A previous article of ours “nVidia alter benchmarks for Fastest Mobile GPU Crown” claimed that nVidia had altered benchmarks to show that the Alienware M17x with dual GTX 280M SLi graphics was the world’s most powerful mobile GPU setup.
The Asus W90 using AMD’s 4870 mobile crossfire setup was the contender with it’s supposed “one year old” drivers, whereas nVidia’s setup was using the “latest drivers”.
According to sources, different CPU’s were run in the comparison as well, using the faster Intel 9300 CPU in the Alienware nVidia system compared to the slower 9000 CPU in the Asus W90.
Upon speaking to Brian Burke (nVidia’s PR spinner), we was told that it’s not their fault that the Asus/AMD competitor came with that particular CPU, as that is all that is available to the consumer. The W90 used to have the 9600 CPU, but is now deemed obselete which is common knowledge Burke explains. All W90’s now include the 9000 CPU. nVidia’s fault? I think not.
We were also told that the drivers for the GPU were the set that were on the system when it came in. AMD does not offer notebook drivers to consumers, as NVIDIA does. A trip to the AMD and Asus website will confirm that the driver tested is the one consumers have access to, which if they were around a year old, go blame AMD. It’s not their fault that nVidia update their drivers and AMD do not. Due to this, it seemed that the second 4870 wasn’t being engaged due to its drivers not supporting it. Whose fault? AMD.
Maybe they are telling us to install hacked or beta drivers, but AMD, Asus or any other manufacturer in the world would tell you not to use such drivers, and why on earth should you have to?
I think the below screenshot says it all, directly from Brian Burke and the Asus test machine.
The plot thickens.
You heard it here first.
Discuss this further here in the eTeknix Forums
Success isn’t permanent, and failure isn’t fatal.
consequently, the benchmark is worth only for marketing purposes. it’s just laptop vs laptop, not GPU vs GPU
1st, where is this review? I’d like to see what everyone’s talking about.
2nd, These companies don’t cook their benchmarks, they don’t have to. They run 100’s of benchmarks so they have lots to choose from. The good one they send to the press and the bad ones they send back to R&D to patch the hardware / software so they don’t look ban next time. (OK, so they may cook their drivers…)
Howaboutthis,
You are correct. Somebody was going to complain no matter what. And all the complaints would have some validity frankly, depending on your perspective. And this is the case most of the time, with most arguments.
The reality is that both Nividia and ATI blow a lot of smoke and both are constantly providing biased info. My hope is that the educated reader will recognize the bias and tricks they play so they can determine the truth.
In this case, in my opionion, Nvidia is effectively deceiving the reader, or at least providing the POV that benefits them most. This is fine for Nvidia, but not fine for the tech/gaming press (eTeknix in this case), again IMO.
It seems people complained because the second 4870 doesn’t get utlilised.
Maybe Asus are to blam for using dual 4870’s when they knew that there was no driver to support it.
Maybe AMD are to blame for having dual 4870’s possible?
Maybe nVidia are to blame for exploiting this?
If they used beta drivers, people would have complained. If they used hacked drivers, people would have complained. If they used desktop drivers on the laptop, people would have complained. What is left? They use the standard available driver (may be 1 yr old but that is what is available),…. omg people still complain.
Andy,
Any gamer worth his or her salt, especially any gamer who forked over for an SLI/Crossfire notebook, is going to be running the latest beta drivers. Frankly, running the latest betas is a comnon practice.
If the latest beta drivers aren’t available or still slow then you/Nvidia have a point.
My opinion is that Nvidia’s response is self-serving and disingenuous, because it ignores common-practice and typical consumer behavior.
At best you can say using the 1-year old drivers is legit for the uninformed user… but these are exactly that would never see the benchmarks to begin with.
I look forward to your response.
Unless the hardware was identical apart from the GPU’s it wouldn’t really be a valid comparison anyway…
I say it is what it is. If it’s faster then that’s the end of it. Surely it doesn’t matter what drivers are used, whether they’re publically available or not. The comparison was between two machines and the nvidia was faster.
Currently awaiting response from AMD. Watch this space.
I’m with Jeebee on this one. I totally agree. What does AMD have to say about all this? That crossfire is not enabled for mobile 4870 at all? Something is amiss here, for sure.
Surely the product should be recalled, as the Crossfire that it is advertised with doesn’t work.
That, or AMD’s standard drivers do actually work when installed, and they don’t need a “special” laptop driver download. Sadly it seems that no website has the balls to actually try this out.
You only have to stop and think “why include two AMD GPUs in a system if the second will never get used?” to realise that NVIDIA are making up some stories here.
What is AMD’s position on this?
Not nvidias fault. LOL! Those guys are amazing…
JoJo, drivers make or break a GPU and, in the case of the W90’s 4870 CrossFire configuration, they break it.
And lets not forget about the CPUs used either. Making this an apple vs orange benchmark test. In either case the results aren’t worth anything.
If this was a review about drivers and not about GPU benchmark results I would agree. However, that’s not the point of the review. It’s to show which GPU(s) are faster. Not which drivers are up to date. That’s were the problem is found. In all you can’t spin it if the drivers weren’t the fore front of the review.
Every company sends beta drivers out, but they always mark that they are exactly that, and they do generally tell you that you should only be using them for “testing purposes”.
The point they are trying to make is the fact that you can go onto the nVidia website and download the latest mobile graphics drivers. Can you do that with AMD/ATI? No.
According to nVidia, they are the latest drivers which can be found on the nVidia website. Believing it or not, is the consumer’s choice.
Then why does nVidia send unreleased drivers to websites with products they want tested?
nVidia will place a non-consumer driver on their partner site for use in reviews. most of the time these drivers will never see the light of day.
what driver version did they use on the Alienwre? I would be willing to bet they were not the ones available on nVidia’s website.