Self-inflicted AMD PR Drama [Nano Fury Edition]

Techreport won't be getting one to review http://techreport.com/news/28971/wanted-for-review-amd-radeon-r9-nano
I suspect AMD is mad since he pointed out the settings used by AMD's internal benchmarks were unrealistic (0x AF for example) in order to make the FuryX and Fury look more favorable than in reality.
Great PR work AMD ...

Interesting. It does seem possible that their little preview blurb the other day may have caused friction. I hope someone else delivers a card to them.
 
Yes, and I think not giving them a card to review is a huge knee-jerk reaction and a mistake, as hopefully there will be backlash against AMD's decision given how respected TR is, as they should be since I don't think they've ever shown bias and they've changed the way people look at GPU benchmarks with their inside the second benchmarks.
 
They were using pre-overclocked nVidia cards in their comparisons without disclosing that fact?
 
They were using pre-overclocked nVidia cards in their comparisons without disclosing that fact?
They told what card they use, but the card that they use is factory oc'ed at around 9%. And no, they didn't mention about the factory oc. I think they should mention that the card is factory oc'ed and also they can mention that 980 is easily oc'ed vs Fury.

Edit: to be fair, I checked 980 cards and most of them sold at least with a tiny bit of oc. On Newegg, only EVGA and PNY sold 980 variant with reference clock. The cheapest 980 Gigabyte variant has around 4.5% oc (vs 9% on the Gigabyte variant TR used). The cheapest 980 card on Newegg (not open box) also factory oc'ed. Not as high as the card TR used but higher than than the cheapest Gigabyte. So if you shop on Newegg, you actually need to pay more to buy a 980 with reference clock if you don't care about brands.
 
Last edited:
It's perfectly okay to put a factory-overclocked card in there because they don't have an obligation to keep all reference cards.

But to use a rather premium Gigabyte 980 G1 version with a 10% overclock, keep quiet about it and then present their graphics like this:

ICzRuUK.gif


... that's some underhanded shit, IMO.
 
Would be nice if reviewers put the clocks next to the names just for quick reference.
 
Would be nice if reviewers put the clocks next to the names just for quick reference.

They put the G1's clocks in the test setup page, but they didn't put the reference clocks for the 980 anywhere in the review, so after looking at the graphics, one could easily think that buying any GTX 980 will result in that performance.
Which it won't, obviously.
 
Eh, well, I'm not gonna memorize all the clocks for every brand of every card when I'm looking at these graphs.
 
Yes, and I think not giving them a card to review is a huge knee-jerk reaction and a mistake, as hopefully there will be backlash against AMD's decision given how respected TR is, as they should be since I don't think they've ever shown bias and they've changed the way people look at GPU benchmarks with their inside the second benchmarks.
Would be nice if reviewers put the clocks next to the names just for quick reference.

Some do it, or at least put the stock version under 980 as baseline and then add the 980 xxx retail OC if they use one, One other thing will be to put the resume of max or average variation on turbo clock speed. if you have a sample who boost at 1200mhz+ and one at 1280mhz thats quite a difference.

This said, i dont know if this is the only reason of TR dont have it, . There's surely many things behind the scene that we dont know. For be honest, lately the reviews of TR are a bit strange,

specially their frametimes report one who are completely opposite at the Fcat mesure on other sites, reviews.
 
I can't believe that people online are actually implying that the TR review was biased.

1) Almost entirely 4K testing (and their Fury X review was actually 4K only) showing the AMD cards in the best light

2) More Gaming Evolved titles than TWIMTBP titles (4 vs. 3)

3) It makes no sense to test reference 980s and 970s against non reference Fury, 390X and 390. There are actually more reference Nvidia cards in the review than there are reference AMD cards. They had non ref Fury, non ref 390X, non ref 390, non ref 980, non ref 970. Sounds perfectly reasonable.
 
The children who come out of the woodwork for commenting on CPU and GPU reviews is terrifying. I also suspect paid shills appear to fuel the fire.

Still, the factory overclocked cards issue has been happening forever. Just be specific about what is being tested. They will still be called biased though. Pitchforks. It is the general public we're talking about here.
 
Last edited:
Not surprising given hardocp ripped the Fury X a new asshole in their review, but TR? I generally avoid hardocp because of Kyle and their silly playable settings methodology, I prefer the in-depth reviews of TR and others who show frametimes and I think know their stuff better than hardocp.
 
Hocp, at least they did mention in their Fury review,

Performance jumped up on the ASUS STRIX R9 Fury. The Fury is now 31% faster than the GeForce GTX 980. The setting holding back performance seems to be the NVIDIA Depth of Field in this game. The GTX 980 can render it much better, the Fury not so much.

Though their method of testing means that they'd test for the best playable settings where 980 is 15% faster than Fury, and even the apples to apples comparison uses it, and hence pushes down the Fury performance to 1/10 of its advantage without it,

In this test we have lowered the settings to a non-bottlenecked "Best Quality" setting. ASUS STRIX R9 Fury is 3% faster than GeForce GTX 980.

http://www.hardocp.com/article/2015/07/10/asus_strix_r9_fury_dc3_video_card_review/5#.VeiT4C6deZ8

And if we had factory clocked 980 masquerading as reference 980 in their charts like techreport did, it would have been faster than the 3% faster Fury and have a clean sweep of the land.
 
Nvidia's had factory overclocked cards for years and factory overclocked cards are likely what the consumer will buy, not TR's or other sites fault for testing those cards, AMD just hasn't had nearly as much success with AIB offering cards clocked much higher than the reference design.
Also Fury and Fury X doesn't look as good as the 980 and 980 ti cards in terms of frame time consistency in a large number of games, AMD has yet to match nVidia in that area for single GPU cards let alone dual gpu cards.
All in all, to be perfectly honest the Fury range of products are underwhelming compared to the competition and might be considered a beta or test version for new technology, HBM of course. I wouldn't call them a huge failure like the Radeon 2900 or FX series, but they're simply not up to par.
 
Nvidia's had factory overclocked cards for years and factory overclocked cards are likely what the consumer will buy, not TR's or other sites fault for testing those cards, AMD just hasn't had nearly as much success with AIB offering cards clocked much higher than the reference design.
Also Fury and Fury X doesn't look as good as the 980 and 980 ti cards in terms of frame time consistency in a large number of games, AMD has yet to match nVidia in that area for single GPU cards let alone dual gpu cards.
All in all, to be perfectly honest the Fury range of products are underwhelming compared to the competition and might be considered a beta or test version for new technology, HBM of course. I wouldn't call them a huge failure like the Radeon 2900 or FX series, but they're simply not up to par.
All would be well if AMD priced the Fury line right, but for $650 you could buy a mother fucking GTX980Ti which is frankly better in every way. Withholding review samples will only make things worse. The reviews will still come out and if AMD thought they were biased before..

On the plus side, AMD saves a few hundred dollars for every card they don't send, so they lose a little bit less money this quarter :)
 
Back
Top