Best Graphics Cards For The Money: Jan. '09

[anonymous x]

[citation][nom]anonymous x[/nom]Wow, once again, no 9800 GTX+, even though it runs faster than the ati 4850 in the majority of benchmarks, it runs cooler (reverence cooler is dual slot, and mine at stock fan speed idles @ 47°C and loads at 58-63°C depending on game), and most review sites get it to overclock higher.(idle temp, my card, stock everything (note that it's factory overclocked))(9800 GTX+ wins 6 benchmarks, 4850 wins 3)(9800 GTX+ wins 6 benchmarks, 4850 wins 3)quoterice????here's one for $144.99 ((after rebate), without rebate it's $164.99)AND it includes Call of Duty World at WarNow once again, why isn't the 9800 GTX+ on your best graphics cards for the money article? It should at least tie the 4850.[/citation]

Oh great the links don't work, there they are:
http://farm4.static.flickr.com/3331/3178641562_72bdc67111_o.png
http://www.guru3d.com/article/geforce-9800-gtx-512mb-plus-review/1
http://www.legitreviews.com/article/732/1/
http://www.newegg.com/Product/Product.aspx?Item=N82E16814130416

 

[pcgamer12]

Shrank is actually a real adjective. It means parched or shrivelled and is used to describe seeds. Looked it up in the online Oxford English Dictionary. (Check your local library's database to gain access to it, you'll need a library card number.)

 

[kelfen]

[citation][nom]smirnovovich[/nom]I believe this to be the first time Tom's acknowledge the existence of the 4870 1GB end even now it is being acknowledged as "It turns out that the Radeon 4870, when equipped with a whole gigabyte of GDDR5 RAM, can up the ante to compete against the GeForce GTX 260+ in some titles."I think it would be better if you would turn this around because GTX 260+ was only launched to match the 4870 1GB.(btw, i am no fanboy cause i am gonna sell my 4870 to get a 260+ because it is slightly better )[/citation]
it is really worth the hassle

 

[c4rl]

Hello! (new here) Can i ask if the 4650 in the hierarchy list is ddr2 or ddr3 version? Just wanna be sure... Thanks in advance!

 

[pauldsr]

I Ii thi[ufasdfasfasgadsgagasgasgasg][/u]
I thinkg

afsdfasf

sdfasdfasfsdfsdfsf







































































































































































i think the asus 3870 X2 should be included herer

 

[pauldsr]

I think the asus 3870X2 should be an included contender, I am using them in twin cards Quad crossfire with 4 monitor output between the 2 cards though i am only using twin 1080p monitors in my setup, they are however getting hard to find as the 4870X2 has replaced it.
souls

 

[smirnovovich]

[citation][nom]kelfen[/nom]it is really worth the hassle[/citation]

what do you mean?

 

[cleeve]

[citation][nom]anonymous x[/nom]Wow, once again, no 9800 GTX+, even though it runs faster than the ati 4850 in the majority of benchmarks[/citation]

Dude, the cards perform right on par.

But notice the title of the article: 'best cards for the money'... the 4850 is $20 cheaper.

End of story. If the 9800 GTX+ price lowers, it'll be on the list. At $20 more, it ain't worth it.

 

[douken]

Whoever did the hierarchy chart deserves a prize

 

[hopkiller]

Well you can't compare the GTX 295 to any of these it will have to be compare to the the 5870

 

[bfstev]

[citation][nom]Cleeve[/nom]I'll address some of your comments here:- This article will never have benchmarks for a direct performance comparison. Why? Because this is an editorial article based on my knowledge and experience, it's purpose is to provide quick refrence for people who don't want to look at mountains of data. To add a ton of benchmarks would be in direct conflict with the purpose and spirit of this article; if you're not satisfied with this general overview or are curious about how a specific card performs, google a review of that particular card.[/citation]

I would need a full review or detailed benchmarks or anything, just maybe a little chart that would give a roundabout of performance at a given resolution. and i mean a broad view here like "oh this card can perform on average between 30 and 50 fps in most games running at 1920x1200. Anything less than that would just be

 

[bfstev]

Anything less than that would just be

 

[Guest]

"- This article will never have benchmarks for a direct performance comparison. Why? Because this is an editorial article based on my knowledge and experience, it's purpose is to provide quick reference for people who don't want to look at mountains of data."

And therein lies the major problem today with Tom's.

There was a time when the vast majority of the content was hard tech.

Now, there's way too much boilerplate.
So I rarely visit the site anymore.

Frankly, *who* are you?

I don't recognize you as any kind of expert.

Probably because I rarely visit the site anymore.

So, if you want me to take you seriously, you need a whole lot more meat in your articles.

Sorry if that sounds harsh - but it is what it is.

Then again, you probably don't give a rip if I visit the site.

