• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

AMD Polaris architecture – GCN 4.0

Mauller said:
Will just have to wait and see, they may surprise everyone like they have been saying since Polaris was announced.

I don't think it will clock anywhere as high as the 1080, but it will certainly get a higher max clock than what people expect.
Click to expand...

AMD has higher IPC and denser transistor die.
Main reason for example maxwell didnt OC as well as the 1080 either as it wasnt built with brute force and high clockspeed.
Pascal do a software async compute and its also a question of HDR support atm.

Unless something really suprising Polaris will be a king with price/performance and OC allows some additional value along the way.
 
flopper said:
AMD has higher IPC and denser transistor die.
Main reason for example maxwell didnt OC as well as the 1080 either as it wasnt built with brute force and high clockspeed.
Pascal do a software async compute and its also a question of HDR support atm.

Unless something really suprising Polaris will be a king with price/performance and OC allows some additional value along the way.
Click to expand...

I'd say maxwell OC'ed far better then fury :p

Have to say if maxwell clocked at 2ghz+ speeds on reference coolers AMD's card would have looked terrible ad way less competive

Hoping AMD wipes the floor with Nvidia personally this round by a good chunk
 
Im so willing AMD to do well this round. I really despise Nvidia. I don't like their business practices, I don't like them lying to customers, I don't like them hiding tessellation to gimp amd cards in games. I don't like how up themselves they are.

However, I am a consumer. And if they have a card that does more FPS for less money than AMD. I will buy it.
 
beany_bot said:
Im so willing AMD to do well this round. I really despise Nvidia. I don't like their business practices, I don't like them lying to customers, I don't like them hiding tessellation to gimp amd cards in games. I don't like how up themselves they are.

However, I am a consumer. And if they have a card that does more FPS for less money than AMD. I will buy it.
Click to expand...

Well then wait while both side release their cards, then decide.
 
So how would polaris not be clocked high (1.3ghz+) on 16nm I don't understand that. Surely it would overclock well if it started at 1ghz?
 
AMD set to host a special live stream for their Pascal architecture on Wednesday
Click to expand...

Amd cannot beat them so they naming them the same to get Nvidia buyers to buy them just hope Lambchop doesn't buy a Amd Pascal he have a heart attack :p.
 
Lokken86 said:
So how would polaris not be clocked high (1.3ghz+) on 16nm I don't understand that. Surely it would overclock well if it started at 1ghz?
Click to expand...

Clock speed depends entirely on the architecture. If AMD decided to use the smaller process to increase IPC then the instructions per a second will go up at the same clock speed.


Given particular fabrication process you can either design a processor with a higher IPC and lower clock speeds, or lower IPC and higher clock speeds. there is no definitive benefit to either approach.
 
D.P. said:
Clock speed depends entirely on the architecture. If AMD decided to use the smaller process to increase IPC then the instructions per a second will go up at the same clock speed.


Given particular fabrication process you can either design a processor with a higher IPC and lower clock speeds, or lower IPC and higher clock speeds. there is no definitive benefit to either approach.
Click to expand...

Show me an instance where clock speeds went backwards on a new node. I can't think of one. When designing a new architecture on a new node they always increase clock speeds as that's a part of gaining performance when a shrink happens.
 
TheRealDeal said:
Show me an instance where clock speeds went backwards on a new node. I can't think of one. When designing a new architecture on a new node they always increase clock speeds as that's a part of gaining performance when a shrink happens.
Click to expand...

doubt you can compare each process being different
tho im not geeky enough to know
nvidia apparently went slightly lower ipc for higher clocks and also sacrificed some power gains for performance
maybe amd will go more for power gains, who knows

super silly to me if they dont put them into mobile parts soon tho!
 
Mei said:
doubt you can compare each process being different
tho im not geeky enough to know
nvidia apparently went slightly lower ipc for higher clocks and also sacrificed some power gains for performance
maybe amd will go more for power gains, who knows

super silly to me if they dont put them into mobile parts soon tho!
Click to expand...

The latest slides show the architecture and for all intents and purposes it is Maxwell with only the clock speed boosted thanks to the new 16nm node.

1080
http://wccftech.com/nvidia-geforce-gtx-1080-official-slides-leak/
http://cdn.wccftech.com/wp-content/uploads/2016/05/NVIDIA-GP104-GPU-Block-Diagram-635x531.jpg


980ti
[URL="http://www.guru3d.com/articles-pages/nvidia-geforce-gtx-980-ti-review,6.html"]http://www.guru3d.com/articles-pages/nvidia-geforce-gtx-980-ti-review,6.html[/URL]
[IMG]http://www.guru3d.com/index.php?ct=articles&action=file&id=15227&admin=0a8fcaad6b03da6a6895d1ada2e171002a287bc

If AMD had gone for a Fiji shrink I'm pretty sure we would get a 50% clockspeed boost and maybe a 40-50% performance boost to match putting it right where the 1070/1080 are. With GCN improvements and a high clockspeed, Polaris may well be a Fury level card with the lower number of stream processors.

[B][COLOR="Orange"]** No Hotlinking **[/COLOR][/B]
 
Last edited:
Mei said:
doubt you can compare each process being different
tho im not geeky enough to know
nvidia apparently went slightly lower ipc for higher clocks and also sacrificed some power gains for performance
maybe amd will go more for power gains, who knows

super silly to me if they dont put them into mobile parts soon tho!
Click to expand...

I don't think they did. I think they mainly scaled down the spec and upped the clocks. We might have an answer soon as i think the gtx1070 will resemble a gtx980 in spec apart from having more memory. An underclocked 1070 could answer the question and i suspect tech sights may do this.
 
You must log in or register to reply here.
Back
Top Bottom