Running a Control with benchmarks

[mallenwho]

Something i noticed with the R7 240 'review' was that there really should be a control on those benchmark graphs. a bar for a rig with no graphics card in it, so we can see how much of a difference each card makes to the rig, along with the difference in how much of a difference it makes. Perhaps the R7 240 was bottlenecking the computer and it was actually getting worse performance. I don't know. There wasn't a control.

 

This is something I think should be implemented in all benchmarking. You conduct them like a scientific experiment (big thumbs up) as to how you run them, always doing the same thing etc. However, something that every experiment needs and you don't have is a control.

 

I posted it on youtube and it got 10 thumbs up with people commenting that they agree, but with new youtube comments you'll never find out. Following your WAN show advice (take it to the forums) i've decided to post it here too.

I don't understand how forums work so i'm putting it here and hope it's right. I also hope that Luke or Linus or someone important can see this and do something about it/redirect it to the appropriate place.

Thank you =)

[upliftsleet1974]

Not all that bad of an idea....I like it!!

[rhyseyness]

@LinusTech @Slick

I would like to second this

[LOLPhilips]

What do they use as the benchmark rig ?? High end CPUs ?? Socket 2011 maybe AM3+.. There is a lot of reasons this wont work..

In other words unless they use a cpu with intergrated graphics this isn't that good an idea.. Besides you have a lot of previous results for comparison and can draw a conclusion that way..

Why would you want to compare a dedicated graphics card to a integrated one, and not just to its general competition..

 

If you're a real gamer you're gonna use a dedicated graphics card anyway..

If not you probably build the pc as a HTPC or for your parents or something..

[SirReallySam]

Following your WAN show advice (take it to the forums) i've decided to post it here too.

[ofr057]

also a point should be made is matching gpu's to the cpu's within the same price point .... i dont think someone would have a 1000 dollar cpu and a 200 gpu

[mallenwho]

@LOLPhillips I meant specifically for this video, to see how much over nothing the card really helps. As more of a satirical thing than anything else.

 

Although it could be interesting to see how well different games run on APU or just CPU, which allows some stingy buyers to realize that for the games they like they don't need to invest in a graphics card.