AMD Radeon R9 Nano Review by TPU...Not

[btarunr]

Just checked the preliminary TPU youtube video. As an AMD product user (not that it affects my opinion). i can sense you guys seem to be a bit upset about not getting r9 nano review sample and i can understand that. You guys should have gotten one to review. However, with all my honest feelings that video seems to be an excuse to target AMD for not getting one to review and not in subtle way but in a noticable way.

Video was made last week, we learned we're not getting a sample yesterday.

 

[Tsukiyomi91]

and that poses a problem for newcomers who wanted something fast, efficient, cheap & doesn't kill their monthly/yearly utility bills while ensuring their bedroom won't ended up getting too warm coz of building a "portable heater". AMD has already lost the efficiency war against both Intel & Nvidia since the global release of 3rd Gen Core Series & Kepler chips. With Haswell, Skylake & Maxwell winning every benchmark tests done by honest, unbiased, no BS reviewers of notable reputation, AMD has cringe over & over again due to the lackluster of not releasing a completely brand new chips to the masses. Instead, they rely on an old, dying chip with lowly tactics, relying on paper sheets & sheer price slash to keep them afloat. Sorry to say but that's how AMD is now....

 

[cuneytcam]

It has been said earlier that the Video test was run before the news about Nano/review.

I did not want to mean that video is 100% kneejerk reaction to review sample things. You could have posted that video years back i would still say the same things except first paragraph.

 

[dorsetknob]

PCWorld got one http://www.pcworld.com/article/2979...-to-the-fury-a-tiny-gtx-970-and-a-pencil.html

Spoiler: Epic Review :)

I think not

 

[Tsukiyomi91]

IMO, I would spend that $650 for either a 2nd GTX970 & a SLI capable board OR a heavily OCed GTX980Ti where it literally killed it's own big brother, Titan X & the whole suite of AMD's finest high end cards.

 

[newtekie1]

Jayz is a system builder with great experience with case building and customization.
He might have cooked up a special rig for the Nano to showcase it. If he haven`t, he is still very good at explaining how it all fits together
Size is mostly the only selling point AMD have of this card, so they try to market it like that.
Too bad GTX 970 ITX is there to ruin the show

That is a very good argument. I also think it is because he is very biased towards AMD, even though he claims not to be because he runs nVidia cards in his main computer(he also openly admits that he only runs those cards because they were given to him). But if you look at some of his video card videos, he is very biased towards AMD. The R9 285 ITX video is a great example of this. He goes on and on praising the card for being so small, even going as far to say it is the only card on the market that small with that level of performance and that price point. But at the time he released the video, the GTX960 had an ITX version out for months, it was/is cheaper than the R9 285 ITX, and performs the same. Also, there has been GTX970 ITX versions from eVGA, ASUS, and Gigabyte for months as well, and it totally kills the R9 285 in performance.

 

[Luore]

I really only signed up just to post this despite I would at some point.
This 'news' post does come out very ranting. Half of it could be removed and the message would be the same.
TPU is biased against AMD most likely unknowingly.
On the normal benchmarking. It is very strange to have something like Project Cars be placed into the mix knowing full well that it won't perform well against AMD. It is very well known that Project Cars developers do not care about any optimization for AMD hardware.
WoW for example is also very heavily biased towards Nvidia hardware. When you had to separate the Fury X's review with "w/o Project Cars and WoW", you guys should realize something is wrong with the test suite.
https://www.techpowerup.com/reviews/AMD/R9_Fury_X/31.html
Despite on Tomb Raider and Witcher 3 disabling the TressFX/Hairworks, the most obvious should've been those two games.

Second part however is power consumption the more recent (relatively) controversial thing.
While it is great for us to know the power consumption on the GPU side it is only half of the equation. It's been known and shown, especially recently, that Nvidia uses a lot of software interventions, especially in DX11 where it recompiles shaders to something more suited. Recently in Kepler and Maxwell they've axed a lot of hardware based functions for software based solutions. This in turn shows less power consumption on the GPU side however due to all the software intervention, it taxes the CPU more and in turn increasing power consumption on the CPU side.

