AMD Polaris architecture – GCN 4.0

[Kaapstad]

Actually we do, it has around 18 billion.

17 billion iirc.

This ^^^^

At least when people talk about Pascal they can do it with a rough idea of the performance. For example full fat Pascal will have about 1.8 to 2x the performance of a Maxwell TitanX.


The problem with this thread is there is no real information whatsoever about Polaris and it's likely performance.

Rumour has it NVidia are going to be calling the GP200 full fat card "The Titan Trident" as in a replacement for Polaris.

 

[Kaapstad]

Can you find Nvidia saying that anywhere? I didn't watch the recent press thing because, auto drive cars = boring. 17mil is a guesstimate that has been around for ages based off taking the current big Maxwell and taking that times the transistor density improvement of 16nm. That number has been around for ages and maybe it's been confirmed but was just a guess initially. The same way people state that big Polaris is between 15-18bil transistors. Nvidia tend to stick with 500-550mm^2 for their big GPU, AMD have only done a big gpu once since the 2900xt and largely only because 28nm was around so long.

With 10nm closer it's more likely for AMD to go back to a 370-450mm^2 core for 14/16nm and we'll be at 10nm before you could bring out a new gen at 14/16nm.

So what are the specs for Polaris or are you running on empty as usual ?

 

[LoadsaMoney]

[UPDATE : 2016 January 5, 10:43 AM ET] The first graphics card belonging to the brand new 14nm Polaris family is reportedly planned to launch within as little as 2 months. This new graphics card is going to be the smallest in the family with major focus on mobile platforms and power efficiency. More powerful graphics cards of the Polaris family are set to debut around mid 2016.

PCPer’s Ryan Shrout reports :

It is likely that this is the first Polaris GPU being brought up (after only 2 months I’m told) and could represent the best improvement in efficiency that we will see.

[UPDATE 2 : 2016 January 5, 11:59 AM ET] PCPer’s Ryan Shrout issued a clarification stating that the Polaris GPU AMD demoed was up and running after just two months from when initial production kicked off, rather than being brought up to market after two months.

lol from Wccf.

 

[Orangey]

Trident was the weapon of Triton, AIO cooler confirmed???

Also why are they bragging on 2 months for a test chip?

 

[Mtom]

17 billion on full fat Pascal

Your turn, how many on full fat Polaris ?

http://www.tweaktown.com/news/47165/amds-next-gen-gpus-feature-between-8-16gb-super-fast-hbm2/index.html

The new Greenland products will pack an insane 15-18 billion transistors,...

 

[Kaapstad]

http://www.tweaktown.com/news/47165/amds-next-gen-gpus-feature-between-8-16gb-super-fast-hbm2/index.html

The new Greenland products will pack an insane 15-18 billion transistors,...

Thanks.

That is good news as it will mean the AMD cards will be up there and competitive with NVidia's.

 

[Boomstick777]

The problem with this thread is there is no real information whatsoever about Polaris and it's likely performance.

Give it some time mate, Mr impatient The Pascal thread is the same way, tons of speculation etc. It's part of the fun. When we know all the facts and the mystery is gone, and also the price tag is revealed the fun bit will be over

 

[drunkenmaster]

This ^^^^

At least when people talk about Pascal they can do it with a rough idea of the performance. For example full fat Pascal will have about 1.8 to 2x the performance of a Maxwell TitanX.


The problem with this thread is there is no real information whatsoever about Polaris and it's likely performance.

Rumour has it NVidia are going to be calling the GP200 full fat card "The Titan Trident" as in a replacement for Polaris.

So what are the specs for Polaris or are you running on empty as usual ?

You're calling me out for specs or running on empty as usual(which is never the case) after you just posted expected performance of Pascal and saying the transistor count is known for it. Both of those pieces of information are RUMOURED, as is the transistor count for big Polaris. I said the Polaris transistor count was rumoured but you're asking me to confirm it, while you're using rumoured Nvidia specs as fact.

 

[drunkenmaster]

This ^^^^

At least when people talk about Pascal they can do it with a rough idea of the performance. For example full fat Pascal will have about 1.8 to 2x the performance of a Maxwell TitanX.


The problem with this thread is there is no real information whatsoever about Polaris and it's likely performance.

Rumour has it NVidia are going to be calling the GP200 full fat card "The Titan Trident" as in a replacement for Polaris.

http://www.tweaktown.com/news/46620...re-over-100-more-transistors-titan/index.html


Oh, and some of the bullcrap rumours that state 17billion also state 50% more performance. It's funny that you just pick and choose the rumours that suit you, dress them up as fact then get uppity when someone gives you a comparable rumour about the AMD cards.

Likewise leading up to a launch of an AMD card we get the Nvidia guys running their mouths about lack of information from AMD while Nvidia are apparently liberal with information, then after a press conference in which Nvidia say nothing useful about Pascal you get Nvidia guys saying Nvidia don't talk about GPUs until right before launch. You also get people criticise AMD for talking rather than doing, but when Nvidia do a press conference talking about products far into the future it's never a problem. Hypocrisy at it's most obvious, as usual.

