Go Back   Hardware Canucks > HARDWARE > Video Cards

    
Reply
 
LinkBack Thread Tools Display Modes
  #71 (permalink)  
Old August 17, 2012, 08:15 AM
great_big_abyss's Avatar
Hall Of Fame
F@H
 
Join Date: Oct 2011
Location: Winnipeg
Posts: 2,210

My System Specs

Default

Quote:
Originally Posted by eyefinity View Post
Here -

[SNIP]

How does that back up any argument that the 660 Ti is on par with the 7950? 3 different overclocked cards can't even win in two TWIMTBP games vs the 7870.
Eyefinity,

1) your first graph is taken from Just Cause 2 using full 8xAA and 16xAF. It's pretty well known that the GTX660Ti slows down with those features turned on, simply because of memory bandwith. Turn them off (like many, many gamers do) and the 660Ti will soar past the 7870 and even beat out the 7950 at times.

2) Dirt Showdown is a game that has been optimised for AMD graphics cards. AMD worked with the developers, and it was designed almost as a showcase for AMD. It has impeccable Eyefinity experiences, and works better on any AMD card than just about any other game. I wouldn't exactly call it a reputable benchmark for that reason.

3) Batman Arkham City is basically a wash. The framerates for ALL cards are so close it really doesn't make a difference, and doesn't prove anything. And, once again, all AA is turned on full, providing a slight advantage to the AMD cards. Throw a GTX670/680 in there and see what happens. They will run away with it.

Face it, Nvidia has winners with ALL the kepler cards. And the GTX660Ti is just another example of it. That's not to say that AMD's efforts aren't relevant, or even a good option. It's just that the new GTX6xx series seem to be a little better. And I'm saying that with my feet firmly in the AMD camp. I've been running various AMD cards (and processors) for the last 10 years, ever since the original Athlon and ATI X800 days.

Eyefinity, you need to give respect where it's due, and stop arguing inane points with cherrypicked arguments.
__________________



HTPC: Z77A-G45; 3770K; Coolermaster GeminII; 2x4GB Kingston HyperX 1600Mhz; MSI R7-260X; 2x 128GB Crucial M4 SSD; 1TB WD Green, 2x 2TB WD Green; PC P&C 750W PS; Fractal Design Node 605;
Son's Rig: M5A97; 1055T; CNPS20LQ; 2x4GB Corsair Vengeance 1600Mhz; ASUS GTX650Ti Boost; 80GB Intel 520 SSD; 320GB WD Black HDD; SPI 700W; Bitfenix Shinobi;
Reply With Quote
  #72 (permalink)  
Old August 17, 2012, 08:24 AM
Banned
 
Join Date: Aug 2012
Posts: 28
Default

Quote:
Originally Posted by mattlef View Post
Look at methodology. Synthetic benchmarks never equal real life gameplay. Ever.
That also has to come into play. Drivers can make synthetic benchmarking look much better than it is.
And we're supposed to trust reviewers with a mouse/keyboard and fraps instead? Do you think that one of the reasons why Nvidia performs so well here?

Run fraps and do 2 similar benchmark runs. On the second one, look around a bit more. See the difference in the results.

That is why you don't trust reviews that are manually doing benchmarks. You cannot guarantee they are done within acceptable differences.
Reply With Quote
  #73 (permalink)  
Old August 17, 2012, 08:28 AM
Skyllz's Avatar
Allstar
F@H
 
Join Date: Sep 2008
Location: Ottawa
Posts: 777

My System Specs

Default

And here you are arguing over a less than 5fps difference in a static benchmark...

Benchmark are like car reviews... Everyone has their own way to compare things. There is no bad method, as long as it's consistent and HWC has always have.
__________________
Is smoke supposed to come out of there?
Reply With Quote
  #74 (permalink)  
Old August 17, 2012, 08:29 AM
Banned
 
Join Date: Aug 2012
Posts: 28
Default

Quote:
Originally Posted by great_big_abyss View Post
Eyefinity,

1) your first graph is taken from Just Cause 2 using full 8xAA and 16xAF. It's pretty well known that the GTX660Ti slows down with those features turned on, simply because of memory bandwith. Turn them off (like many, many gamers do) and the 660Ti will soar past the 7870 and even beat out the 7950 at times.

2) Dirt Showdown is a game that has been optimised for AMD graphics cards. AMD worked with the developers, and it was designed almost as a showcase for AMD. It has impeccable Eyefinity experiences, and works better on any AMD card than just about any other game. I wouldn't exactly call it a reputable benchmark for that reason.

