AMD Polaris architecture – GCN 4.0

[Poneros]

Intel has already proven they will play dirty in order to stay top dog. We're talking about billion dollar businesses here, you think they won't try anything and everything under the sun to stay in the spotlight? Oh, the naivete..

The only question is if the risk is worth it, but we don't have the available metrics to comment. The companies certainly do, on the other hand.

 

[DarrenM343]

most review sites have paid for advertising banners on their website, if Asus take out a banner ad and then just happen to get a good review... ?

if Asus never suggested they wanted a good review it's different. The reviewer however does not deserve to be a known as a good reviewer if they review a product favourably because a company used their paid for advertising . I'm sure most respectabe reviewers wouldnt do it because they are offering a service which requires a certain degree of honesty and integrity.

Both sides have to take responsibility. So the reviewer should not allow the company to buy the advertising space(which would be daft) or should make it clear that doing so will not affect the outcome of any reviews. If Intel said "said fine, we're not advetising then" then that is fine too (I believe) as no offence has been made although of course a little cringe worthy, but Intel never asked for it anyway but then backed away when the reviewer stated it would not affect their relationship.

BTW, Intel lowering the cost of their chips to be less than AMD to win more customers is not illegal, even doing so at a loss for a while. If Intel products are so good that the stockists have to accept lower margin in order to stock them vs AMD chips that wouldn't be illegal either.

And I'd be surprised if any reviewer confessed to taking a payment if they had done.If they have, are they still reviewers now?Laws aside I'd never trust them again. As I said, both parties have to take responsibility. Taking a bribe is as disgusting as offering it in this modern world so if any foru members are reading and trusting the reviews from people who knowingly took a bribe then more fool you

 

[Mei]

Sorry guys, you are stating without any proof that both of these companies offer payment for something, such as preferential treatment. Well,I don't know how up to date you are with laws affecting business, especially in the UK, but you are acusing the two companies of BRIBERY & CORRUPTION! That is against the law.

Something as simple as taking a client out to an expensive lunch these days can be seen as bribery which is against the law. Offering money to review sites would definitely be against the law and very controversial. I believe the US law is not as strict as the UK,which is a shame for a supposedly "world leader" but I'm sure both nVidia and AMD are careful as they sell their products internationally where such laws are more strict. If for example a UK based person or company accepted money from the AMD or Nvidia then the UK party would be breaching the law for accepting the bribe so the law applies to both the giver and the receiver.

I'm surprised someone hasn't picked up on this as many companies in the UK HAVE to provide anti-bribery and corruption training and covers gifts and hospitality, both the giving of and the receving of. Blatantly paying someone for something to "help get what they want" would be the worst thing they could do. "High profile" companies have to be especially careful.

well all reviewers that went to the nvidia event got a little holiday with horseriding and ofcourse all expenses paid, the cards free - does that make their opinions of the product and the company tainted? i think you would have to say yes

BUT and i think its a big but, not to be confused with my butt, unlike video games and some other things opinions i dont think matter so much when you can look at the facts of benchmarks to make up your own mind, being sceptical of all reviewers opinions is pretty healthy, same as fans of a company on a forum i guess

sadly yes benchmarks can be manipulated in so many different ways so it does mean you have to be careful, clockspeeds temperatures, drivers, all kinds of ways but you can get to the truth eventually

 

[DarrenM343]

Intel has already proven they will play dirty in order to stay top dog. We're talking about billion dollar businesses here, you think they won't try anything and everything under the sun to stay in the spotlight? Oh, the naivete..

The only question is if the risk is worth it, but we don't have the available metrics to comment. The companies certainly do, on the other hand.

What are the penalties?
For people, up to 10 years in prison and an unlimited fine. Companies can also face unlimited fines.

Mistakes of the past may not be indicative of future behaviour.

Not defending anyone here but given the risks to businesses and employees I'm not so paranoid about the companies.

If after the training I have received I am caught bribing an employee of another company, I could fact 10 year in prison and unlimited fine. My company or my client, an unlimited fine. If the other person accepted the bribe, the same for them too I believe and their company.

 

[Troezar]

Anyone remember the recent emissions scandal? We've all got different risk thresholds. Some people actually get a buzz out of it so never say never

 

[DarrenM343]

well all reviewers that went to the nvidia event got a little holiday with horseriding and ofcourse all expenses paid, the cards free - does that make their opinions of the product and the company tainted? i think you would have to say yes

It does but we have to rely on the integrity of the reviewer to not let it affect the reviews , hence both sides have to take responsibility. Nvidia were not directly affecting the reviews or the sales of the product to the reviewer. If Nvidia later went on to offer them a 1to1 bribe that would have course be a bribe. If a reviewer asked for another or else they would get a bad review that would be a bribe.