Clearly, your core demographic has moved from the hard core hobbyist to the casual user who is more inclined to click all the ads.

 

[bfstev]

Anything less than that would just be less than 30 or n/p(not playable). You already do it with the top saying "play most game at whatever resolution with decent quality", so why not make it a quick glance chart with just the basic road-about fps to resolution comparison.

Anyway just a suggestion that i think would definently make the purchase decision easier, at least for me anyway.

this kept breaking for me when i used the symbol for "less than", sorry

 

[cleeve]

[citation][nom]xcfdsacdsacdsa[/nom]And therein lies the major problem today with Tom's.[/citation]

Honestly, I don't see it as a problem. The bottom line is, this particular article isn't for you, as you crave benchmarks. That's fine, we have a lot of articles like that. They're important; I spend a lot of time writing them. That's what gives me the experience I need to write an article like this one, actually.

I do think it's kind of ridiculous to frame your personal dislike of an editorial-style article as a sweeping conviction of Tom's Hardware in general. Frankly, I see your post as just another guy jumping on the 'Tom's Hardware is sooo 1999' bandwagon.

The fact that this monthly editorial gets a hell of a lot of hits demonstrates that there's a public desire for it - perhaps even a need. What's your attitude, that people who aren't willing or able to spend the time to do the research don't deserve a monthly breakdown? A little elitist there, aren't we?

[citation][nom]xcfdsacdsacdsa[/nom]Frankly, *who* are you?[/citation]

I'm the guy who wrote the article you read.

Who does that make you? The guy who complained about it as well as an entire website without mentioning any specific disagreements with my conclusions? The guy who decided that editorial articles have absolutely no merit at all for everybody because he doesn't personally like them?

If you have a problem with my conclusions or facts, go ahead and engage me, I'll be happy to discuss my motives for making the recommendations I have.

But if you want to whine and moan about how THG sucks because you don't personally find value in an article of this kind... then yeah, you're right! I probably don't give a rip if you visit the site.

 

[Guest]

The AGP version of the Radeon HD 2600 XT doesn't seem to have official drivers from ATI available. Is this just not a problem with the "hotfix" drivers or vendor supplied ones since the article mentions driver issues being resolved?

 

[anonymous x]

[citation][nom]Cleeve[/nom]Dude, the cards perform right on par. But notice the title of the article: 'best cards for the money'... the 4850 is $20 cheaper. End of story. If the 9800 GTX+ price lowers, it'll be on the list. At $20 more, it ain't worth it.[/citation]
"right on par"
I repeat, most reviews show the GTX+ winning on most benchmarks. I go by what I see with my own eyes. If Tom's hardware somehow got identical performance scores with the two cards, so be it. When I read reviews, I counted which card won more benchmarks. The 9800 GTX+ card nearly always won more.
"it ain't worth it"
better performance, and lower temperatures (which means more OC headroom) is worth $20.
Sure the difference between the two cards is not worth $100, or even $50, but a mere $20? (which, btw, is enough to make the "Best PCIe Card For $150" price point you set for the 4850 512 MB).
If, for some reason, you failed to see the $144.99 card (free shipping) I showed you that also came with a $45 game,
http://www.newegg.com/Product/Product.aspx?Item=N82E16814130416
why wouldn't it make (or tie) the $170 price point, where there is a 4850 1GB card. Most people who buy sub $200 cards aren't looking to game on giant 24" LCDs were the extra video ram will be useful, and by the time 1 GB of video ram is recommended for lower resolutions, the cards will be outdated anyway.

 

[werepossum]

[citation][nom]Cleeve[/nom]Honestly, I don't see it as a problem. The bottom line is, this particular article isn't for you, as you crave benchmarks. That's fine, we have a lot of articles like that. They're important; I spend a lot of time writing them. That's what gives me the experience I need to write an article like this one, actually.I do think it's kind of ridiculous to frame your personal dislike of an editorial-style article as a sweeping conviction of Tom's Hardware in general. Frankly, I see your post as just another guy jumping on the 'Tom's Hardware is sooo 1999' bandwagon.The fact that this monthly editorial gets a hell of a lot of hits demonstrates that there's a public desire for it - perhaps even a need. What's your attitude, that people who aren't willing or able to spend the time to do the research don't deserve a monthly breakdown?
SNIP
[/citation]
I recommend these guides quite often to people looking for a new video card. So many people either don't have the time or inclination to go through all the benchmark articles or, frankly, get confused. I use them myself to keep abreast of general levels of performance, knowing that if I recommend one of your recommended cards I'll be reasonably safe unless the person only wants to play one or a few related games and has really bad luck. Even then, the speed between one recommended card and another in a particular game engine is seldom really significant or noticeable in the real world. Bottom line, read the benchmark articles if you enjoy them, but if you only use the recommendations you'll be pretty safe.

