GTX480 arrived [evga forums]

[MrK6]

Yeah I love the belittling attitude but I've done the research myself. The BBC2 benchmark is not being reproduced like that across the majority of benchmarks. I can point to more instances where the gtx480 is faster than the hd5870.

gtx480 is faster

http://www.pcgameshardware.com/aid,...-Fermi-performance-benchmarks/Reviews/?page=7
http://www.hexus.net/content/item.php?item=24000&page=9
http://news.firingsquad.com/hardware/nvidia_geforce_gtx_480_470_performance/page14.asp
http://www.hardwarecanucks.com/foru...s/30297-nvidia-geforce-gtx-480-review-14.html
http://benchmarkreviews.com/index.p...sk=view&id=480&Itemid=72&limit=1&limitstart=6
http://benchmarkreviews.com/index.p...sk=view&id=480&Itemid=72&limit=1&limitstart=6
http://www.guru3d.com/article/geforce-gtx-470-480-review/16


Here's a few where the gtx480 is slower or faster, depending on settings:

http://www.brightsideofnews.com/new...rce-gtx-480--gtx-480-sli-review.aspx?pageid=4
http://www.tomshardware.com/reviews/geforce-gtx-480,2585-13.html

Here's two where it's slower

http://www.techreport.com/articles.x/18682/12
http://www.techspot.com/review/263-nvidia-geforce-gtx-480/page7.html

See the general trend? The gtx480 is usually faster in that game. What I wanted was for you to show me where a game consistently scores 10% slower with a gtx480 vs. an hd5870 but it appears you can't.

No, what you want to do is prove me wrong because I pointed out your failure in understanding graphics cards and simple math. However, what you have done here is just continued to prove my point that you have no idea how graphics cards work. No game is going to show consistent "percentage increase" across the board. The two graphics cards have very different architectural designs, and one design will excel where another fails and vice versa just due to the complexity and nature of the specific scene being rendered. The simple fact, as I've already stated and provided sources, is that the GTX480 is 10-15% faster across the board. You can have extremes in the range of -20% to +70%, but the average shows the trend.

My calculations took minima into account, and I also ended up with 15% ish.

Btw, for what it's worth MrK6, Bad Company 2 was also faster with a GTX 480 in my bench. Which is easily reproducable, since it's the first snow-level. Cold War I think it's called. It's after the cutscene, and starts when you have to use a 40mm-grenade on a armored truck. It's quite an intensive part with lots of explosions, soft particles going everywhere, which is were the cards really struggle.

Thanks for replying and I have no reason to doubt your testing. The GTX480 might be faster than the 5870 in other scenes/areas of the game. I provided examples of what I pointed out in regards to understanding averages that met tviceman's arbitrary and ridiculous standards and as you can see he's still sulking, so I'm done with that argument.

Yeah, it is like somebody buying a Ferrari and worrying about the gas mileage.

If that Ferrari howled like crazy due to highway noise every time you went over 60MPH, people still wouldn't buy it

 

[Genx87]

If that Ferrari howled like crazy due to highway noise every time you went over 60MPH, people still wouldn't buy it

Right, because sports cars are known for being so quiet.

 

[Madcatatlas]

Are you comparing high end GPU`s to "high end" car brands?

I do hope your not saying Fermi products are the Ferraries of the GPU world...

 

[Genx87]

Are you comparing high end GPU`s to "high end" car brands?

I do hope your not saying Fermi products are the Ferraries of the GPU world...

Why yes, yes we were. Go back and read why.

 

[Madcatatlas]

No thank you It just fails, no matter how you put it. They are that different in application.

 

[MrK6]

Right, because sports cars are known for being so quiet.

Sounds like you've never ridden in a Ferrari .

No thank you It just fails, no matter how you put it. They are that different in application.

That's what I was trying to point out, even using the same analogy. I like your direct approach even better .

 

[MrK6]

For all parties interested, this summary was posted over at GPUReview:

http://www.gpureview.com/gpureviews-gtx-480-and-gtx-470-overclocking-roundup-article-834.html

Personally, I always think review sites tend to "wimp out" when it comes to graphics cards overclocking, and I have a funny feeling the "true" averages might be a little higher. I'd still like to see more results on the GTX470. If it can be shown to consistently overclock to the 750MHz range (with voltage or not) and have reasonable heat/noise characteristics, it'd be a good buy at $350.

