Radeon HD 5770, Radeon HD 4890, And GeForce GTX 275 Overclocked

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.

cinergy

Distinguished
May 22, 2009
251
0
18,780
[citation][nom]siliconchampion[/nom]Whoa, people, people!Perhaps in stead of flaming the author out about his choice of Nvidia favoring titles, perhaps you could make some helpful suggestions of game titles you would like to see benchmarked...Personally, I would love to see some CoD4 and MW2 benchies, but that's just me.[/citation]

Flaming author because this ISN'T the first time and it's not just about selecting game titles. I'm sure he's very well aware of the situation but just doesn't give a damn.

And ranking cards with cherry picked game titles is very, very bad idea. Radeon HD4890 losing GTX260? You can call that a fair estimate? Leaving Last Remnant out alone would raise ATI cards positioning many notches.
 

1898

Distinguished
Oct 13, 2009
249
0
18,690
It's beyond me why almost all top-notch nVidia cards have been included while leaving out the 5000 series except for the cheapest one.
 

Jenoin

Distinguished
Sep 29, 2009
37
0
18,530
Haha, I wasn't sure about all the people saying that this was totally biased (except for the The Last Remnant that benchmark is totally retarded) but then a simple google search alerted me to the fact that the GTX 275 you were using is actually as expensive as an HD 5850. You should have definitely pointed this out to your readers. The 5850 will thrash the GTX 275 in every way. (except possibly in The Last Remnant)
 

shubham1401

Distinguished



GTX295 isn't that good.
Firstly they haven't put the benchies for ATI 5XXX cards.
Secondly GTX295 doesn't play many games at 2560X1600 resolution while 5XXX series can do so.

Power consumption of GTX295 is fairly high while newer ati cards are so energy efficient.
It doesn't even justifies its 500$ price.

 
G

Guest

Guest
yeah guys,

take it easy. i find it far more believable that it was an oversight rather than an intentional slighting that led to showing games that favored nvidia over amd.

its okay to be right and be nice about it at the same time.
 

techrep

Distinguished
Nov 24, 2009
1
0
18,510
1.)Check this and the price vs. the other cards.
http://www.newegg.com/Product/Product.aspx?Item=N82E16814127453

2.)Check this card and the price vs. the cards you tested. I own and have OC'ed it past your MSI part while it stays QUIET!. Ordered another for CF!!
http://www.newegg.com/Product/Product.aspx?Item=N82E16814161299

3.)Check this OC review of the 4890 for another perspective.
http://www.anandtech.com/video/showdoc.aspx?i=3555

4.)Why doesn't anyone test using monitors with a 1920 x 1080 resolution as they are widely adopted now rather than 1680 x 1050??? OK, keep the 1680 x 1050 but at least add 1920 x 1080.

5.)The pricing of each card is not being taken seriously in this review, sadly, as it is the bottom line for most of us regarding bang for the buck and the point of many quarterly articles here. this article is a contradiction to that philosophy. IMHO...
 

geok1ng

Distinguished
Jun 25, 2008
111
0
18,690
it is not Tom's fault if most games are NVIDIA favorable. "The way it was meant to be played" is an initiative of NVIDIA to offer help to game developers, so far without a similarly effective response from ATI. Ever since AMD bought ATI and fired personnel ATI has been unable to gain the upper hand in the drivers and game developing front. I speak as the happy owner of a 4870X2.
 

jennyh

Splendid
I have never seen such a PATHETIC excuse for a review as this one. Some of this stuff is disgraceful.

When did the gtx295 EVER beat the 4870x2 in Farcry2? Check out Anands bench for the real story :-

http://www.anandtech.com/video/showdoc.aspx?i=3679&p=9

No 5850, 5870, 5970? Is that because you wouldn't have a green bar at the top in every one?

This reviewer is bought and paid for and should never be allowed to review anything again.

Total FARCE, you should be bloody well ashamed of yourself. How much did you get paid?
 
Hopefully the author will take some of this into consideration.
We have EOLd parts representing ATI for highest end, while nVidias part costs as much as the mew ATI part, which goes unmentioned.
Mentioning this several times would put a better context on perf/price in your article
 
Looks like the 5770 is generally a bit behind the 4870 at stock speeds and just edges it out when over clocked. Of course, even overclocked it's often a bit behind the ol 4890 even at the latter's stock clocks. Still, the 5770 makes sense for ATI since it's cheaper to produce, at least when yields are decent, and can thus commands a higher margin than the GTX275 or the 4890.
 

liquidsnake718

Distinguished
Jul 8, 2009
1,379
0
19,310
[citation][nom]siliconchampion[/nom]Whoa, people, people!Perhaps in stead of flaming the author out about his choice of Nvidia favoring titles, perhaps you could make some helpful suggestions of game titles you would like to see benchmarked...Personally, I would love to see some CoD4 and MW2 benchies, but that's just me.[/citation]

Well you have to put into consideration that TOms did not put in the newer 5850's, 5870's and the new king, 5890. This isnt really biased as Nvidia did hold on its own and it clearly showed during these benches.... Ive actually seen some of these benchmarks on older tests and I think they just wanted to test budget "high" end GPU's that are shown in the BLACK bars and comparing it to the older tests in which the 5870's and 5890's werent a part of...
 
G

Guest

Guest
Where are the 5850 and 5870 ATI cards?? Why your ATI 5770 perform so bad? You have Nvidia favoritism?? I think there's no fair play in these tests...
 

grimjester

Distinguished
Sep 24, 2009
24
0
18,510
In our second chart, we use normalized results for each benchmark, converting the scores into percentages and then calculating the final result. This ensures that individual benchmarks in which a single card may score much higher than the remaining models do not skew the results. Effectively, all benchmarks count the same.

If you set the slowest to 100%, the last sentence is not true. Quite often the slowest card has serious problems with lack of memory or driver support for a game. For example, the highest settings on Far Cry 2 choke the 4770 completely, giving the average card a score of 700% while the 1920x1200 without AA/AF gives the average a more reasonable 150%. This makes the cases where a game is completely unplayable at the tested settings outweigh five cases where it's working fine.

Normalizing on the sum or average of all cards would ensure that every benchmark contributes equally to the overall score, but that would also mean the ranking of the fastest cards is sensitive to how they do on tests where some random inconsequential cards fail completely. E.g. the comparison between 295 and 4870x2 would be mostly determined by cases where the 4830 chokes.

Normalizing on the fastest card would be far better, imo.
 

airgreek

Distinguished
Nov 8, 2009
22
0
18,510
Another great example of how people always think there is a hidden agenda. Thank you TH crew. Your work is the best on the net
 

MasterDOOM

Distinguished
May 30, 2008
25
0
18,530
But Remmber guys the ATI Radeon 5970 is the Best Videocard in the World in this moment ! Nice Job Tomshardware and AMD - ATI Radeon :)
 

tacoslave

Distinguished
Apr 3, 2009
704
0
18,980
looking at these results it seems the 4870x2 is the most powerful card (dx10) you can buy because of the last remnant results among others weight the results in nvidias favor
 
G

Guest

Guest
May is say i think it's silly to waste 25watt on average for a maegre 5% oncrease in FPS. All this tells me that a OC GTX275 isn't worth the extra cash you throw at it.
 

vianescute

Distinguished
Oct 14, 2009
4
0
18,510
like everyone said before, why dont test the other higher performance card like 5970? coz people wants to see the all new high card to be compared for next pc building plan.
 
Status
Not open for further replies.