Incorrect review testing

[LukeTN50]

Have just been looking at some gpu reviews as i am currently on the hunt for a new card (just posted a thread) and just noticed a MAJOR mistake with the testing at OC3D (TTL). He has changed test setup yet his results are the same!!!

How can an i7 3960x @ 4.5ghz with 16gb dominator platinum on a rampage 4 get the EXACT same results as an i7 950 @ 4ghz 6gb mushkin on a G1 assassin????

Check out the MSI GTX 560 results, original review scores are included with the new GTX650ti results:

Old review test setup: http://www.overclock3d.net/reviews/gpu_displays/msi_gtx560ti_448_core_review/3
New review test setup: http://www.overclock3d.net/reviews/gpu_displays/nvidia_gtx650ti_boost_review/3

Now check out Batman results for each review: they are exactly the same. I imagine the rest of the results are the same too when you do comparisons...

How can something like this be trusted? It's very misleading....and obvious cutting corners.

 

[andybird123]

your links are broken

also, he does in fact make reference to the fact that old results are... old results, on older drivers, hence why he mentions that the 650ti boost on current drivers scores higher than the original Nvidia sample GTX 680 did on the drivers it had at the time

no, TT does not retest every GPU everytime he does a new one, but as far as I can see he is quite clear about that

 

[rankftw]

Links are broken.

 

[LukeTN50]

sorry guys fixed links!!

 

[LukeTN50]

your links are broken

also, he does in fact make reference to the fact that old results are... old results, on older drivers, hence why he mentions that the 650ti boost on current drivers scores higher than the original Nvidia sample GTX 680 did on the drivers it had at the time

no, TT does not retest every GPU everytime he does a new one, but as far as I can see he is quite clear about that

I don't think so buddy. There is nothing about using an old driver and nothing about a different platform there which actually shows very inaccurate analysis. It's very misleading.

Drivers you can forgive because reviewers are forced to use new drivers on new cards but using different cards for comparison on different platforms is very poor. They cannot be compared. Apples with apples, pears with pears.

 

[panyan]

also, he does in fact make reference to the fact that old results are... old results, on older drivers, hence why he mentions that the 650ti boost on current drivers scores higher than the original Nvidia sample GTX 680 did on the drivers it had at the time

no, TT does not retest every GPU everytime he does a new one, but as far as I can see he is quite clear about that

+1

He mentions this in numerous videos

 

[Rroff]

Assuming its the same GPU and same GPU clocks that wouldn't be entirely unusual - for a lot of benchmarks my Q9550 @ 4GHz puts up identical framerates with 470 SLI as a highly clocked i7 3770K. Just depends where the CPU and GPU performance hits a wall.

 

[andybird123]

I don't think so buddy. There is nothing about using an old driver and nothing about a different platform there which actually shows very inaccurate analysis. It's very misleading.

Drivers you can forgive because reviewers are forced to use new drivers on new cards but using different cards for comparison on different platforms is very poor. They cannot be compared. Apples with apples, pears with pears.

he mentions it all over his site, in the review you've linked to he mentions it again at the top of the arkham city page