 

[Kaapstad]

You're calling me out for specs or running on empty as usual(which is never the case) after you just posted expected performance of Pascal and saying the transistor count is known for it. Both of those pieces of information are RUMOURED, as is the transistor count for big Polaris. I said the Polaris transistor count was rumoured but you're asking me to confirm it, while you're using rumoured Nvidia specs as fact.

You could have just said you don't know.

You could have also said you don't know when it came to HBM1 as you got very little right there either.

How has HBM1 played out -

1. Unreliable.

2. Short supply.

3. Terrible 1080p performance due to low clockspeed.

You missed that lot above didn't you lol.

 

[Kaapstad]

Likewise leading up to a launch of an AMD card we get the Nvidia guys running their mouths about lack of information from AMD while Nvidia are apparently liberal with information, then after a press conference in which Nvidia say nothing useful about Pascal you get Nvidia guys saying Nvidia don't talk about GPUs until right before launch. You also get people criticise AMD for talking rather than doing, but when Nvidia do a press conference talking about products far into the future it's never a problem. Hypocrisy at it's most obvious, as usual.

How many AMD GPUs do you own DM ?

I own 12 AMD GPUs so I think you have scored another own goal lol.

 

[Gregster]

http://www.tweaktown.com/news/46620...re-over-100-more-transistors-titan/index.html


Oh, and some of the bullcrap rumours that state 17billion also state 50% more performance. It's funny that you just pick and choose the rumours that suit you, dress them up as fact then get uppity when someone gives you a comparable rumour about the AMD cards.

Likewise leading up to a launch of an AMD card we get the Nvidia guys running their mouths about lack of information from AMD while Nvidia are apparently liberal with information, then after a press conference in which Nvidia say nothing useful about Pascal you get Nvidia guys saying Nvidia don't talk about GPUs until right before launch. You also get people criticise AMD for talking rather than doing, but when Nvidia do a press conference talking about products far into the future it's never a problem. Hypocrisy at it's most obvious, as usual.

No need to be like that and talk objectively. No extra points for rudeness and I would like to think we are all adults. I never expect any info till close to launch but it is always good to speculate about what is what. You have a nasty habit of accusing people of this and that and then play the victim. Surely you want to see some healthy competition from both AMD and Nvidia? I personally enjoy all the rumours and what people surmise and it is harmless.

I am really excited to see what AMD bring and what Nvidia bring and will probably give them both a fair run out.

 

[Kaapstad]

Give it some time mate, Mr impatient The Pascal thread is the same way, tons of speculation etc. It's part of the fun. When we know all the facts and the mystery is gone, and also the price tag is revealed the fun bit will be over

I think with these type of threads it may be a good idea next time to grab the first 5 posts so you can cut and paste in all the news, rumours, videos and anything else so people can read them all in one place.

 

[Orangey]

There's no new info so it'll just be the same endlessly parroted stuff we've had for the last 3 years already.

Don't you all get bored of making the same posts again and again and again? That's why this place is dead lately.

 

[drunkenmaster]

You could have just said you don't know.

You could have also said you don't know when it came to HBM1 as you got very little right there either.

How has HBM1 played out -

1. Unreliable.

2. Short supply.

3. Terrible 1080p performance due to low clockspeed.

You missed that lot above didn't you lol.

I categorically was right about HBM1 ALL along, everything I said was accurate. I'm fairly sure I said it would be short supply also talking in depth about it being the first interposer product and it being a proof of concept product for both HBM1, interposers and the whole supply chain. But don't forget you argued against everything I said about HBM1 for 6 months before HBM1 was available and you were entirely wrong on absolutely everything, again.

How many AMD GPUs do you own DM ?

I own 12 AMD GPUs so I think you have scored another own goal lol.

Ah, because people can't own gpus and be biased, the argument with no logic in it. You consistently talk absolute crap about AMD, only ever talk up Nvidia, you jump into every argument to have a go at AMD and defend Nvidia. That you own AMD gpus doesn't prevent you doing any of this, obviously. Does give you a convenient excuse though.

No need to be like that and talk objectively. No extra points for rudeness and I would like to think we are all adults. I never expect any info till close to launch but it is always good to speculate about what is what. You have a nasty habit of accusing people of this and that and then play the victim. Surely you want to see some healthy competition from both AMD and Nvidia? I personally enjoy all the rumours and what people surmise and it is harmless.

I am really excited to see what AMD bring and what Nvidia bring and will probably give them both a fair run out.

Sorry but what is this post meant to say, literally every thread leading up to Fury X was completely full of Nvidia trolls screaming about AMD, bad marketing, refusing to say anything, literally full of it. When Nvidia do the same they defend it, this is fact, it is seen in every thread about an upcoming AMD release like clockwork. For years every single time it's "why are AMD refusing to talk about it", it's always the same 5-10 specific people and when it's explained why and how Nvidia do the same thing it's ignored, explained around or they go quiet.