 

[tviceman]

No, what you want to do is prove me wrong because I pointed out your failure in understanding graphics cards and simple math. However, what you have done here is just continued to prove my point that you have no idea how graphics cards work. No game is going to show consistent "percentage increase" across the board. The two graphics cards have very different architectural designs, and one design will excel where another fails and vice versa just due to the complexity and nature of the specific scene being rendered. The simple fact, as I've already stated and provided sources, is that the GTX480 is 10-15% faster across the board. You can have extremes in the range of -20% to +70%, but the average shows the trend.

I asked you to show me a game that is consistently 10% or slower on a gtx480 vs. an hd5870 across the board and you cannot. I can, however, show you plenty of games that consistently score 20% or better on a gtx480 vs. an hd5870 across the board which is exactly the same principle of what I am asking you to do. But since an hd5870 just isn't fast enough to make your argument true, you just say I don't understand graphics and simple math. Ok.

 

[ZimZum]

If that Ferrari howled like crazy due to highway noise every time you went over 60MPH, people still wouldn't buy it

When the Dodge Viper first came out. It had horribly uncomfortable driving position, no AC, no radio, an interior door latch that consisted of a cable, no side windows, no roof, no door handles, etc etc. It was fast as hell, but everything else about it sucked. Thats probably a better analogy than a Ferrari.

 

[Genx87]

No thank you It just fails, no matter how you put it. They are that different in application.

Ha how would you know unless you go back an entire page and read why the comparison was made?

/facepalm

 

[Genx87]

Sounds like you've never ridden in a Ferrari .

I have not however I have been in the following "sports" cars.

Corvette(76,94,95,03)
Dodge Viper
Mustang

If you think these things are quiet you are on crack.

And for what it is worth a buddy of mine has an Audi RS6 with the turbo V10 that tops out about 500 HP. That car while a luxury sedan isnt exactly quiet either.

My analogy is fine. People worrying about 5 dollars a month in electrical bills for a high end gpu would be like somebody wanting to buy a ferrari but worry about the poor gas mileage.

It is just silly when you think about it to worry about 5 bucks a month on a 500 dollar GPU.
Just like it would be silly to worry about paying more for gas with a Ferrari.

 

[badb0y]

I have not however I have been in the following "sports" cars.

Corvette(76,94,95,03)
Dodge Viper
Mustang

If you think these things are quiet you are on crack.

My analogy is fine. People worrying about 5 dollars a month in electrical bills for a high end gpu would be like somebody wanting to buy a ferrari but worry about the poor gas mileage.

It is just silly when you think about it to worry about 5 bucks a month on a 500 dollar GPU.
Just like it would be silly to worry about paying more for gas with a Ferrari.

For me personally, higher power consumption means more heat in my case, I could care less about the monthly bill.

 

[OILFIELDTRASH]

Anyone worried about an extra $5.00 to their monthly power bill shouldn't be spending $500.00+ on a video card to begin with.

So if I spent $500 on a videocard I should just continue to waste more money on my electric bill? Why should I have to settle for that when I'm paying $500 for a graphics card?

Thankfully my gtx280 still chews up console ports and spits them out!

 

[Madcatatlas]

If you think these things are quiet you are on crack.

And for what it is worth a buddy of mine has an Audi RS6 with the turbo V10 that tops out about 500 HP. That car while a luxury sedan isnt exactly quiet either.

great example, the audi rs6 is a monster when it comes to sound, infact it is illegal to buy one if you live in a rural area in Norway, they compare it to the sound jets make and so you have to live out in on the countryside to buy one.

I could facepalm back, but ..lol

edit: http://www.youtube.com/watch?v=RngrdmFdKW8
loud

 

[MrK6]

I asked you to show me a game that is consistently 10% or slower on a gtx480 vs. an hd5870 across the board and you cannot. I can, however, show you plenty of games that consistently score 20% or better on a gtx480 vs. an hd5870 across the board which is exactly the same principle of what I am asking you to do. But since an hd5870 just isn't fast enough to make your argument true, you just say I don't understand graphics and simple math. Ok.

No, you created scenarios of increasing specificity and complexity to escape being called "wrong." Since you like to rewrite history just as well change the variables of an argument to hide your ignorance, let's review where this started, you stated:

It gets 30% as often as it gets 8%. So, in my post to the original person, I augmented his comment to say it gets anywhere from 8-30%.

Which I corrected and said no, it gets 10-15% on average, and there will be -10%'s to balance out those 30%. You conveniently ignored the facts that you were A) wrong about the cards performance (it's "30% as fast just as often as it's 8% as fast") and B) have no idea how averages work, instead changing the focus on some arbitrarily defined scenarios and requirements for me to prove my point. I provided the links, including a summary of averages (showing the 10-15% figures I stated) as well as specific examples of the GTX480 falling behind 5870 by 10% or more. Since then you have tried to change the requirements of "proof" further to avoid being "wrong." This is childish, enjoy your pity party.