It's not as blatant a bribe but I have to admit, in other countries that would I believe be a bit more frowned upon probably but not sure. As it involved multiple reviewers and part of the nature of the business(to show off products at such events), and they were given a review sample not a gift which is required to carry out the reviews, and done so publicly (it was part of a presentation) it's less of an issue (not as if it was done behind closed doors).

The whole event did cross my mind but the public nature of it makes it fine and the fact the reviewers were treated equally etc.

When AMD stopped sending review examples (rumour?) to someone who did a bad review on a Fury X ages ago that wasn't bribery either but I frowned on that as it is kinda suggesting of it (give our product a bad review and you too may not receive anymore samples).

Off topic but when Clarkson used to test cars given to him by the manufacturers I'm very sure the BBC would have kept a close eye on things, paying up where they needed to etc towards the cost but who knows. If Ferrari wanted to put him in a really expensive hotel, would the BBC pay half, would they deny it and put him in a cheap hotel? No idea really but interesting given that he could make or break the sales of a really expensive car. Or as a professional car reviewer would it all be acceptable because integrity is a huge part of the job?

 

[Mei]

It does but we have to rely on the integrity of the reviewer to not let it affect the reviews , hence both sides have to take responsibility. They were not directly affecting the reviews or the sales of the product. If Nvidia later went on to offer them a 1to1 bribe that would have course be a bribe. If a reviewer asked for another or else they would get a bad review that would be a bribe.

It's not as blatant a bribe but I have to admit, in other countries that would I believe be a bit more frowned upon probably. As it involved multiple reviewers and part of the nature of the business, and they were given a review sample not a gift, and done so publicly (it was part of a presentation) it's less of an issue (not as if it was done behind closed doors).

The whole event did cross my mind.

When AMD stopped sending review examples (rumour?) to someone who did a bad review on a Fury X ages ago that wasn't bribery either but I frowned on that as it is kinda suggesting of it.

my point was more when something can be measured, a metric, as opposed to something subjective, like if something is fun or ugly

tho they not called "youtube influencers" for nothing

do the masses have the mindeset to ignore all the opinions and weed through to the facts, i doubt it but they are atleast there in a non subjective thing

denying samples for review is much different to selectively choosing?, i duno it feels to me one day instead of hardware reviewers we should just have a app or something on steam to cut to the facts lol - a subjective thing like say headphones that would not be possible, but gfx cards and cpu's why not!?

 

[andybird123]

It was the FuryX's where they only sent review samples to sites that they felt would give them "fair" reviews, and several of the big review sites didnt get one

 

[CAT-THE-FIFTH]

Why don't some of you thread thrash the Pascal thread - why don't you go and bicker about Nvidia banning sites,etc in that one??

It just shows how much issues with E-PEEN some of you have,by ON PURPOSE try and keep the Pascal thread on track and not this one,by burying any news on Polaris in this thread.

 

[CAT-THE-FIFTH]

Just been meandering around the Pascal threads and spotted a post that is claiming that a leak of info states that the Gigabyte version of the 1080 (GTX 1080 Gaming G1) can reach up to 2.4Ghz.

Now that is fairly impressive, especially if they do come in @ $599 and not $699.

Now before you all flame me for coming into an AMD thread and saying this....please read on...

What some of those posters in the Pascal thread either didnt mention, didn't see or just completely ignored (take your pick)....was the info on the R9 470X that was below it (Actually the Headline of the actual piece was about the 470X NOT the 1080). Oh this was translated from Chinese so bear with it.

IT Home News May 13 message, just IT House exposure of the public version of the GTX overclocking 1080 , I believe many people have been core frequency of 2114MHz GTX sturdy performance of 1080 scared, but the latest news is that Gigabyte non-public version of the card is the GTX 1080 Gaming G1 core frequency can reach 2.4Ghz, compared to the public version of the top 13% frequency increase again.
Of course GTX 1080 5399 Yuan high price is not the average consumer can afford, we still tend to choose more graphics thousand or so, AMD is currently a low-end graphics card Radeon R9 470X exposure, alleged that the card uses the new Polaris 11 architecture, TDP of only 60W, that is to say no external auxiliary power supply to normal operation.
AMD R9 470X exposure: TDP is only 60W
It is reported that, R9 470X will use 1280 stream processors, 1GHz core frequency, 80 texture units, 40 raster units, the use of 4GB GDDR5 memory, memory frequency 7Ghz, memory interface is 128Bits, 112GB / S memory bandwidth .
The graphics performance is expected between R9 380 to R9 380X, as for the price, about 1,100 yuan, will be released in the second quarter of this year. I believe that focusing on power players, R9 470X is a good choice.