3) Batman Arkham City is basically a wash. The framerates for ALL cards are so close it really doesn't make a difference, and doesn't prove anything. And, once again, all AA is turned on full, providing a slight advantage to the AMD cards. Throw a GTX670/680 in there and see what happens. They will run away with it.

Face it, Nvidia has winners with ALL the kepler cards. And the GTX660Ti is just another example of it. That's not to say that AMD's efforts aren't relevant, or even a good option. It's just that the new GTX6xx series seem to be a little better. And I'm saying that with my feet firmly in the AMD camp. I've been running various AMD cards (and processors) for the last 10 years, ever since the original Athlon and ATI X800 days.

Eyefinity, you need to give respect where it's due, and stop arguing inane points with cherrypicked arguments.
So basically what you're saying is, benchmark weaker settings and the Nvidia cards perform better. What happens in a year when games are a bit more complex and those same settings are now that bit harder on cards with lower bandwidth?

Funny how nobody mentions that hm? So we get a bunch of lower IQ settings and you don't bother to benchmark AMD favoured games because Nvidia says so, then lets try out overclocked Ti's vs stock Radeons as well at some sites, I mean why the hell not? Furthermore, why not throw your dummy out of the pram and refuse to benchmark the 7950 with it's new boost? How many of those can you say about this site? 3 out of 4?

If you can't see what's going on here you are more than a little bit "green".

Anandtech tried to tell us that the 560 Ti was faster than the 6950 on review day, then 6 months later people started noticing how far behind it was in their bench. The *exact* same thing is going to happen here, and this 660 Ti will be found wanting, very, very far behind the 7950.
Reply With Quote
  #75 (permalink)  
Old August 17, 2012, 08:33 AM
SKYMTL's Avatar
HardwareCanuck Review Editor
 
Join Date: Feb 2007
Location: Montreal
Posts: 11,779
Default

Quote:
Originally Posted by eyefinity View Post
Yeah and you've never used games that wildly favoured Nvidia? We all know who uses their initiative the most to hobble the opposing cards.

As for not using Dirt Showdown - it couldn't possibly be that you don't want to lose Nvidia's huge advantage in Dirt 3 could it?

Or maybe it was because Nvidia asked you nicely in the reviewer guide? Yes this is the point - you can tell which sites are Nvidia biased because they are the ones that didn't benchmark Dirt Showdown.
Can you show me a single game we are currently using that is a TWIMTBP title and uses specific routines that are meant to handicap the competitors' products? And actually, AMD is using their initiative FAR MORE right now to hobble NVIDIA's performance. Sniper Elite, Dirt Showdown and Shogun 2 have ALL received either updates or were launched with visual features that were specifically designed for this purpose.

As for Dirt 3, you do realize that it is a Gaming Evolved title, right? NVIDIA's performance advantage was attained through driver optimizations after (once again) being confronted with AMD-sponsored features.

No, sites that used Dirt Showdown should be considered biased against NVIDIA. But....most aren't.


Ironically, you show the HD 7870 running neck and neck with the GTX 660 Ti at high resolutions. As do we. You show the GTX 660 Ti falling by the wayside as MSAA is increased at 2560....we show the same and call it out in the conclusion.

Is the HD 7870 faster than the GTX 660 Ti on average at 1080P / 1200P? Hell...NO. Any sane person that's been looking at the overall picture wouldn't think otherwise.
__________________
Reply With Quote
  #76 (permalink)  
Old August 17, 2012, 08:45 AM
great_big_abyss's Avatar
Hall Of Fame
F@H
 
Join Date: Oct 2011
Location: Winnipeg
Posts: 2,210

My System Specs

Default

Quote:
Originally Posted by eyefinity View Post
So basically what you're saying is, benchmark weaker settings and the Nvidia cards perform better. What happens in a year when games are a bit more complex and those same settings are now that bit harder on cards with lower bandwidth?

Funny how nobody mentions that hm? So we get a bunch of lower IQ settings and you don't bother to benchmark AMD favoured games because Nvidia says so, then lets try out overclocked Ti's vs stock Radeons as well at some sites, I mean why the hell not? Furthermore, why not throw your dummy out of the pram and refuse to benchmark the 7950 with it's new boost? How many of those can you say about this site? 3 out of 4?

If you can't see what's going on here you are more than a little bit "green".