I have not however I have been in the following "sports" cars.

Corvette(76,94,95,03)
Dodge Viper
Mustang

If you think these things are quiet you are on crack.

Then you should have said one of those, not a Ferrari. There's a reason a Ferrari is in a class by itself. I own a '69 Camaro that I love and through my mods is loud as hell, but I'm still not comparing it to a Ferrari.

Anyway, I agree with the argument. I already stated that with power being so cheap in the US, it isn't a valid concern for the card. But as others have said, the heat and noise that come from the higher power consumption are more of a problem.

 

[Genx87]

No, you created scenarios of increasing specificity and complexity to escape being called "wrong." Since you like to rewrite history just as well change the variables of an argument to hide your ignorance, let's review where this started, you stated: Which I corrected and said no, it gets 10-15% on average, and there will be -10%'s to balance out those 30%. You conveniently ignored the facts that you were A) wrong about the cards performance (it's "30% as fast just as often as it's 8% as fast") and B) have no idea how averages work, instead changing the focus on some arbitrarily defined scenarios and requirements for me to prove my point. I provided the links, including a summary of averages (showing the 10-15% figures I stated) as well as specific examples of the GTX480 falling behind 5870 by 10% or more. Since then you have tried to change the requirements of "proof" further to avoid being "wrong." This is childish, enjoy your pity party.


Then you should have said one of those, not a Ferrari. There's a reason a Ferrari is in a class by itself. I own a '69 Camaro that I love and through my mods is loud as hell, but I'm still not comparing it to a Ferrari.

Anyway, I agree with the argument. I already stated that with power being so cheap in the US, it isn't a valid concern for the card. But as others have said, the heat and noise that come from the higher power consumption are more of a problem.

The car was a metaphor for all sports cars.

 

[Genx87]

great example, the audi rs6 is a monster when it comes to sound, infact it is illegal to buy one if you live in a rural area in Norway, they compare it to the sound jets make and so you have to live out in on the countryside to buy one.

I could facepalm back, but ..lol

edit: http://www.youtube.com/watch?v=RngrdmFdKW8
loud

I find it amusing you had time to search on youtube for that video but couldnt be bothered to go back a whole page in this thread to understand why the example was used.

 

[Keysplayr]

Anyway, I agree with the argument. I already stated that with power being so cheap in the US, it isn't a valid concern for the card. But as others have said, the heat and noise that come from the higher power consumption are more of a problem.

More of a problem? What would your problem be? It's not as if we are talking about the difference between a BBQ and a thermal nuclear reactor here. It's a hotter running, louder fan noise, higher performance graphics card. But dude, it's not as if the competition doesn't make any noise and doesn't throw out any heat. They all do. Some more than others. It has always been this way. Always.

 

[tviceman]

No, you created scenarios of increasing specificity and complexity to escape being called "wrong." Since you like to rewrite history just as well change the variables of an argument to hide your ignorance, let's review where this started, you stated: Which I corrected and said no, it gets 10-15% on average, and there will be -10%'s to balance out those 30%. You conveniently ignored the facts that you were A) wrong about the cards performance (it's "30% as fast just as often as it's 8% as fast") and B) have no idea how averages work, instead changing the focus on some arbitrarily defined scenarios and requirements for me to prove my point. I provided the links, including a summary of averages (showing the 10-15% figures I stated) as well as specific examples of the GTX480 falling behind 5870 by 10% or more. Since then you have tried to change the requirements of "proof" further to avoid being "wrong." This is childish, enjoy your pity party.

I did say you were right about the 30%, 8% situaiton. Or did you forget to read that?

I agree with the 10-15%. I never disagreed with that. And I do stand corrected, it doesn't get 30% as often as it is 8% or lower. It is in the 20-25% or better as often as it is 8% or lower, but not 30%.

HOWEVER, you did not back up your claim that I called you out on, and that was to link several reviews with the same game reproducing gtx480 being 10% (or more) slower than a 5870 on that game.

The fact is you claim there are games where the gtx480 is 10% slower. To me, that means more often than not a game is going to be slower on the gtx480 whether it's CPU bound or GPU bound. Yet you can't show me consistent results of a single game being 10% slower on a gtx480 than an hd5870. There is nothing arbitrary when I say show me a game that is consistently scoring 10% faster on an hd5870 than a fermi across the majority of reviews.