So we have the R9 470X with roughly equivalent performance between 380/380X and this comes in at 60W TDP with no power connectors so powered through the PCIe bus for 1,100 Yuan (Yes, I did the exchange rate for you) = £117 (Not sure if this is without taxes/Vat or whatever though). Not bad performance per £ considering the current 380/X is still selling on OCUK for between £158.99 - £199.99 and needs minimum 2x 6pin power connector.

Link is here http://www.ithome.com/html/digi/225244.htm

also more interesting info here on P10/P11 http://videocardz.com/59903/possible-polaris-10-and-polaris-11-specifications-emerge

Oh and just to let you know the exchange rate on the GTX1080 came in at approx £575

AMD debunks Polaris delay rumour from guy in Zoo:

http://www.eteknix.com/amd-debunks-recent-negative-rumours/

Mobile Polaris 10 and 11 GPU specs leaked:

http://videocardz.com/59903/possible-polaris-10-and-polaris-11-specifications-emerge

AMD to give an "inside look" at Polaris on May 18th:

http://videocardz.com/59923/amd-to-give-inside-look-at-polaris-on-may-18th

 

[DarrenM343]

Why don't some of you thread thrash the Pascal thread - why don't you go and bicker about Nvidia banning sites,etc in that one??

It just shows how much issues with E-PEEN some of you have,by ON PURPOSE try and keep the Pascal thread on track and not this one,by burying any news on Polaris in this thread.

Well, if people must make daft accusations of bribery in the thread supposedly aboout Polaris then expect someone to intervene.
Fair point tho but I had no idea which thread this was anyway.
I however hope some of you do care about ethics

 

[Mauller]

That 470x info if true at 1ghz, then the Polaris 10 will easily be furyx/overclocked 980ti performance. Just depends on at what clocks the Polaris 10 would be. That little Polaris 11 looks like it could reach 390-390x performance when clocked up.

 

[JediFragger]

It was the FuryX's where they only sent review samples to sites that they felt would give them "fair" reviews, and several of the big review sites didnt get one

Yup, that was dodgy as hell. They should let the product stand on its own merits!!

It's obvious that they knew that the FuryX was weak in certain areas and wanted to minimise the collateral damage. That worked out well for them...

 

[Mei]

i would but that thread is mostly about 21:9 monitors right now
seriously tho i think it was a healthy discussion that was talking about stuff that will impact the pascal launch and reviews like it has before

anyway i wont trash your thread more by arguing about im not trashing the thread lol - carry on!

 

[CAT-THE-FIFTH]

Polaris 10 launching at Computex:

http://videocardz.com/59947/amd-to-launch-polaris-10-at-computex

So both Polaris and Bristol Ridge will be launched at Computex. That means the Zen compatible AM4 platform is also launching then.

That 470x info if true at 1ghz, then the Polaris 10 will easily be furyx/overclocked 980ti performance. Just depends on at what clocks the Polaris 10 would be. That little Polaris 11 looks like it could reach 390-390x performance when clocked up.

http://videocardz.com/59932/amd-launches-radeon-m400-series

Supposedly most of the M400 series are rebrands but the AMD site does not state this.

 

[Mei]

why rebrand mobile parts when polaris uses have the power
looks like a perfect mobile chip to me
crazy

 

[CAT-THE-FIFTH]

why rebrand mobile parts when polaris uses have the power
looks like a perfect mobile chip to me
crazy

That's the thing - there is no information on the page saying what GPUs are being used. So the parts might be Polaris but it is hard to say.

Its all rather muddied ATM.

But at least we know the launch is happening.

 

[AlamoX]

that link is weird, there is no way they will rebrand bonaire again, polaris 11 have the same CU, it's supposed to replace it as M470, i might get behind the tonga because they must have a lot of stock as M480, so they might keep it around, and poalris 10 as M490, then M460 and lower get changed within the year.
but noway they will rebrand Bonaire, perf/watt is crap on that chip and it's so old it isnt even funny.

 

[CAT-THE-FIFTH]

The rumour is originally from WCCFTech,who looked at the AMD site and said they looked like rebrands. We actually don't know,since none of the cores were listed AFAIK.

 

[FoxEye]

The rumour is originally from WCCFTech,who looked at the AMD site and said they looked like rebrands. We actually don't know,since none of the cores were listed AFAIK.

I'd like to know why - if Polaris 11 is tiny, cost-effective to produce, low-power, good perf, and up-to-date with modern features - why you would want no less than TEN rebranded models below it in your mobile line-up?

Seems like a lot of wasted effort, as well as making no sense.