Anandtech tried to tell us that the 560 Ti was faster than the 6950 on review day, then 6 months later people started noticing how far behind it was in their bench. The *exact* same thing is going to happen here, and this 660 Ti will be found wanting, very, very far behind the 7950.
Your argument is that the 7870 far surpasses the 660Ti. No, it does not. Yes, the 7950 is slightly faster than the 660Ti. It also costs a little more (about $30), uses more power, and is noisier. But then again, nobody is arguing that. They're just saying that for the money the 660Ti is competitive with the 7950!

Oh, and Dude, do you work for AMD? Do you know a fellow named Raghu78? Are you related to him?

Because at the rate you're going, calling everybody out and accusing reviewers here of being monetarily rewarded for bias, you're on the same track as he is. We're a peaceful community here, and we don't like groundshakers. We can all have an argument and resolve differences of opinions peacefully and without slander.
__________________



HTPC: Z77A-G45; 3770K; Coolermaster GeminII; 2x4GB Kingston HyperX 1600Mhz; MSI R7-260X; 2x 128GB Crucial M4 SSD; 1TB WD Green, 2x 2TB WD Green; PC P&C 750W PS; Fractal Design Node 605;
Son's Rig: M5A97; 1055T; CNPS20LQ; 2x4GB Corsair Vengeance 1600Mhz; ASUS GTX650Ti Boost; 80GB Intel 520 SSD; 320GB WD Black HDD; SPI 700W; Bitfenix Shinobi;
Reply With Quote
  #77 (permalink)  
Old August 17, 2012, 08:48 AM
SKYMTL's Avatar
HardwareCanuck Review Editor
 
Join Date: Feb 2007
Location: Montreal
Posts: 11,779
Default

Quote:
Originally Posted by eyefinity View Post
And we're supposed to trust reviewers with a mouse/keyboard and fraps instead? Do you think that one of the reasons why Nvidia performs so well here?

Run fraps and do 2 similar benchmark runs. On the second one, look around a bit more. See the difference in the results.
Oh yes, it is a VAST conspiracy.

We've always been completely transparent in how we benchmark games. AMD has received just as many positive reviews here as NVIDIA. So where's the bias?

Quote:
Originally Posted by eyefinity View Post
So basically what you're saying is, benchmark weaker settings and the Nvidia cards perform better. What happens in a year when games are a bit more complex and those same settings are now that bit harder on cards with lower bandwidth?
Actually, you have it wrong. As DX11 or any API for that matter matures, complexity increases but developer and driver optimizations vastly outpace it. This is why most new DX11 games (other than the ones that are artificially hobbled by poor coding) actually perform BETTER in DX11 mode. As such, hopefully we will need LESS POWERFUL hardware as time goes on. We should all be happy for that. Note that I am NOT bringing the GPUs in question into this since it is besides the point.

Quote:
Funny how nobody mentions that hm? So we get a bunch of lower IQ settings and you don't bother to benchmark AMD favoured games because Nvidia says so, then lets try out overclocked Ti's vs stock Radeons as well at some sites, I mean why the hell not? Furthermore, why not throw your dummy out of the pram and refuse to benchmark the 7950 with it's new boost? How many of those can you say about this site? 3 out of 4?
Again with the conspiracy theories.

We have two TWIMTBP games: Metro 2033 and Batman
We have THREE Gaming Evolved games: Dirt 3, Deus Ex and Shogun 2

All of the other titles are either joint projects for Eyefinity / Surround initiatives (Crysis 2 and Witcher 2) or claimed by NO ONE (Battlefield 3, Wargame and Skyrim


Our refusal to benchmark the HD 7950 Boost was well publicized: a lack of time and a real lack of pricing information. Oh, by the way, the Boost version is going for $30 more than what we had the reference cards pegged at. Where's your outrage about that??
__________________
Reply With Quote
  #78 (permalink)  
Old August 17, 2012, 08:53 AM
Banned
 
Join Date: Aug 2012
Posts: 28
Default

Quote:
Originally Posted by SKYMTL View Post
Can you show me a single game we are currently using that is a TWIMTBP title and uses specific routines that are meant to handicap the competitors' products? And actually, AMD is using their initiative FAR MORE right now to hobble NVIDIA's performance. Sniper Elite, Dirt Showdown and Shogun 2 have ALL received either updates or were launched with visual features that were specifically designed for this purpose.
Err..Batman and Metro? And Crysis 2? And I'm sure you're well aware of the "special relationship" between Bethesda and Nvidia so you might as well add Skyrim to that.

