The Fury Nano Thread

[Gikaseixas]

Question is what is worse from AMDs point of view. The bad PR from the review or the bad PR from not sending them a card.

This

 

[Silverforce11]

Temps data in:

https://www.youtube.com/watch?v=ofJidWb1CDM

~65C under load.

 

[guskline]

Silverforce, what is load? Last night on the Pc
Perspective live podcast there was some mention that Nano actually runs UP to 1000 Mhz But goes as low as 800 Mhz to stay within the 175 W threshold. I'm anxious to see what performance truly is.

 

[MeldarthX]

Silverforce, what is load? Last night on the Pc
Perspective live podcast there was some mention that Nano actually runs UP to 1000 Mhz But goes as low as 800 Mhz to stay within the 175 W threshold. I'm anxious to see what performance truly is.

this is what was said on youtube.

+Brian Fisher we have 3x3 measuring test. (video has been edited to 6x speed) 1-step. 3 minutes stay idle at windows desktop. 2-step. run Metro : Last Light - D6 Test. it takes 3 minutes around. (DX11, FHD, Very High Quality, 16x AF, Normal Blur, Very High Tessel) 3-step. return to windows desktop cooldown the temperature. it also takes 3m.

 

[Silverforce11]

Silverforce, what is load? Last night on the Pc
Perspective live podcast there was some mention that Nano actually runs UP to 1000 Mhz But goes as low as 800 Mhz to stay within the 175 W threshold. I'm anxious to see what performance truly is.

Temp target of 75C. Up to 1ghz.

Metro LL load, 68C highest, ~65C average.

Zlatan has exp with a Nano first hand, he said it maintains 4 digit clock speeds (before we knew the 1ghz specs, we all thought it was 850mhz) all the time in his tests. Maybe it's very cold in Eastern Europe.

 

[Elixer]

Stream is going on now... They are talking about HBM, comparing 290 to Nano, Sept 10 is on shelf date.

75C max, they will show thermal solution...
NO THROTTLING at 75C, only at 85C. (Which means, it should never throttle, unless you have a really, really hot case).

They will release 3d CAD file for the heating solution, in case you want to mod it...

Power draw is 150W (from 1 8 pin connector) so, total is 175W max. (8 pin + PCIe power) which means, it is under the 225W max total possible.
1000 times/sec recalibration for the power system.
Dual BIOS switch.
UMC is for the HBM stuff.
"Ass kicking in DX12"
They talk about Async compute... AMD didn't cut corners on the design, Maxwell did, and Async compute performance shows.

 

[positivedoppler]

GOLD!!

http://techreport.com/news/28971/wanted-for-review-amd-radeon-r9-nano

These same guys who repeatedly find results that are so much worse for AMD than most other sites...

These same bunch who threw out Dirt Showdown for "bias" due to poor NV performance... but lavish praise on Project Cars..

The same site that finds XDMA CF stutters worse than SLI when every other site find the opposite...

About time AMD. Well played you guys, now you're catching on!

This is laughable and embarrassing. If you can't afford $650(probably a tax deductible expense too) then you have no business running a hardware review site. Begging the public for help makes them look like a little Mickey Mouse operation.

 

[LTC8K6]

What happened to 175W?

Now we are going to have the same performance with only 150W?

I guess that's great, but why was it 175W in all the release articles?

 

[Arachnotronic]

This is laughable and embarrassing. If you can't afford $650(probably a tax deductible expense too) then you have no business running a hardware review site. Begging the public for help makes them look like a little Mickey Mouse operation.

They're trying to get one ahead of availability; I don't think this is a matter of the TR not being able to afford one.

 

[tential]

Question is what is worse from AMDs point of view. The bad PR from the review or the bad PR from not sending them a card.

As a consumer I'm even more annoyed I have less reference points. I want to say I'm 100% used market at this point but I still want a nano/furyx lol....

 

[3DVagabond]

You realize that this is not good PR, right?

Do you actually care whether or not Tech Report gets a review sample? I remember the hack job they did with the 7950. I don't know if that had anything to do with AMD's decision, but I'm not concerned that they don't seem to be on the top of AMD's review partner list.

They've obviously worked in consort with nVidia in the past. Maybe that's where they got their info on frame times and ways to make Nano look bad. Because that's exactly what happened with the 7950, IMO. Considering they changed their entire review suite for that review.

 

[Silverforce11]

What happened to 175W?

Now we are going to have the same performance with only 150W?

I guess that's great, but why was it 175W in all the release articles?

