i7-720QM + 5870 CF, 3dmark06 = 11578?!

Hello everybody, bought a M17x recently, and tried to benchmark it with 3dmark06 basic version, the result is really heartbreaking... (no OC, factory setup)

I have read a lot of reviews with exact same setup as mine to reach around15k while my first bench is 11.5k?!

The ORB link is here ORB - World of Performance

3DMark Score 11578
SM 2.0 Score 4294
SM 3.0 Score 6116
CPU Score 2851

Both CPU tests went with 0-1 fps, I guess that's the reason of low score aye? This m17x R2 has i7-720QM, 8 Gb x 1333 mhz RAM, Raid 0 1 TB, 5870 crossfire, Catalyst 10.6, etc.

Could any1 please give me some advice about further benchmarks or possible problems this laptop might have? Thank you very much in advance.

 

Download GPU-Z and ensure that your system is running in crossfire. I just did a run benching my i5 520 and scored over 13K. My guess is it's something to do with your GPUs.

As far as the 0-1 fps for the cpu test thats normal.

Also make sure your not upping any settings and running 3DMark06 on the default settings.

 

yeah it sounds like you got stealth mode enabled. the cpu score isnt going to be high the 720 is a weaker cpu, it runs pretty slow ghz wise, and 06 is a CPU lovin benchmark. i think in stock mode you should get low 15s easy. check out to see it your x fire is enabled like DR650SE said, and make sure your not in stealth mode.

 

Double checked, it is indeed running with crossfire enabled, and I just downloaded 3DMark 06 basic version, so I am not allowed to up any setting at all. I wonder if it will do better with 10.7...

 

ORB - World of Performance second try, same old... stealth mode is disabled...

 

Have a look at the power settings and ensure its set to high performance and that the gpus are set to max perf instead of power saver.

Also, check that catalyst ai is enabled in ccc 3d menu.

 

Yeah to above, and ai is enabled, but I put it in standard instead of advanced.

3DMark Vantage results:

ORB - World of Performance

Any comments welcome, thank you again for your help!

 

run gpuz in the background as you do the benchmark. after it finishes check the max clocks recorded for each gpu on the sensor page. this will tell you if one isn't being used or if they arent hitting 3d clocks.

 

Both gpu checked with GPUZ, whenever the benchy starts, core becomes 700 and memory 1000, for the duration of each test (1 min+), temp. gets to 70 and GPU load is usually around 30-40%. That's normal right?

 

Called Dell support, their answer is that this performance issue is not a tech issue, thus they can not give further instructions... Damn, what can I do now? Open the case and check for the cables? Any other hints please?

 

I just tested it with crossfire disabled, here are the scores:

3DMark Score 10889 3DMarks
SM 2.0 Score 4358
SM 3.0 Score 5283
CPU Score 2817

Only SM 3.0 Score suffered a minor drawback, does this mean that my crossfire is not working as intented?

 

Have you tested any games to see if performance is correct there? If it isn't working correctly in games then it most certainly IS a tech issue and it's their responsibility to sort it out.

 

Hi Noche,

Furmak products are qaint they remind me of games I use to play a decade ago. Your card is Dx11 not Dx10 and Furmark products also look for PhysX which is highly outdated at this point unless you are a Nvidia customer. However if you want a higher score that is simple enough, change your Catalyst Control Center settings. If you are looking for an up to date benchmark try Unigine as it utalizes Dx11. Any game properly optimized for Dx11 will run much more efficiently than games on Dx10. Again change ccc settings and you can fool the benchark higher. FurMark with 5xxx series cards is a classic apples/orange comparison. The real test is how many FPS can 4xxx card get on Dx11.....00000000000000000000000000000000!


AMD Cypress based cards, implemented a hardware solution to protect the VRM. They dedicate a very small portion of Cypress’s die to a monitoring chip. This chip monitors the VRM. If the chip detects a dangerous situation (current overload), the chip will immediately throttle back the card by one PowerPlay level. Once the dangerous situation has disappeared, the card goes back to its performance level. With FurMark this back and forth continues as the VRMs permit.


P.S. Dell keeps a record of all your service calls and why.

 

First of all furmark is a separate program and nothing to do with this case.
Second 3dmark doesnt push the components nearlly as hard as a furmark run so it's unlikely that they would get pushed to the temp limits unless it's sitting on a bed or something.

I dont think 3dmark 06 uses physx, i know vantage does but so far he's referred to 06 mainly.

Also with his system in 06 (even though its not the best benchmark anymore) he should be getting between 15.5k-16.5k in 06 not 11k so something isn't happening right.

 

This is funny. I want you to run metro 2033 in dx11 at max spec and see what kinds of fps you get.

 

My apologizes for the confusion, was in a hurry this morning ment to type "FutureMark products". Noche said he was using 3DMark. Yes it does rely on Physx, simply google, Nvidia smokes 3Dmark thanks to physX. Again Dx10 doesn't come close to Dx11 so again we are comparing apples to grapes. Anyone who is currently playing Crysis in Dx10 and Battlefield Bad Company 2 in Dx11 with 5xxx series cards knows what I am talking about. You can (hypothetically) string a dozen 4xxx series along and you will still have 10.1 graphics. Amd have been tweeking their Catalyst 4xxx drivers for near two years. No suprise that they run well on FUTUREMARK products. I said current limits not temp limits although I grant you they do go hand in hand. AMD and Dell are not oblivious to warranty issues resulting from cronic benching and have built protection into the cards. Having a wonderful time playing metro in Dx11. thxs.