On the flipside you've managed to find the AMD sponsored titles that Nvidia does better in. Shogun and Dirt 3 at least. However you look at it in terms of "hobbling", you are using 5 extremely Nvidia-centric titles and 2 out of 3 AMD titles are heavily favouring Nvidia in benchmarks.

Quote:
As for Dirt 3, you do realize that it is a Gaming Evolved title, right? NVIDIA's performance advantage was attained through driver optimizations after (once again) being confronted with AMD-sponsored features.
No, sites that used Dirt Showdown should be considered biased against NVIDIA. But....most aren't.
Yeah right. Let's tesselate the hell out of some jersey barriers or hidden underground water and benchmark that instead, that's a lot fairer, right?

Exactly what is it that AMD is locking Nvidia out of regarding Dirt Showdown? If the cards aren't up the the task then you simply have to bench at lower settings and inform the reader that the Nvidia card's aren't capable of running the game acceptably at highest quality settings. Isn't that what most sites have been doing regarding PhysX for years?

AMD should be leveraging their superior architecture in order to gain an advantage - Nvidia has been doing it for years, on top of plain bribery and threats.
Reply With Quote
  #79 (permalink)  
Old August 17, 2012, 09:00 AM
great_big_abyss's Avatar
Hall Of Fame
F@H
 
Join Date: Oct 2011
Location: Winnipeg
Posts: 2,210

My System Specs

Default

Quote:
Originally Posted by eyefinity View Post
On the flipside you've managed to find the AMD sponsored titles that Nvidia does better in. Shogun and Dirt 3 at least. However you look at it in terms of "hobbling", you are using 5 extremely Nvidia-centric titles and 2 out of 3 AMD titles are heavily favouring Nvidia in benchmarks.
Um, every benchmark shows every single AMD card VASTLY outperforming all the Nvidia cards. Now you're just plain lying to suit your purposes.

Oh, and Dude, take off your tinfoil hat and open your eyes!

Also, Why is this such a big deal to you? Do you have shares in AMD? Is your life in any way directly affected by how well AMD does?

Really, who cares. Ultimately these are consumer products that we're talking about. You buy one, use it for a couple of years, then buy another one. It's kind of like the equivalent of a calculator. Both AMD and Nvidia cards will do the job. I'm not sure what the big deal is...


-Consecutive threads merged. In the future, please edit previous post rather than creating a new one within a few minutes. (sswilson)
__________________



HTPC: Z77A-G45; 3770K; Coolermaster GeminII; 2x4GB Kingston HyperX 1600Mhz; MSI R7-260X; 2x 128GB Crucial M4 SSD; 1TB WD Green, 2x 2TB WD Green; PC P&C 750W PS; Fractal Design Node 605;
Son's Rig: M5A97; 1055T; CNPS20LQ; 2x4GB Corsair Vengeance 1600Mhz; ASUS GTX650Ti Boost; 80GB Intel 520 SSD; 320GB WD Black HDD; SPI 700W; Bitfenix Shinobi;

Last edited by sswilson; August 17, 2012 at 09:20 AM.
Reply With Quote
  #80 (permalink)  
Old August 17, 2012, 09:25 AM
mattlef's Avatar
Hall Of Fame
 
Join Date: Apr 2008
Location: Toronto, Ontario
Posts: 1,976

My System Specs

Default

Okay, guys lets not feed the troll anymore.
He's just trying to be a keyboard warrior and stirr up stuff where there is none. We all know what to beleive and who to put our faith and trust in for accuracy.
__________________
Co-Host of the Dam Good Show

My Rig: Motivation 2.0
----------------
Build log - Motivation: Sponsored by Galaxy and Corsair
---------------
Check out the HardwareCanucks Youtube Channel for News/Reviews and more!
Instagram Twitter
Reply With Quote
Reply


Thread Tools
Display Modes

Similar Threads
Thread Thread Starter Forum Replies Last Post
NVIDIA GTX 680 2GB Review Comment Thread SKYMTL Video Cards 148 April 3, 2012 06:56 PM
Comment Thread for AMD HD 6990 Review SKYMTL Video Cards 73 March 22, 2011 07:21 PM
BFG GTX 285 OCX Review Comment Thread SKYMTL Video Cards 8 January 21, 2009 02:55 PM
BFG GTX 260 Review Comment Thread SKYMTL Video Cards 33 November 15, 2008 11:26 PM
BFG GTX 280 OCX Review Comment Thread SKYMTL Video Cards 15 August 19, 2008 09:07 PM