Marketing and Prejudice Get the Better of Consumers with PC Processors: Test

[btarunr]

At the AMD & HardOCP Game Experience event held in Texas, gamers were asked to participate in a blind test. The test involved gaming on two sets of gaming PCs with two PCs each, in each set is an AMD-powered PC, and an Intel-powered one. Participants weren't disclosed which PC was driven by what, as they were assembled in identical-looking cases (no window), with identical monitors and other peripherals. The first set is of budget single-monitor HD gaming, while the second set is high-end three-monitor gaming.

After gaming on both rigs in each set, respondents were asked to tick on a sheet of paper, which rig gave them a better gaming experience, or if gaming both had no observable difference. AMD went into this exercise expecting that most respondents will select "no difference" as their option, and so that would bring good PR to AMD, but to their surprise, most respondents selected the rigs that was powered by AMD processors.



In the budget single-monitor gaming machine, the AMD machine (system B) was powered by AMD A8-3850, ASRock A55 chipset motherboard; the Intel machine (system A) was powered by Intel Core i3-2105, and ASRock H61 chipset motherboard. The goal was to configure the PCs to cost under US $500. Both machines were made to use CPU-integrated graphics Ofcourse the respondants were not told which machine was driven by what. The results are as follows:

• System A (Intel Core i3-2105) better: 5 votes
• System B (AMD A8-3850) better: 136 votes
• No difference: 2 votes

With the high-end gaming machines, the Intel machine (system A) was powered by Intel Core i7-2700K, with ASRock P67 Fatal1ty motherboard; the AMD machine (system B) was powered by AMD FX-8150 with ASRock 990FX Fatal1ty motherboard, both machines were given Radeon HD 7970 graphics driving 3-monitor Eyefinity. Again, the respondents didn't know which system was driven by what processor. Results are as follows:

• System A (Intel Core i7-2700K) better: 40 votes
• System B (AMD FX-8150) better: 73 votes
• No difference: 28 votes

The results of this little experiment modeled along the lines of PepsiCo's famous blind-tests of the 1970s shows that marketing and prejudice get the better of consumers, at least in the case of PC processors.

View at TechPowerUp Main Site

 

[amd/atifiend]

nothing new here

it's all about marketing and displays which of the 2 companies excel at it.

 

[H82LUZ73]

They should try this at E3 or some trade shows ,I bet the out come will be the same.Also 3 monitors is better then one cpu clock for clock.

 

[MikeMurphy]

I can't quantify it but my Llano setup impresses me and many others who have seen it in action.

Otherwise, I would simply chalk up the first test to better image quality.

 

[silkstone]

For the hi-end machine the sample size is far to small to show anything conclusively.

 

[MikeMurphy]

I thought the sample sizes were reasonable.

I'd be really interested to know WHY they thought the high-end AMD system was better

 

[btarunr]

Some might argue that since the respondents might have known that at least one of the two rigs had Intel, it might have had a "placebo effect" on them, influencing their choice; but then the results are clearly in favor of the AMD rig in both tests, so that could not have been a factor.

 

[xenocide]

For the hi-end machine the sample size is far to small to show anything conclusively.

Definitely. I think (from the picture at least) on the 3-Screen setup the System B is a little more spacious as well. Could just be the angle, but it looks like the guy playing at System A is a little more cramped in. Things like this do affect people perception.

I think the Low-End system is a given though. The AMD Machine was overwhelmingly picked because the Fusion Platform has way better integrated GPU's. Nobody is really surprised by that fact. The only way they could prove marketting prejudice is if they did 2 groups of equal size, and told 1 which setup was which, but not the other. If the one that was told skews in the opposite direction as the one that wasn't, they are clearly having clouded judgement.

 

[Sir Alex Ice]

Blind test with PCs is not possible, they can find out with a simple Dxdiag command which CPU is in which system. So the test is not relevant, unless they do it double blind and in a controlled environment so that they really do not know and do not have a way to find out which system is AMD powered.

 

[reverze]

no surprise AMD is the better choice. shame for those people who spend so much money on rigs just to benchmark.

 

[HumanSmoke]

Definitely. I think (from the picture at least) on the 3-Screen setup the System B is a little more spacious as well. Could just be the angle, but it looks like the guy playing at System A is a little more cramped in. Things like this do affect people perception.

Pure marketing stunt. Apart from the subtle setup differences, there's no way you can eliminate observer and control bias in this scenario. For the results to mean anything, you'd need both everything verified identical (including software in the backround), ABX testing ( 1 x Intel, 1 x AMD and 1 x random Intel or AMD rig) and supervision from someone who did not know the setups (double blind testing).
Call me cynical but I've learned lately that AMD's viral marketing needs to be taken with the contents of the Bonneville Salt Flats (Did anyone check to see if John Fruehe was in attendance?)

I think the Low-End system is a given though.

I'm surprised that the Intel system garnered 5 votes...which kind of throws the other results into a dubious light

 

[xenocide]

no surprise AMD is the better choice. shame for those people who spend so much money on rigs just to benchmark.

My i5-2500k is cheaper and outperforms an FX-8150 in about 90% of gaming benchmarks. I don't feel cheated in the slightest.

 

[thereturnfc]

Blind test with PCs is not possible, they can find out with a simple Dxdiag command which CPU is in which system. So the test is not relevant, unless they do it double blind and in a controlled environment so that they really do not know and do not have a way to find out which system is AMD powered.

No surprise , these word means that you never done any experiment/LAB. otherwise...what is the purpose of this test to begin with?!?!