AMD's hardware have essentially well, more hardware. This causes more GPU power consumption while being able to execute more on the hardware side. This is one reason why DX12 more favorable towards AMD due to not needing driver intervention, because there are next to nil while having more on the hardware to execute the needed function. This frees up developers to abuse the CPU more along with the GPU.

If it was me I'd remove the two games and include a full system power consumption along with GPU consumption. It'd really clear things up and add more needed information.

Unrelated but concerning CrossFire and SLI. Put it in spoiler to not be a gigantic link blob.

Spoiler

To clear up CF has worse frame time, Nvidia is now on that side of worse frame time as has been for the past year. Is AMD's perfect? No, but it is better and improved. Of course the damage is done except for the more informed.
http://www.sweclockers.com/test/20216-nvidia-geforce-gtx-titan-x-i-sli/
http://cdn.sweclockers.com/artikel/...4ZDFkZGVkOTkyNTZjNDMwMmRkZWM4ZjU1YWEzYmIwOSJ9
vs
http://cdn.sweclockers.com/artikel/...zZTkzY2NjM2ZjMWU2Y2I4NDk4MmE4ODcxNTVlNzJjMyJ9
Just as an example.

 

[dorsetknob]

On the normal benchmarking. It is very strange to have something like Project Cars be placed into the mix knowing full well that it won't perform well against AMD. It is very well known that Project Cars developers do not care about any optimization for AMD hardware.
WoW for example is also very heavily biased towards Nvidia hardware. When you had to separate the Fury X's review with "w/o Project Cars and WoW", you guys should realize something is wrong with the test suite.

The range of games is probably what the average Gamer plays So its a FAIR REPRESENTATION of games to bench mark with

 

[Tsukiyomi91]

If u think that TPU & other reputable reviewer's testing methods prove to be "biased" as what you claim, you are either high or just simply in a cringe that AMD isn't getting what it deserves... Then please do a more proper testing methods that would deemed as no BS in the eyes of the blinded fools.

 

[Luore]

The range of games is probably what the average Gamer plays So its a FAIR REPRESENTATION of games to bench mark with

Doesn't really work that well if you want to be unbias. You don't incorporate biased material while trying to be unbiased. This is also why I said unknowingly biased. If they had to show a different set of information without Project Cars and WoW it should've been suspect to begin with. Why bother having it if you have to remove it.

 

[cadaveca]

I have been a long stalker of TPU, but if you are going to use your media power for some weird agenda i highly doubt it will help you in the long run.

We have no agenda other than to give our readers the info they want, the way they want, while remaining 100% transparent and honest. Any "bias" from news postings is not our own... most often those are written by other sites, and we merely link to them. If our readers want to know, we try to inform them. It could be good, bad, or in-between, it's simply aggregation, not editorials. We very rarely post editorials, and every thing mentioned so far as being problematic in postings isn't our writing.

For me, this isn't a job. It's a hobby, 100%. It doesn't pay bills, it merely takes up some of my time, and lets me play with and write about hardware based on my experiences, which I love doing. I also get the chance to help others with problems, and in return, I get some support, and when I have problems, there's always someone to help me out in return.

If that's a weird agenda, well, I have to tell you, we're pretty successful at it already.

Doesn't really work that well if you want to be unbias. You don't incorporate biased material while trying to be unbiased. This is also why I said unknowingly biased. If they had to show a different set of information without Project Cars and WoW it should've been suspect to begin with. Why bother having it if you have to remove it.