AMD's TDP is always above their gaming power usage. R290X TDP was 300W. Avg gaming load: 230-250W.

Fury X is 275W TDP, avg gaming load is less than R290X.

175W Nano TDP => ~150W gaming load sounds spot on.

 

[tential]

Do you actually care whether or not Tech Report gets a review sample? I remember the hack job they did with the 7950. I don't know if that had anything to do with AMD's decision, but I'm not concerned that they don't seem to be on the top of AMD's review partner list.

They've obviously worked in consort with nVidia in the past. Maybe that's where they got their info on frame times and ways to make Nano look bad. Because that's exactly what happened with the 7950, IMO. Considering they changed their entire review suite for that review.

My HD7950 feels stupidly fast still. With how high that thing OCs too?

I definitely got my money's worth out of this card. If it could do any resolution VSR, I'd probably not upgrade.

 

[Silverforce11]

Oh, check this out: This is gold too.

http://hardforum.com/showpost.php?p=1041819470&postcount=6

http://hardforum.com/showpost.php?p=1041819539&postcount=15

Who says [H] isn't biased? A few years ago they were quite neutral. Now it seems like NV $ has reached them.

AMD better not give these fools a Nano sample.

http://hardocp.com/article/2015/08/27/amd_radeon_r9_nano_video_card_paper_launch#.VejoGxGqpBc

"Today is just a paper launch as is AMD's new habit."

 

[3DVagabond]

Oh, check this out: This is gold too.

http://hardforum.com/showpost.php?p=1041819470&postcount=6

http://hardforum.com/showpost.php?p=1041819539&postcount=15

Who says [H] isn't biased? A few years ago they were quite neutral. Now it seems like NV $ has reached them.

AMD better not give these fools a Nano sample.

http://hardocp.com/article/2015/08/27/amd_radeon_r9_nano_video_card_paper_launch#.VejoGxGqpBc

"Today is just a paper launch as is AMD's new habit."

If you read through the thread at (H) you'll see that except for the performance graphs the review has been written.

 

[Silverforce11]

If you read through the thread at (H) you'll see that except for the performance graphs the review has been written.

I've basically said that since last year. They write their article with conclusion already made, then fill in blanks.

Their Asus Fury Strixx for example, they claim it wasn't power efficient compared to their 980. But their own results show the Fury Strixx uses ~11% more power, it delivers ~15% more performance (with GameWorks off in Dying Light & Witcher 3). It is therefore MORE power efficient.

And they are trolling AMD on their forums, editors, trolling.. and they want AMD to give them a sample?

 

[LTC8K6]

AMD's TDP is always above their gaming power usage. R290X TDP was 300W. Avg gaming load: 230-250W.

Fury X is 275W TDP, avg gaming load is less than R290X.

175W Nano TDP => ~150W gaming load sounds spot on.

The post was fixed. Nano is back to 175W now. :biggrin:

 

[AnandThenMan]

nm misunderstood the post.

 

[Gikaseixas]

I've basically said that since last year. They write their article with conclusion already made, then fill in blanks.

Their Asus Fury Strixx for example, they claim it wasn't power efficient compared to their 980. But their own results show the Fury Strixx uses ~11% more power, it delivers ~15% more performance (with GameWorks off in Dying Light & Witcher 3). It is therefore MORE power efficient.

And they are trolling AMD on their forums, editors, trolling.. and they want AMD to give them a sample?

They clearly are butthurt it shows by reading that thread. So juvenile LOL
AMD did well to not send them a sample when everybody already know the outcome.

 

[Eymar]

They clearly are butthurt it shows by reading that thread. So juvenile LOL
AMD did well to not send them a sample when everybody already know the outcome.

Yeah, it's really surprising how immature the posts are. I remember when their reviews were positive for AMD (7970 period), now they treat AMD with such vitrol. Pretty backwards thinking too, why dump on a card that showcases an interesting and possibly emerging market (sff power pcs) that would bring more visitors to your site.

 

[raghu78]

I've basically said that since last year. They write their article with conclusion already made, then fill in blanks.

Their Asus Fury Strixx for example, they claim it wasn't power efficient compared to their 980. But their own results show the Fury Strixx uses ~11% more power, it delivers ~15% more performance (with GameWorks off in Dying Light & Witcher 3). It is therefore MORE power efficient.

And they are trolling AMD on their forums, editors, trolling.. and they want AMD to give them a sample?

exactly. I have said the same too. Recent hardocp reviews on Fury X and Fury look like they have written the conclusion first and then try to give justification for that. Moreover their current game suite is pathetic and heavily Nvidia biased given its small size. Moreover they make it a point to use Gameworks features which are known to run poorly on AMD cards.

