The AMD Mantle Thread

[Gloomy]

Huh, it shows the 780 decreasing by similar amounts to the 290X under DX when CPU speed decreases.

But that can't be, because Nvidia has better drivers??

 

[jj109]

Huh, it shows the 780 decreasing by similar amounts to the 290X under DX when CPU speed decreases.

But that can't be, because Nvidia has better drivers??

Maybe a less confusing representation here:

Feels good to know that I was right about the drivers though.

 

[ShintaiDK]

Source: http://www.betazeta.com/#!/necronomicon/chw/post/amd-cumple-las-expectativas-con-mantle

BTW, how much more does the GTX 780 GHz cost? Looks like $550+ vs. ~$450 for the standard edition. I don't see many listings here.

You do know that site/blog is created by a person listed 277 on AMD influencer leaderboard? So lets not use those numbers at all since their credibility is zero.

 

[Skurge]

You do know that site is created by a person listed 277 on AMD influencer leaderboard? Zero credibility for linking shills.

He obviously isn't doing a very good job.

 

[jj109]

It's funny seeing Mantle make the i3-4330 a better buy for BF4 than the A10-7850K or even the FX-8350

 

[Gloomy]

Maybe a less confusing representation here:

Feels good to know that I was right about the drivers though.

Interesting to note another review that gets similar scores with the "780GHz" as pclab.pl

In fact, those scores are identical to the stock CPU scores pclab got.

Because those are the same EXACT scores.

http://pclab.pl/art55953-3.html

 

[jj109]

Interesting to note another review that gets similar scores with the "780GHz" as pclab.pl

In fact, those scores are identical to the stock CPU scores pclab got.

Because those are the same EXACT scores.

http://pclab.pl/art55953-3.html

Now that I read the article, looks like it's just Spanish blog spam reposting other people's benchmarks.

 

[ShintaiDK]

Interesting to note another review that gets similar scores with the "780GHz" as pclab.pl

In fact, those scores are identical to the stock CPU scores pclab got.

Because those are the same EXACT scores.

http://pclab.pl/art55953-3.html

Yep, exactly same.

 

[RampantAndroid]

Yep, exactly same.

Is the data good, or is he mashing data together so it means nothing in the end?

 

[Gloomy]

Is the data good, or is he mashing data together so it means nothing in the end?

Besides the fastest 780 on the planet, it seems okay.

Please go to the original source: http://pclab.pl/art55953-3.html

It has overclocked CPU results. That 780 GHz is bottlenecked with a stock 4770K, and probably with the 4.5GHz 4770K too from the looks of it.

 

[ShintaiDK]

Is the data good, or is he mashing data together so it means nothing in the end?

He forgot OC data, didnt specify what map is used. MP or SP etc.

 

[3DVagabond]

Maybe a less confusing representation here:

Feels good to know that I was right about the drivers though.

So when did the 780GHz become 30% faster than the 290X using a 4770K? Sorry, not buying it.

See any discrepancies?

He obviously isn't doing a very good job.

lol :thumbsup: I think AMD needs to change the criteria.

 

[JoeRambo]

If Triple buffering is implemented properly you should get smooth fps all the way from 0 to 60 (or whatever the refresh rate of your monitor is).

Triple buffering is dead due to input lag it causes, seeing a frame rendered tens of miliseconds away is not that great. Only real solution in smooth fps stage is Nvidia's Gsync - displaying frame as soon as it is ready, making FPS=refresh rate hz within capabilities of panel.

 

[Gloomy]

Recent patch boosted Nvidia performance lots. I doubt it's 30% though I've yet to see any other site's bench to compare with.

If that 780GHz exists, I want it. Bad.

 

[ShintaiDK]

Recent patch boosted Nvidia performance lots. I doubt it's 30% though I've yet to see any other site's bench to compare with.

If that 780GHz exists, I want it. Bad.

Well you can buy both GTX780 and 780TI with 1Ghz+ base clocks

 

[3DVagabond]

Besides the fastest 780 on the planet, it seems okay.

Please go to the original source: http://pclab.pl/art55953-3.html

It has overclocked CPU results. That 780 GHz is bottlenecked with a stock 4770K, and probably with the 4.5GHz 4770K too from the looks of it.

The fastest 780 on the planet and the slowest 290X.

 

[Gloomy]

The fastest 780 on the planet and the slowest 290X.

No, the 290X numbers seem accurate to me.

Here's another site (compterbase.de) with a stock 4770K:

Both the DX and Mantle numbers are close enough for me.

Like I said though, that 780 they used is something special. Blessed by Jason Hung Swang himself and anointed with the sweat from his massive arms...

 

[Silverforce11]

Would like to see some 64MP R290X CF action with Mantle, see if its true what Johan released, i7 + R290X CF = 50% faster with Mantle vs DX.

 

[jj109]

The fastest 780 on the planet and the slowest 290X.

The 780 GHz is almost as fast as a 780 Ti and the 290X might be throttling due to load. It's certainly possible that they also have a lemon. Or it might be the new beta drivers being wonky. Remember this?

In any case, those 780 GHz numbers are real. Siege of Shanghai is incredibly CPU intensive, but the only time I dip consistently below 80 FPS is while looking across the bay from the US side. I say CPU intensive because I couldn't get GPU utilization above 90% unless the FPS was near 100. I recorded a short clip here:

http://www.youtube.com/watch?v=9cVRCQmN0XA

My FPS should also be a bit higher than shown since ShadowPlay has its small but measurable performance hit.

 

[desprado]

Would like to see some 64MP R290X CF action with Mantle, see if its true what Johan released, i7 + R290X CF = 50% faster with Mantle vs DX.

I have done that and i am only getting 7% of improvement but what is funny is that i am getting 101fps with Dx in the same map.

 

[omeds]

So when did the 780GHz become 30% faster than the 290X using a 4770K? Sorry, not buying it.

Since win8.1 when CPU bound. Mantle certainly closes that gap though, or surpasses it.

 

[Silverforce11]

I have done that and i am only getting 7% of improvement but what is funny is that i am getting 101fps with Dx in the same map.

Thats pretty meh since users on other forums have massive gains with CF.

Thats why I wanna see some legit sites do it..

 

[Dresdenboy]

You do know that site/blog is created by a person listed 277 on AMD influencer leaderboard? So lets not use those numbers at all since their credibility is zero.

I don't care about the site's content, since the author used data of 2 reviews as he stated at the end of his article and in the text: pclab and wccftech. So all discussion about that being the same scores is void. And it removes any influence of the blog's history. I just found it interesting, that the author scaled the scores based on 4770K = 100%. This at least is his own work. And mistakes should be easy to detect.

Where is this leaderboard, btw? I didn't know this blog before, just found it in google image search.

 

[Gloomy]

I stand corrected. This was an amazing patch for Nvidia cards then.

 

[Enigmoid]

Is there any difinitive word on whether there has been an IQ regression with mantle? The mantle screenshots that I've seen seem to have more fog.