• 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

GFX-Kid said:
That is all great, but will any of it translate into gaming performance? Hopefully yes :)
Click to expand...
N19h7m4r3 said:
Let's hope! My main concern is the single compute. It's much slower than the 980Ti.

Single Compute 5312.60 vs 9005.12
Click to expand...

Unfortunately you are quite right. We all know AMD technically is superior to Nvidia when it comes to dx12, due to to having Async Shaders support at hardware level, when Nvidia existing cards don't and is only supported at software level, along with their upcoming card won't have support at hardware level but only a workaround. Because of this, but we know AMD will have higher power consumption and heat, even if delivering the same frame rate/performance.

But you can bet of Nvidia to throw their weight around with developers to keep the use of Async Shaders to as low as possible, so AMD's hardware advantage won't mean much in reality, and Nvidia will pull ahead in general performance and win in benchmarks, due to not bothering with Async Shaders but continue to focus only on improving performance for dated approach, at the cost holding back the one of the potential biggest benefit of dx12 for PC gaming for the sake of winning against the competition and maximising their profit.
 
Last edited:
Marine-RX179 said:
Unfortunately you are quite right. We all know AMD technically is superior to Nvidia when it comes to dx12, due to to having Async Shaders support at hardware level, when Nvidia existing cards don't and is only supported at software level, along with their upcoming card won't have support at hardware level but only a workaround. Because of this, but we know AMD will have higher power consumption and heat, even if delivering the same frame rate/performance.

But you can bet of Nvidia to throw their weight around with developers to keep the use of Async Shaders to as low as possible, so AMD's hardware advantage won't mean much in reality, and Nvidia will pull ahead in general performance and win in benchmarks, due to not bothering with Async Shaders but continue to focus only on improving performance for dated approach, at the cost holding back the one of the potential biggest benefit of dx12 for PC gaming for the sake of winning against the competition and maximising their profit.
Click to expand...

A fair number of the most intensive Gameworks effects tend to push tessellation since Nvidia cards are stronger at geometry. If AMD has made big improvements to tessellation with Polaris they might not have to rely so much on Async to be competitive.
 
Last edited:
Marine-RX179 said:
But you can bet of Nvidia to throw their weight around with developers to keep the use of Async Shaders to as low as possible, so AMD's hardware advantage won't mean much in reality, and Nvidia will pull ahead in general performance and win in benchmarks, due to not bothering with Async Shaders but continue to focus only on improving performance for dated approach, at the cost holding back the one of the potential biggest benefit of dx12 for PC gaming for the sake of winning against the competition and maximising their profit.
Click to expand...
Got the excuses in the post already I see.
 
Take this with lots of salt:

Khalid on wccf has commented in his comments section that he is very confident that P10 uncut reference core clock will be between 1300 and 1400 MHz because he has "seen them."
Click to expand...

That is the senior editor saying that - could be a load of tosh,but 2304 shaders clocked at 1.3GHZ+ would make the core faster than a Hawaii one with 2816 shaders running at 1.05GHZ!!

JPR says the following too:

Interestingly, JPR found that only 15% of all GPUs sold are priced at $349 or higher. And of that 15% only 3% are priced at $449 or higher. This means that AMD’s Polaris GPUs are aimed at market segments where the volumes are inherently significantly larger compared to Nvidia’s GTX 1080 and GTX 1070.
Click to expand...

http://wccftech.com/amd-takes-gpu-share-nvidia-q1-2016/
 
Last edited:
Just realised AMD that Polaris 10 might do better than an R9 390X - now it could be that level,but I had forgotten about the HD6870. The HD6870 had 1120 shaders at 900MHZ and was less than 10% slower than an HD5870 with 1600 shaders running at 850MHZ which had nearly 15% more bandwidth.
 
CAT-THE-FIFTH said:
Just realised AMD that Polaris 10 might do better than an R9 390X - now it could be that level,but I had forgotten about the HD6870. The HD6870 had 1120 shaders at 900MHZ and was less than 10% slower than an HD5870 with 1600 shaders running at 850MHZ which had nearly 15% more bandwidth.
Click to expand...

I think they went from a VLIW5 architecture with the 5xxx series to a VLIW4 architecture with the 6xxx series, meaning the shaders were also a bit smaller.
 
Worried the 1070 is going to be over £400 now at launch.

Come on AMD, give me something to go on before June 10th! Just get around 980Ti performance for sub £300, lower power, heat, and noise and we can talk!

Oh, and sort out the Linux drivers.
 
Tute said:
Worried the 1070 is going to be over £400 now at launch.

Come on AMD, give me something to go on before June 10th! Just get around 980Ti performance for sub £300, lower power, heat, and noise and we can talk!

Oh, and sort out the Linux drivers.
Click to expand...

If they manage that then I'll be buying 2 of em. It's a big ask though.
 
You must log in or register to reply here.
Back
Top Bottom