That's not true, unfortunately. Those are titles with big followings, so using the benchmarks we do allows those that play those games to have an accurate representation of what they would get from a product, in the way they already use their current products. It's about providing relevant data, not about bias. We cannot help that certain software runs better on certain hardware. That's up to the devs, and suggestions that we might change benchmarks to remove bias and thereby not provide relevant info, kind of cancels each other out. Maybe you should ask the devs to write software without bias.

It also provides the opportunity for hardware makers to "show-up" the competition, if they were performing poorly before, and now they do well... it's a bar they should strive to achieve if they wish to increase market share.

(this is my personal opinion, not the opinion of TechPowerUp as a whole. I can never speak for TechPowerUp as a whole, only W1zzard can.)

 

[tabascosauz]

Doesn't really work that well if you want to be unbias. You don't incorporate biased material while trying to be unbiased. This is also why I said unknowingly biased. If they had to show a different set of information without Project Cars and WoW it should've been suspect to begin with. Why bother having it if you have to remove it.

It's not like it affected the overall outcome for the Fury X...there was a figure with Pcars included but it obviously was dismissed in favor of the number excluding Pcars for the final score. Just because it's Gameworks doesn't mean that people don't play it. And it doesn't matter if AMD cards suffer as a result. Reviewer simply puts less emphasis on that benchmark for the AMD card.

If that's what you think constitutes a clear bias against AMD, it's apparent that you haven't been around long enough to read AMD reviews in the past. The 270X cards were given rather high scores, for one, because they were well designed.

 

[fec32a4de]

...WoW
[/spoiler]

great, now you got me posting on this thread/discussion.

Im only gonna focus on one thing: AMD and WoW.
They're not friends. AMD cards suck on WoW. Always did, always will [Maybe DX12 will save us? Gamon didn't]. So do their CPU/APU. (Tomshardware mentions the CPU part).
(i.e most basic intel parts "Pentium" and i3 perform well beyond AMD's comparable CPU/APU in regards to WoW.

Fact is, either blizz focuses on Nvidia (3d vision, Nvidia HBAO+ [actually says Nvidia in Settings] etc support - oh shit they do!) or AMD has some terrible implementation of DX11 drivers in regards to WoW.

When WoW with 390 or 290 gets less FPS in real life tests than 960. You know something up. But this is all besides the point.

Me? I'm an AMD fan, a big one. I will always be. I will always root for the "underdog". I want AMD to succeed and I need them to, we all do. Otherwise monopolies and shit like that.
The fact that they keep refreshing the same GPUs over and over and over again isn't doing them any favors. And overclocking their APUs and CPU offerings isnt doing anything at all.

I am looking forward to Zen, to new CPU/APU offerings on 16/14nm node and new videocards. AMD can make a comeback, but they are running out of time and money.
I may be running on a Core i7 at the moment, but I am always looking at AMD, I want them back.

TL;DR version
TPU isnt overly critical of AMD on games like WoW. Its just that AMD sucks in this department. And Nvidia and Blizz work together.
And its nice to see TPU use WoW in their reviews, no one besides them and TomsHardware do. Thank you TPU!

This whole damn thread has been derailed anyway.

 

[Tsukiyomi91]

Gonna rest this case once & for all. Will give these AMD fanboys like Sony time to get their shit together & think in a rational manner. Maybe one day they will slowly shift into the Green camp once all the lies AMD has been brainwashing them with is gone. Oh, the Blue camp would accept them with open arms & show them the beauty of power, efficiency & ingenuity in which AMD has failed to accomplish that.

 

[Luore]

That's not true, unfortunately. Those are titles with big followings, so using the benchmarks we do allows those that play those games to have an accurate representation of what they would get from a product, in the way they already use their current products. It's about providing relevant data, not about bias. We cannot help that certain software runs better on certain hardware. That's up to the devs, and suggestions that we might change benchmarks to remove bias and thereby not provide relevant info, kind of cancels each other out. Maybe you should ask the devs to write software without bias.

