AMD may be cherry picking review samples

[Dylan522p]

http://www.tomshardware.com/reviews/radeon-r9-290-review-benchmark,3659-20.html

Toms hardware expclicitly tested and showed this. Even Linus' video shows this. I think this is terrible and we really should punish them by votimg with our wallets if it is true. That said we must wait for someone to get a reference retail cards to test this.

[Hans Christian | Teri]

No surprise, everyone else is doing it so why shouldn't AMD?

[Freaky_spider]

NVIDIA does it to, whats your point? that companies never lie? what dream world do you live in?

[Joshyy]

Yeah, like Nvidia doesn't. 

[M-ursu]

Dude, they all do it.

[Kilerbomb]

Doesn't Nvidia do the same thing?

[Guest]

They all do it

[Bluejay0]

Intel, NVIDIA, EVERYONE DOES IT!!!

 

Recently, Intel sent engineering samples of the Haswell chips to reviewers and that backfired since all chips sucked like hell. Overclocking and heat output was monstrous.

[ajoy39]

lol this is the worst article ever and just shows that Tom's Hardware doesn't understand how the card works. 290s are performing the same as 290xs because they have more thermal headroom, and run more consistently at higher boost clocks than the 290x can with the reference cooler. 

[Lukiose]

lel @ the late realization.

Pretty damn sure the 290 is performing as such because they bumped the fan up to 47%, allowing it to remain at desired clock speeds without throttling.

Edit: Totally Ninja'd by ajoy

[TheRuke]

No shit! Every other company does this...

[Alex780]

I don't care if they cherry pick their gpus. At least they will be on an equal footing with Nvidia, who already do it.

[delet my acc pls]

Scoop of the year!!!

[Cacao]

Just because nvidia or anyone else does, it suddenly makes it ok? What happened to getting an accurate representation of the product from reviewers?

[chaozbandit]

A company nowadays that sends out review samples would be out of their minds to not at least do a thorough check on those samples before they get sent out. So what if AMD does it? It's been made very clear to the community that Nvidia does it as well (see TTL's rant on AMD's R9 GPUs in regards to how well the overclock + thoughts on ASUS Matrix cards). That being said, a lot can still happen between the time a sample gets picked and when it actually arrives at the reviewer's doorstep and the actual long-lived performance is always an unknown. Heck, this is even apparent in areas outside of the tech industry. Let's take a brief look at the auto industry.

 

Everyone remembers a while back when that Tesla Model S was sent to a reviewer and it received a terrible score. Tesla Motors would have done checks on the vehicle to ensure it performed, hence why they installed a black box prior to sending it out. Was the bad score a reliable representation of the product? No, it was clear that the reviewer did not know how to handle an electric vehicle in accordance o manufacturers specification. That's a fault on the part of the reviewer; however, manufacturers are not entirely out in the clear either. Take a look at SRT (Dodge, for those who don't follow). They sent their new Viper to Motortrend for a large-scale car review. They knowingly sent them a base car, let alone one they knew was not perfect, and still chose to let Motortrend review it in the competition (it did not perform as theirprevious Vipers and actualyl fell apart / dismantled itself during the test). Is this the fault of the reviewers (Motortrend)? The answer is no.

 

So it's pretty damn clear that "cherry picking" by the manufacturer is evident everywhere (to some degree) in today's consumer society. This should not be a surprise to anyone; just because a company like AMD hasn't made headlines regarding this issue before, does not mean it hasn't happened before. In the end, it always goes back to the vicarious relationship between the manufacturer and they each do their respective jobs in providing the latest products for affiliates to look at such that they can give the public an accurate representation of what they received. It's on the responsibility of both parties to provide what they believe would be best for the end consumers.

[helping]

Just because nvidia or anyone else does, it suddenly makes it ok? What happened to getting an accurate representation of the product from reviewers?

There is no room for morals in capitalism

[Mo5]

So that's what they've been doing during that week, not the drivers

[Cacao]

There is no room for morals in capitalism

Morals? Isn't that like borderline fraud? What if a keyboard maker would give reviewers an amazing version of a keyboard handmade just for them, then sell a crappy one to the public? It's not that much different.

[ajoy39]

Morals? Isn't that like borderline fraud? What if a keyboard maker would give reviewers an amazing version of a keyboard handmade just for them, then sell a crappy one to the public? It's not that much different.

thats not whats happening, and there really isn't a good comparison between GPUS and keyboards. Every single R9-290 is the same chip, on the same PCB, with the same cooler, but silicon performance varies from card to card. They are all built exactly the same, the variance comes from the silicon lottery. 

[Freaky_spider]

Morals? Isn't that like borderline fraud? What if a keyboard maker would give reviewers an amazing version of a keyboard handmade just for them, then sell a crappy one to the public? It's not that much different.

uhm, it only affects OC performance.... not much more, its not like they gave it an extra 300 cores or something...

[RatedBlam]

Morals? Isn't that like borderline fraud? What if a keyboard maker would give reviewers an amazing version of a keyboard handmade just for them, then sell a crappy one to the public? It's not that much different.

Well I'm not going to question someones moral, but what if AMD want's reviewers to fully test the overclocking potential and how good the cooler is and then the card doesn't overclock, what then. But ofc there is more to it but you get the idea, and I'm not sure how to make changes to keyboards that aren't obvious

[helping]

Morals? Isn't that like borderline fraud? What if a keyboard maker would give reviewers an amazing version of a keyboard handmade just for them, then sell a crappy one to the public? It's not that much different.

Nope, samples wouldn't fall under advertisement and I believe technically the results produced by the person with the engineering sample are not officially endorsed and are completely exclusive of the company. It falls under a grey area where they are inadvertently doing misleading advertisement by providing a cherry picked version to an indiscriminate third party which is free to do whatever they want that is not under NDA. That's completely legal and I don't see why it would be illegal.

Trying to police something like that, while it is real, is too difficult and moves too far from neocapitalism to almost socialist (loosely speaking).

Thinking that they don't, and hoping that they don't, and preventing them from cherry picking is too idealist.

[Cacao]

thats not whats happening, and there really isn't a good comparison between GPUS and keyboards. Every single R9-290 is the same chip, on the same PCB, with the same cooler, but silicon performance varies from card to card. They are all built exactly the same, the variance comes from the silicon lottery. 

 

 

uhm, it only affects OC performance.... not much more, its not like they gave it an extra 300 cores or something...

Well obviously my comparison is a pretty extreme example (couldn't think of anything better off the top of my head ). But really, what chances do you have of getting a cherry picked GPU without paying extra for it? Pretty slim. I think reviewers should get a totally random GPU to represent the randomness that consumers would also receive.

[Warcony]

I bet if this article was for Nvidia people would be outraged, but because AMD is everyone's prodigy son these days its cool. Regardless, every company would at the very least, check each review sample.  Sending out a dead sample would be very unprofessional.  Cherry picking is a little unfair but like benchmark cheating, almost every company is guilty. 

[Freaky_spider]

Well obviously my comparison is a pretty extreme example (couldn't think of anything better off the top of my head ). But really, what chances do you have of getting a cherry picked GPU without paying extra for it? Pretty slim. I think reviewers should get a totally random GPU to represent the randomness that consumers would also receive.

yes ofcours, that is true, i even find that reviewers should get equal cards, for obvious reasons