 

[v12dock]

System A (Intel Core i3-2105) better: 5 votes
System B (AMD A8-3850) better: 136 votes

No kidding the A8 would give a better gaming experience

 

[ViperXTR]

wtf 5 votes on i3 2105 lol i was expecting zero/nada.

 

[badmau5]

I'd probably buy this if test was made by 3rd party, but @ the event hosted by AMD - meeh. And what drivers they used for both machines? personally im not sure if both pc's were identical other than cpu/mobo. Looks shady to me. What i know for sure from other independently made reviews is that FX8150 bottlenecks 7970 performing worse than nehalem quads.

no surprise AMD is the better choice. shame for those people who spend so much money on rigs just to benchmark.

lol yeah but dont forget that those "stupid" benchmarks actually measure performance.

 

[HalfAHertz]

Erm it was an AMD show... Let's think a bit, what are the majority of people that usually go to these kinds of shows? That's right - fans! So if AMD fans go to an amd show to see some AMD systems and also have the chance to play on said systems, they'd feel happy and compelled to root for the people that gave them that chance, no?

 

[btarunr]

Erm it was an AMD show... Let's think a bit, what are the majority of people that usually go to these kinds of shows? That's right - fans! So if AMD fans go to an amd show to see some AMD systems and also have the chance to play on said systems, they'd feel happy and compelled to root for the people that gave them that chance, no?

But then they didn't know which system was AMD, which was Intel. It was a blind-test.

lol yeah but dont forget that those "stupid" benchmarks actually measure performance.

Yeah, and this test shows that outperformance in benchmarks means squat as long as games are smooth enough.

 

[ZoneDymo]

Erm it was an AMD show... Let's think a bit, what are the majority of people that usually go to these kinds of shows? That's right - fans! So if AMD fans go to an amd show to see some AMD systems and also have the chance to play on said systems, they'd feel happy and compelled to root for the people that gave them that chance, no?

Are you serious?
And if you are, I am at a loss for words other then "please learn to read".

 

[MikeMurphy]

I'd probably buy this if test was made by 3rd party, but @ the event hosted by AMD - meeh. And what drivers they used for both machines? personally im not sure if both pc's were identical other than cpu/mobo. Looks shady to me. What i know for sure from other independently made reviews is that FX8150 bottlenecks 7970 performing worse than nehalem quads.

You're assuming vsync is off.

With it capped at 60fps is there really much of a perceived difference?

 

[naoan]

Erm it was an AMD show... Let's think a bit, what are the majority of people that usually go to these kinds of shows? That's right - fans! So if AMD fans go to an amd show to see some AMD systems and also have the chance to play on said systems, they'd feel happy and compelled to root for the people that gave them that chance, no?

My thought exactly.

But then they didn't know which system was AMD, which was Intel. It was a blind-test.

Easy, pick the one with more obvious fps drops from time to time.

 

[the54thvoid]

Like others have pointed out it's fairly obvious AMD would win a cpu gfx test. They have the best on chip gfx by far of the two companies.

As for the bigger test, using an AMD gfx card (7970) with either an Intel chip or an AMD chip isn't a very good test scientifically. It would need various combinations with AMD cpu + AMD gfx, AMD cpu + Nvidia gfx, Intel cpu + AMD gfx and Intel cpu with Nvidia gfx. That gives four options and an analysis of variance would provide more insight.

All that being said, i dont think there would be a perceivable difference bewteen cpu's using a 7970 (unless there is inherent programming bias from the same brand combo). The sample size is too small and also, if Intel did the same test and came up with the same results (but favouring themselves - on the expensive rig) we'd all be suspicious too.

They shoulda stuck a phenom II in there instead of the Intel but they'd have probably known what would've happened. (Point being, despite brand loyalty they are still trying to flog BD, therefore using a Phenom II might have upset that sales trip).

 

[NC37]

wtf 5 votes on i3 2105 lol i was expecting zero/nada.

Meanwhile in Intel's impenetrable fortress of Core doom...

"Did you five succeed?"

"Yes master. We successfully sabotaged AMD's event by casting our votes for the i3 system."

"Excellent! Those five votes will be the start of our revolution!! If people will stop and think about why those five chose the i3...they can be made to think anything!!!"

"But, we lost on both tests!"

"Yes well...out my sight! I don't pay you to think! Go back to IGP R&D!"

"Oh boy! Hey guys, lets celebrate with an wafer fight!!"
"Yay! Silicon on silicon action!!!"
"Dibs on HD3000 wafers!!"
"Awww, no fair those fly better!"
"Oh I am so pulling out my GMA950 ninja star wafers!!"

And then...

"Useless engineers. They'll never see the grand picture! I will show them with my evil laugh!"
"Muhhahahahahahaha hahahah ahahahahah haaaaaaaaaaaaaaaaaaaaaaahhhahaaaaaa!!"

 

[HumanSmoke]

Win-Win for AMD in any case. If the majority of people pick i7 over Bulldozer then it's not really earth shattering news, and that isn't the case then I'm sure the guy that does their PR/waters the office plants gets some good copy.

How much (if any) difference is largely going to come down to the game being played. AMD and their fans are very quick to point out BF3 and F1 2011 as being "representative" of the overall abilities of each CPU (games that along with AvP don't really care what CPU is in the system...anyone care to guess what the people were playing?)....I'm pretty certain Civ 5 wasn't on the menu.

AMD should have been ballsy...pit Bulldozer against i3 2300 or an 1100T and see if anyone noticed the difference.

 

[Bundy]

This test would be very interesting/compelling if it was done by someone independant. As the info was provided by their marketing dept., I don't believe.