I understand that, but it shouldn't be showing up as part of the performance summary, something that a lot of people reference to.
If it doesn't then by all means have it in the review but not in the relative performance.

 

[rtwjunkie]

Gonna rest this case once & for all. Will give these AMD fags like Sony time to get their shit together & think in a rational manner. Maybe one day they will slowly shift into the Green camp once all the lies AMD has been brainwashing them with is gone. Oh, the Blue camp would accept them with open arms & show them the beauty of power, efficiency & ingenuity in which AMD has failed to accomplish that.

You also show an extreme that is no more helpful and just as disruptive as Sony's extreme AMD zealotry.

 

[cadaveca]

I understand that, but it shouldn't be showing up as part of the performance summary, something that a lot of people reference to.
If it doesn't then by all means have it in the review but not in the relative performance.

Do you work for AMD?

To be fair to both parties, it DOES need to be included. Removing software that many people use, but favors a different brand, would be biased, the exact thing you are saying we shouldn't be. Just because one brand does poorly in a test, doesn't mean we exclude it, or 100% anyone could claim bias.

Since each tested app is on it's own page, we know which tests are popular, and what people read out of the reviews, and what they want to see, and what they ignore.

 

[Luore]

Gonna rest this case once & for all. Will give these AMD fags like Sony time to get their shit together & think in a rational manner. Maybe one day they will slowly shift into the Green camp once all the lies AMD has been brainwashing them with is gone. Oh, the Blue camp would accept them with open arms & show them the beauty of power, efficiency & ingenuity in which AMD has failed to accomplish that.

That's not really helping your case by insulting people and hostility.
Personally I don't care for being aligned to any company nor do I find it helpful to ones' self and others. It's not being a fan of any sort. It's being reasonable and if something seems off to you it should be criticized. Doesn't matter what company or media it is. Whether you're on camp A, B, or C, or some guy down the street in the corner.

Do you work for AMD?

I do? Cool, when did that happen.
No, I'm just someone that's interested in hardware and software of computers.

to be fair to both parties, it DOES need to be included. Removing software that many people use, but favors a different brand, would be biased, the exact thing you are saying we shouldn't be.

Removing bias doesn't make bias, that's just fallacious at best. It fits into this
https://en.wikipedia.org/wiki/Paradox_of_tolerance

 

[cadaveca]

I do? Cool, when did that happen.
No, I'm just someone that's interested in hardware and software of computers.

Removing bias doesn't make bias, that's just fallacious at best. It fits into this
https://en.wikipedia.org/wiki/Paradox_of_tolerance

My job as a reviewer is to meet the needs of our readers, and provide relevant info to their uses, not cater to the hardware makers. We set the standard, and it is up to the hardware maker to meet or exceed our expectations, or to fail, and the reader to judge a product's merit based on that. Otherwise we aren't reviewing anything.. we would then be marketing shills. It is also my job to represent our readers to the companies, and ensure that what the end user wants is conveyed and understood, not to educate on reader on what they should want.

I did have to ask, since your statements seem more friendly to AMD than anyone else, and are contrary to what being a reviewer is about, in my books. If hardware doesn't show up on my door step, I simply tell me readers DO NOT BUY. I can only recommend what I have tried myself. I could never get another piece of hardware, and it would not matter to me one bit. W1zzard could "fire" me, ban me, whatever... I do my job without fear of any repercussions.

That's not having bias, in my books.

 

[dorsetknob]

I understand that, but it shouldn't be showing up as part of the performance summary, something that a lot of people reference to.
If it doesn't then by all means have it in the review but not in the relative performance.

How oh How do you judge whether AMD cards have improved if those games you say are unfaverable to AMD Cards are not used for comparasion bench marks

Use a Crystal Ball and Tarot cards
sorry but in order to pass judgment in review you have to use a wide range of games some favorable for AMD and Some for Not
Thats life and untill games are made Hardware neutral thats the way it will be

 

[tabascosauz]

Removing bias doesn't make bias, that's just fallacious at best. It fits into this
https://en.wikipedia.org/wiki/Paradox_of_tolerance

A slightly more logical Sony here, boys.

