would be strange, as 3dMark06 weighs in their CPU scores very heavily, which drags my score down due to the Athlon Turion 64 ML44 processor. With a nice desktop processor from Intel, the scores should be much higher at default 1280x1024.

So two questions to clear things up for me:
- Are the scores being provided in 3dMark06 using the demo settings of 1280x1024 res?
- If not, is there a D900C score of 3dMark06 using the demo settings of 1280x1024 res and perhaps some individual test scores within it?

 

yes my test were ran with the download demo version 3dmark06 score: 9107

http://forum.notebookreview.com/showthread.php?t=125244

 

I actually have spoken with Cisco Cheng (editor of PC Mag that reviewed it) before.

Most of Ziff Davis [as I was told] rarely gets a chance to test custom/boutique brand notebooks to review. So their experiences with them would be average and not as in depth as with an enthusiast reviewer. They only get to review the notebooks that are available to them... mainly the typical OEMs (Dell, HP/Compaq, Sony, Toshiba, etc..)

 

Wow, what a monster machine. The one thing I'd complain about is the "smallish" 17" display in a 14.5 lb laptop.

 

Justin says I should get my new lappy (NP9260) shipped next Tuesday (6/5/0) I'll dual boot the puppy Raid 0 with XP Pro on the first two drives, and Vista Ult. on the third. Is there anything you "all" can advise me on as I set this baby up. Optimized drivers and XP Multicore hotfix, anything else???:

 

Here is another review:

http://www.pclaptop-review.com/2007/06/04/eurocom-d900c-phantom-laptop-review/

 

Torque, just have the unit shipped to me for a week or two.. I'll get it setup and uhh.. "test" it's stability while gaming for you

Since you got all that money towards this system (hopefully your rig didn't cost your full budget), maybe you could be brave and try a Q6600 for us? Check with Justin to make sure it's OK on the warranty if you can (I'm sure it is, but check anyways)

 

That isn't another review, it is just a cheap advertisement! A review should not link the company to pawn off the systems, it is not exactly good practice. A side from that, all it does is highlight the features. I wish a more in depth review was available.