My only request would be a little more info on each class relative to the classes above and below. For instance, when the 8800GT 512MB first came out there as little reason to buy anything else. (Although to be fair I think that was mentioned at the time.)

Thanks for doing these articles. I find them enjoyable and quite useful.

 

[bestliu]

why don't you think about HD3850? Antique but more efficient with 256bit than HD4670~

 

[TheGreatGrapeApe]

I would need a full review or detailed benchmarks or anything, just maybe a little chart that would give a roundabout of performance at a given resolution...

Based on what criteria? A single resolution or small sample of games shows only a slice of performance, and for any chosen title there are others that would negate those.

This has always been a quick look based on pooled information from Cleeve's personal experience and collected information from other sources.

You want digits, read the GPU charts, you want well supported opinion on the subject then this is a good source.

 

[TheGreatGrapeApe]

Frankly, *who* are you?

And who are you?

So, if you want me to take you seriously, you need a whole lot more meat in your articles.

I don't think anyone's looking for anything from you. You don't like it, don't read them. You want more 'meat', read the GPU charts.

Then again, you probably don't give a rip if I visit the site.

Yep, and neither do the hardcore hobbyist who never had much tolerance for whinning.

This article never pretended to be an architectural investigation or a collection of a bazillion benchies, it's maintained what it always was, a good look at the state of the market offering. Changing now would be like taking something that works and changing it for those who never were involved enough in the site to understand it's origin to begin with, so really, just how hardcore are you if you're that out of it?

 

[JonnyDough]

[citation][nom]pcgamer12[/nom]Shrank is actually a real adjective. It means parched or shrivelled and is used to describe seeds. Looked it up in the online Oxford English Dictionary. (Check your local library's database to gain access to it, you'll need a library card number.)[/citation]

Those silly English blokes don't know anything about my language. You act as if it originated in Europe or something.

 

[cleeve]

[citation][nom]anonymous x[/nom]Sure the difference between the two cards is not worth $100, or even $50, but a mere $20?[/citation]

I don't think it is based on what I've seen; you do, and you're certainly entitled to that opinion.

From the benches I've seen and my experience with the cards, they trade blows. To me, it's more than the number of benchmarks each cards wins, but the margins they win with; and they're incredibly slim. Also, if the 4850 wins fewer benches but by a bigger margin, that factors into my conclusion.

I still am of the opinion that if you want to spend $165, you'd be better off to spend an extra $5 for a 1GB 4850...

So, one more time, not worth the extra $20 for the 9800+ IMHO. In yours it is, and you're welcome to buy a 9800 GTX+.

P.S. I've said it a number of times now, we don't consider mail-in rebates. And I'm sure as hell not going to recommend a card over another one because it comes with a game when my criteria is price/performance.

 

[fledgling101]

There was a time when the vast majority of the content was hard tech.

Although I'm sure the vast majority of people here don't like to hear it. This site is not nearly as technical as it once was. Personally I can't say that this is entirely a bad thing as I enjoy the current articles however whenever I feel the need for stats or more data (which is fairly frequent), I'm forced to begin googling for reviews of questionable origin and bias.

I'm not saying that Tomshardware should trash this method and get go back to the former more technical form (don't get me wrong though, I still really miss those old articles even if I did skip past the information to the conclusion half of the time). However I am saying that Tomshardware is no longer the same as it once was, and this has caused me to lose a fair amount of confidence in regard to the quality of the articles (this is not referring to this particalur article but rather Tomshardware as a source of data and opinions clearly affected by benchmarks and tests, as I mentioned earlier, I personally appreciate this GPU article). While I agree that this article has no reason to become inundated with benchmarks, graphs and the such that used to commonly populate any technical article from Tomshardware, I must say that I can no longer find these articles, filled with benchmarks and data, in any form on this website in the last year or so. It is this in combination with the added addition of various articles hurried appearances (they look as though the whole thing was a rough draft that they wanted to finish and move on to something else, keep in mind I'm not talking about all of the articles here) that causes me to believe that Tomshardware is a lesser site then it once was.

You can say I'm wrong or even pick my English abilities apart, but I don't see any way to refute the base of this post, merely that Tomshardware no longer publishes articles of it's previous quality.

 

[cleeve]

I'd have a look at the recent 4830 crossfire review, the triple-sli 260 vs double-sli 280 review, and the 295 preview.

All recent articles, and lots of benchmark data there.

Having said that, we are conscious that our game benches could be better and we're trying to get a system in place where we will be reviewing them often and updating them on a more regular basis. As well, we're making an effort to always add power usage benchmarks to new articles, and we're working on temperature standards as well.

So we're not perfect, but we're working on it!