I don't understand why you cannot comprehend the fact that Pcars' presence in the benchmarks doesn't instantly assume Pcars' influence in the final score. I don't know why you can't figure this out. If you are oblivious to the fact that it's a Gameworks game and has a huge AMD disadvantage, and are unable to let it go as a result, then I'll leave it up to you to figure it out. The reviewers and the readers are all aware of this fact.

Again, just because Pcars fared horribly on the FuryX, you shouldn't be coming to the conclusion that the low overall score for the FuryX was due in part to Pcars. It was priced to beat the GTX 980 Ti, and it failed in that regard, not only in Pcars.

 

[Osjur]

Here's the most likely reason why TPU didn't get a sample. They operate in Germany as a worldwide english site. There just isn't that many samples and they most likely went to the german sites.

Also, AMD is not directly sending these samples, they come from retailers and its not just a card they are sending. It's a whole ITX form factor computer.

 

[cadaveca]

Um, I'm not in Germany? Only W1zzard and DarkSaber are in Germany AFAIK. We are a global website, with hosting and staff in many countries, with our primary readership being in the US.

Also, AMD is not directly sending these samples, they come from retailers and its not just a card they are sending. It's a whole ITX form factor computer.

Now that... THAT is relevant. So AMD wanted testing done their way only.

 

[Kanan]

Wow, worst news article I read here ever and by far.

1st: please stop the crying, its a new innovative tech thats in the childhoods, you sound like a NV-fanboy anyway, thats not an 7870, or GCN 1.0 product, you obviously think its easily produced, but its the exact opposite. Please go and search for some descriptions how Fiji is designed, how many steps and companys are involved to the finished product.

2nd: 350$ for Fiji? Are you serious? Again you sound like an AMD hater, 0 respect für their accomplishments and 350$ is just crazy, sorry.

Please shut off this NV biased whining. You should delete the whole Newspost and accept that its not justified to cry that loud about things that are pretty normal considering a brand new tech, HBM + Interposer + connections etc. that is.

 

[the54thvoid]

That's not really helping your case by insulting people and hostility.
Personally I don't care for being aligned to any company nor do I find it helpful to ones' self and others. It's not being a fan of any sort. It's being reasonable and if something seems off to you it should be criticized. Doesn't matter what company or media it is. Whether you're on camp A, B, or C, or some guy down the street in the corner.

I do? Cool, when did that happen.
No, I'm just someone that's interested in hardware and software of computers.

Removing bias doesn't make bias, that's just fallacious at best. It fits into this
https://en.wikipedia.org/wiki/Paradox_of_tolerance

You don't seem to understand the point of a review and the gaming market whatsoever.
The test suite at TPU includes a host of games that are very popular. To dismiss a game because of vendor intervention would be an unfair reflection of the state of affairs. If game "x" is popular with a vast proportion of gamers, we like to know how it plays. To remove that game is to deny the reviewers audience of information.
What you purport to do is censorship of intelligence by proxy of 'biased' coding. It's not TPU's fault if a popular game has been coded favourably to a certain vendor. Dice worked closely with AMD on BF4, perhaps that should be excluded just in case? BioShock series as well? Thief, Tomb Raider. All AMD titles, surely must be filtered out. It's not TPU's problem that in DX9 or DX11, Nvidia managed to software optimise better.
No matter how cerebral your argument is put across and how civil you are (and thanks for being polite) your point is necessarily misguided. The hypocrisy of asking for a title to be removed because it favours a vendor is denying the reality of gaming hardware outcomes. DX12 will remove Nvidia's coding advantages so perhaps moving on we shouldn't need to care but frankly what you are suggesting is just as bad as Nvidia asking Star Dock to remove Async from the AoS bench...