AS for your post it was rude directly to me while I was posting factually about how posters behave in such threads. I wasn't being rude, Kaap was being rude. He both had a go at me in a rude manor in the first place and he was being a hypocrite. Complaining about lack of hard numbers/info from AMD and using Nvidia rumours as fact to again as the theme of my posts, push the idea Nvidia is giving out real information while AMD isn't which brings us full circle to how it's absolutely accurate because I'm directly responding to someone doing just that.

 

[drunkenmaster]

There's no new info so it'll just be the same endlessly parroted stuff we've had for the last 3 years already.

Don't you all get bored of making the same posts again and again and again? That's why this place is dead lately.

I've mostly stayed out of this forum for ages because every single time anything is posted about AMD, Kaap and others are there to bash everything, ask silly questions and take a silly stance. When I say something half sensible in defence of Nvidia in the other thread but Kaap and Gregster instantly jumped on my post as being negative for absolutely no reason at all.

Then in defending my point of view Greg decides to call me rude or angry while ignoring that both he and Kaap in both threads were the person who was being rude. This is their standard behaviour, hound anyone not talking up Nvidia into leaving the forum. Of course, didn't we see a image captured of him calling for back up. It's clear as day a clique of them hound anyone remotely positive about AMD or negative about Nvidia off this section of the forum. It's painful and the reason why I barely stay around here any more.

It's pitiful how certain people on here act, they've destroyed any ability to talk about technology. HBM1 threads were trashed by Kaap talking utter nonsense, as every other thread I see on here gets trashed, instigators are always the same group of people.

EDIT:- the longer some people go on and consistently use the "I bought a card from XXX thus I can't be biased" the more I'm convinced Nvidia fanboys after years of insults and attacks bought some AMD gpus just to use that excuse.

 

[Boomstick777]

^^

Yeah I must admit, I can see what DM is saying here it's getting a bit too salty lately. Kaap's even critiquing the way I posted this thread. Others don't get this level of scrutiny lol. Just posted the latest news, it all gets updated along the way.. Exactly the same in the Pascal thread. Oh well what can you do. Maybe lack of real progress in hardware is making people irritable

 

[Kaapstad]

I categorically was right about HBM1 ALL along, everything I said was accurate. I'm fairly sure I said it would be short supply also talking in depth about it being the first interposer product and it being a proof of concept product for both HBM1, interposers and the whole supply chain. But don't forget you argued against everything I said about HBM1 for 6 months before HBM1 was available and you were entirely wrong on absolutely everything, again.

You talked a great deal about HBM1 and very little of it came to pass.

Ah, because people can't own gpus and be biased, the argument with no logic in it. You consistently talk absolute crap about AMD, only ever talk up Nvidia, you jump into every argument to have a go at AMD and defend Nvidia. That you own AMD gpus doesn't prevent you doing any of this, obviously. Does give you a convenient excuse though.

I am equally critical of NVidia hardware but you fail to notice that.

Some of the things that I have been critical of NVidia lately have been.

TWIMTBP - are they serious.
SLI - or rather the lack of it.
NVidia pricing.
G-Sync - I hate closed systems.
NVidia drivers - numerous reasons.
9 series bus size - not good for high resolutions.

Perhaps you could share with us what you don't like about AMD, or are you biased lol.

 

[Kaapstad]

Kaap's even critiquing the way I posted this thread.

That was a suggestion not criticising.

 

[Scougar]

I've mostly stayed out of this forum for ages because every single time anything is posted about AMD, Kaap and others are there to bash everything, ask silly questions and take a silly stance. When I say something half sensible in defence of Nvidia in the other thread but Kaap and Gregster instantly jumped on my post as being negative for absolutely no reason at all...... *SNIP*

I realized long ago that this forum rarely discusses things sensibly, or rarely gets anything achieved with the discussion. This place is almost the equivalent of reddit at times, with people spouting nonsense and arguing totally off-topic points for goodness knows what reason (JUST LIKE THIS POST I'M MAKING!!). I like it because it's relatively friendly due to the rules and highly active. If the rules weren't here and good moderation, this place would turn into a cesspit in a very short period of time.

It is laughable people trying to act adult and tell others to be more mature when in their recent past they have acted outright childish. Smell your own hole people and have some humility.

------------------

Anywho... Kaap, what do you mean about HBM being unreliable? Can you give some examples? On a separate note, did you ever get your watercooled fury x rig together for benching?

To me, quite often a new tech in some area's is a backwards step until it matures a little, that's just life.

With talk of culling more triangles from the scene, to me that seems really weird, because you would think they would have already tried all that kind of stuff when trying to make things more efficient, and not do unnecessary work. Dunno, but it just struck me as odd.

I'm looking forward to the next gen AMD/Nvidia, and thanks to this news have stopped looking at other cards for now and will continue to save for a next gen card even though I know I'll be an early adopter.