Here is what brent had to say about how he would redesign AMD's product lineup. It was completely devoid of logic. He wanted a highly binned ultra low volume Nano SKU to be priced at USD 329. Does he even realize that the Fiji die size is right on par with GTX 980 Ti. This is why people like him should not give advice as they have no proper logical thinking.

http://hardforum.com/showpost.php?p=1041820972&postcount=201

This was my reply

http://hardforum.com/showpost.php?p=1041835077&postcount=601

That guy Brent has no idea that pricing is not determined arbitrarily. It is determined by the cost structure of your product and that of the competition. AMD's pricing on Fiji is determined by both the manufacturing cost and the limited supply situation. Right now I am very disappointed with AMD's products this year but the last thing AMD want to do is get into a price war. Nvidia has the better cost structure (better perf/sq mm and perf/transistor) and the more efficient architecture (perf/watt). AMD cannot afford to get into a price war this gen. That was possible in the past like in 2008 when AMD had a massive perf/sq mm advantage and the HD 4870 at 260 sq mm beat the 576 sq mm based GTX 260 and the 282 sq mm HD 4890 competed well with the 470 sq mm GTX 275.

https://en.wikipedia.org/wiki/GeForce_200_series

http://www.anandtech.com/show/2556/2

But right now all AMD can do is try and keep the market share loss as low as possible since I don't think AMD can gain back market share this gen.

 

[tential]

exactly. I have said the same too. Recent hardocp reviews on Fury X and Fury look like they have written the conclusion first and then try to give justification for that. Moreover their current game suite is pathetic and heavily Nvidia biased given its small size. Moreover they make it a point to use Gameworks features which are known to run poorly on AMD cards.

Here is what brent had to say about how he would redesign AMD's product lineup. It was completely devoid of logic. He wanted a highly binned ultra low volume Nano SKU to be priced at USD 329. Does he even realize that the Fiji die size is right on par with GTX 980 Ti. This is why people like him should not give advice as they have no proper logical thinking.

http://hardforum.com/showpost.php?p=1041820972&postcount=201

This was my reply

http://hardforum.com/showpost.php?p=1041835077&postcount=601

That guy Brent has no idea that pricing is not determined arbitrarily. It is determined by the cost structure of your product and that of the competition. AMD's pricing on Fiji is determined by both the manufacturing cost and the limited supply situation. Right now I am very disappointed with AMD's products this year but the last thing AMD want to do is get into a price war. Nvidia has the better cost structure (better perf/sq mm and perf/transistor) and the more efficient architecture (perf/watt). AMD cannot afford to get into a price war this gen. That was possible in the past like in 2008 when AMD had a massive perf/sq mm advantage and the HD 4870 at 260 sq mm beat the 576 sq mm based GTX 260 and the 282 sq mm HD 4890 competed well with the 470 sq mm GTX 275.

https://en.wikipedia.org/wiki/GeForce_200_series

http://www.anandtech.com/show/2556/2

But right now all AMD can do is try and keep the market share loss as low as possible since I don't think AMD can gain back market share this gen.

I linked to read. Exited when he said at at 450 th fury would give the 980 a run for its money.
At 450 it's a stupid amazingly good deal it's not even funny. That was a post written as if he had never seen a gpu before.....

 

[Silverforce11]

I linked to read. Exited when he said at at 450 th fury would give the 980 a run for its money.
At 450 it's a stupid amazingly good deal it's not even funny. That was a post written as if he had never seen a gpu before.....

He wants Nano for <$350? PFFT...

Fury for $450?

Well sure, if 980 drops to $400, Fury should be $450. Faster in DX11 already, more "future-proof" (can I use that term, now?)...

And while Brent is at it, he can fix his Asus Fury review, the conclusion was all wrong. Fury Strixx is MORE power efficient, faster, similarly priced to the 980 in that comparison based on his own data.

 

[Hitman928]

raghu, it looks like your post was deleted. Get used to it if you want to post over there and have a differing opinion.

I used to like Hardocp because of their different testing methodology as another data point that looked at things from a different angle, then following the site and the reviewers closer, I stopped going there entirely. It wasn't even so much the perceived bias and such as it was the complete dismissal of any community feedback and the we're always right no matter what type of attitude. . . no thanks.

 

[positivedoppler]

They just did a 980 Ti review with almost nothing but Gameworks titles except for BF4. That plus the rants I wonder is the reason amd didn't provide them a nano.