 

You still didnt answer the question what are your temps and fps playing metro 2033 in dx11 it is a simple question i get at the lowest 30 fps in metro on my rig and 60+ in the high end on high in dx10.My temps NEVER get out of the 70's on my gpu how about you?

 

Simple? Catalyst settings? Tessellations off/on? Depth of field off/on? AAA? 4x MSSA? Physx on off? Have never really bothered to look at the temps and FPS. Next time I am playing I will take a look at it. This game has Nividia written all over it to highlight the 480's so it never will play to its full potential even with Dual 5970's. From what I have read Metro 2033 is not very well optimized on many accounts. They have only added marginal optimization for Dx11. The game relies heavily on Nvidia Physx. If you don't have a Nvidia product it will run PhysX off CPU based on a lower standard. If you have Nvidia than you get FULL Physx effects off of GPU. Not really a great test for any AMD product. Not many really good Dx11 titles yet except a handful. Crysis 2 won't even be out until Christmas. Running 5870's on an old benchmark is like running a BMW on low octane, performance will suffer.

 

First of all, thank you so much everybody for your kind comments. Allow me to answer some:

Oogamar, I did try some games, but due to lack of time, the most I have got is flickering (like everyone else) with cf on playing certain games, and sometimes game crashes when I have played for awhile on 1920 x 1200 res with all high. But no BSOD nor anything, I had my doubts when I got BSOD with everest ultimate, but the last BETA of it finally works with cf.

FalcomMachV and claxdog, thank you for your comments, could you kindly advice me on some other benchmark to test my PC please? Thank you very much.

Well, at least I have got this with vantage:

3DMark Score P11223
CPU Score 9930
Graphics Score 11732

Is that score ok?

SO... long story short, if we gonna use cars to compare, it is like having a BMW running at a max speed of 120 km/hr, does it work? Yes, Finely? sorta, as expected for the amount of money paid? NO.

I for hell hope dell notice about my phonecall, coz I still consider this not working as it should.

 

Unigine supports Dx11
Unigine | Unigine (3D engine for games and real-time virtual worlds)

3DMark 11 (for Dx11) isn't due until the "third quarter" (technically July 1 through September 30).

GPU makers optimize their products to score well on benchtests. However real world game play is often a far different experience.

 

Noche what are your system specs, games with flicker, and country of purchase, and drivers? Have you patched the games properly. Example Crysis has Patch 1.1, 1.2 and hotfix 1.21. Really old games might not have support for CF but I assume the games you are running do. If you are using stock drivers the game shouldnt be crashing. Are you using any utilities in the backgroung that could be conflicting?

 

Gonna run Heaven benchmark and give some scores asap, let's see how it goes...

 

System specs as stated: i7-720QM, 8 Gb 1333mhz RAM, 1 TB Raid0 HDD, Ati 5870 cf, shipped 4th of july, with latest bios I believe 012.019.000.006.036284 (same as silver's on the other thread). Country of purchase: China, drivers = out of box stock ones, apparently 10.6 (ccc doesn't show it, GPU-Z's driver version: 8.692.2.0/win7 64). No utilities in background, unless you mean daemon tools.

I have some minor problems while using cf with old games as settlers VII (textures blinking), SW: force unleashed, sc ii (that's a known issue). Most of which will go off with CF disabled. Have yet to find a BSOD playing with cf enabled, but of course, I haven't had the time to play extensively (never more than 2 hrs straight).

I will run heaven benchmark shortly, any other benchmark you would recommend me to do? Or maybe some other tools to show up more info? Windows experience showed 7.2 on graphic (aero and 3d). I am rather new to all this benchmark so I get lost easily. Thank you very much in advance.

 

WEI scores lower with Samsung SSD. My WEI is 5.9 because of my solid state drive. It is just the way WEI treats HDD vs SDD in some cases. Some older games simply run better with CF off unless they have special CF drivers. I am not familiar with those games. I am running stock Catalyst drivers, Driver Packaging Version 8.692.2-100203a1-095371C-Dell The only utility I run is Microsoft Security Essentials.

 

Your driver packaging version is exactly the same as mine, and I run MSE + daemon tools, but no more other utility.

On a side note, heaven benchmark scores:

Heaven Benchmark v2.1
FPS: 30.9
Scores: 779
Min FPS: 7.7
Max FPS: 77.4

Hardware
Binary: Windows 32bit Visual C++ 1500 Release May 21 2010
Operating system: Windows 7 (build 7600) 64bit
CPU model: Intel(R) Core(TM) i7 CPU Q 720 @ 1.60GHz
CPU flags: 1596MHz MMX SSE SSE2 SSE3 SSSE3 SSE41 SSE42 HTT
GPU model: ATI Mobility Radeon HD 5870 8.692.2.0 CrossFireX 1024Mb

Settings
Render: direct3d11
Mode: 1920x1200 fullscreen
Shaders: high
Textures: high
Filter: trilinear
Anisotropy: 4x
Occlusion: enabled
Refraction: enabled
Volumetric: enabled
Replication: disabled
Tessellation: normal

As I have stated in the other thread, few days left before I have to live exchanging parts with this R2, so I have to make sure that what I have at hand is a normal (as no tech problem or card malfunction issues like cables etc.) before tweaking or I am out of time, thank you all for your comments.