That's pretty loose. I didn't say what game, I didn't say what resolution, I didn't say what AA settings, I didn't say what version of directx....

But it's cool. I mean, you know, since I never retracted my 30% - 8% performance gap and called you out when you were OBVIOUSLY right and provided ample proof and pinpoint explanations so yeah I must be childish.

 

[OILFIELDTRASH]

More of a problem? What would your problem be? It's not as if we are talking about the difference between a BBQ and a thermal nuclear reactor here. It's a hotter running, louder fan noise, higher performance graphics card. But dude, it's not as if the competition doesn't make any noise and doesn't throw out any heat. They all do. Some more than others. It has always been this way. Always.

I refuse to accept the mindset of well it's a hot card and it's just part of it. Nvidia has a dud on their hands and I'm not okay with paying higher electric bills due to increased power consumption and heat generated by this thing. You'll probably say well it's $5 extra a month. Well sorry that $5 could go towards a saving account or something. Fermi is not okay and nvidia needs to come out with a fast card that won't heat up my house (within reason) in the Texas summer time.

 

[1ManArmY]

So if I spent $500 on a videocard I should just continue to waste more money on my electric bill? Why should I have to settle for that when I'm paying $500 for a graphics card?

Thankfully my gtx280 still chews up console ports and spits them out!

Nice!

 

[Madcatatlas]

Car analogies are fail as I said. I have read this entire topic from the start, i just didnt need to "go back and find out why you" (plural) starting using them.
Let me put it into words for you, instead of leaving it up in the air so you can grab it (yes i do think you need it clarified this much)

Audi RS6 is not loud in the world of cars, naturally you could break it up even more by saying "world of sports cars" or "world of <500 HP cars", (I just wrote that so you wouldnt pick on it..)

Cars..are not GPU`s. Cars are not pocket change attainable objects (noone carries several hundred thousand grand in pocket change, i do mean hard cash and not credit.. see i had to write this so you wouldnt PICK on it..:hmm


Also, since it seems so freaking impossible to fathom (for some), there are nVIDIA fans out there who are PISSED off at the current nVidia offerings. There are netcafe owners out there (you guessed it, we are talking the serious ones) who are PISSED off at nVidias current offerings.

They are nVidia fans and want nVidia products, but LOOK at the negatives. Its like a roadblock. So instead of just writing stuff off, pointing fingers saying "ati/amd fanboy", how about you actually get the point that is being made: showing displeasure at the offerings a company you are a fan of (their products in this case) hoping they get the message and fix it/make better products in the future.

And lets not even start with people who have small serverparks (think smaller companies here) who want the compute power of Fermi and have to rethink or compensate for the negatives of the offered product. And lets by all means NOT forget they were promised this product much earlier and then promised the same product a month later, and a month later..

yes, i do feel cheated.


oo ati fanboy

 

[tviceman]

This is sure to get the Nvidia haters even more riled up.

Alienbabeltech took the time to find the max overclocks of both their gtx480 card and hd5870 card and is finishing up their article now. As I said mentioned earlier, http://forums.anandtech.com/showpost.php?p=29662246&postcount=107 , the gtx480 may in fact scale better with overclocking than an hd5870.

http://alienbabeltech.com/abt/viewtopic.php?p=36348#p36348

 

[MrK6]

The car was a metaphor for all sports cars.

Well, poor analogy nevertheless, but I get your point.

More of a problem? What would your problem be? It's not as if we are talking about the difference between a BBQ and a thermal nuclear reactor here. It's a hotter running, louder fan noise, higher performance graphics card. But dude, it's not as if the competition doesn't make any noise and doesn't throw out any heat. They all do. Some more than others. It has always been this way. Always.

Nice spin attempt. Heat form computing equipment is always a problem. And when it's excessive, as in the case of the GTX480, it's a big problem. That's what half of this thread has been about. The GTX 480 is using 40-50% more power for only 10-15% more performance.

I did say you were right about the 30%, 8% situaiton. Or did you forget to read that?

There's a failure at reading comprehension. I said you conveniently ignored it and started changing variables and requesting benchmarks. It wasn't until I posted that chart that you finally admitted it.

The fact is you claim there are games where the gtx480 is 10% slower.

Where? Show me once where I ever said that. Go and quote it, right now. Once you've realized you're making things up, you'll hopefully stop posting.

 

[Skurge]

Tviceman, look at the stalker benches, the 5870 is consistantly 10&#37; faster than the 480.