Tom's Hardware Video Card guide getting worse and worse..

[LAAkuma]

Tom's Hardware is supposed to be the best of the best at reviews and benchmarks. I come to Toms Hardware every single day and read every article, watch every video. And in the last few months, something is totally clear about Tom's Hardware versus other review sites. Its all in the details. Tom's video card reviews are almost dead last in terms of how complete, informative, etc. This is for the following reasons:

Lets take the HD 3870 card for example. Now here is where Tom's failed.

#1 No crossfire benchmarks? How can you have a ATI video card review without crossfire? its totally absurd at this day and age. Totally unacceptable.
#2 And this made me LOL, and probably everyone else too. What the hell is the point with the dual video card Dx10 2007 http://www23.tomshardware.com/graphics_2007.html?modelx=33&model1=1060&model2=853&chart=318
what is the point of this nonsense page? I mean games like MS simulator, Doom 3, and Might and Magic?

Are you serious? How about sticking to your review, and including games that matter like perhaps Crysis, ut3, Quake Wars, Orange Box, Hellgate london, and things people actually play in droves?

I don't know who is doing the work for these 2007 VGA charts, or what have you, but it looks like pure laziness to me. Maybe the people doing the reviews are not being supervised like they should and are just playing TF2 all day long while on Tom's dollar. I don't know.

But there has got to be some serious updating to the VGA charts when your including cards like the G92, RV670, crossfire, SLI and all the newest technology.

Not stupid MS flight simulator, might and magic, nonsense!

So here is what I think is worthy of Tom's site.

#1 Full market roundup of all the G92 cards with benches
#2 Full market roundup of all the HD3800 series cards with benches
#3 VGA charts that people actually can learn something from.
#4 Updates of all the cards EACH time a new driver comes out for it.

You do those 4 things, your #1 in the world at hardware video card wise.

That is the key to what makes a good site right? lets not get away from the basics.
#1 thing that makes a successful sight is to generate content people who visit CARE about and WANT to read!

This is where the site is failing in the video card review department.

Now I laid it out right there what people want to see. 1-4 above. If the people getting paid by Tom putting up these video card reviews can't do it, then I think Tom should hire somebody who can do it. Its business, Tom is in this to be the best.

I visit no less than 8 hardware review sites a day reading up on everything, and on video cards and I know what makes a good review.

If your going to review anything new from ATI without crossfire, don't even bother. After all, I think its safe to say most everyone with a X38 wants to know, and X38 is pretty much the dominant platform for the enthusiast market.

Thats my opinions, I hope things get fixed. If they don't or can't get fixed, I will freaking quit my job, go work for Tom and fix it myself. I know exactly what to do, and can do it easily.

LAAkuma

 

[Pollux]

+1


Leave Battlefield 2142 as I play it

 

[teh_boxzor]

+2 i mean how ridiculous is it that i have to use doom 3 to see how much fps my games average on steam games (because some steam games use the doom 3 engine) and yeah how will i know how my 8800gt will fare at playing crysis on 1280 x 1024 should i decide to upgrade?

come on tom's stop being so close-minded.

 

[SirCrono]

I think oblivion performance still shows the overall power of any videocard.

But the chart should be updated, it's not that much work.

 

[San Pedro]

Yes newer games would definitely be a plus.

 

[TheGreatGrapeApe]

Some valid points, but two things that are off base;

No one does #4 anywhere. The closest is the driver reviews from places like TweakTown and such and that's for 1 or 2 cards in limited number of games (mainly 3Dmarks). Also would you use Betas or WHQL only? If using Betas the cards wouldn't be done testing before another one leaked out. And if not you might still go 3-6 months between some driver updates.

As for M$ it belongs in there as much as anything else because it's such a different game, and you want consistency too. 3-4FPS , 2 sims (FSX and a driving one IMO) and 2 MMO / RPGS would give you good balance within a review. Hellgate doesn't belong in a big review, there's just not enough need. You go with the titles that have consistency and are game-engine defining, so Crysis, QET, HL2:E2/TF2 , UT3 all make sense because of their application to other games in general and because of how they strain PCs. If anything I'd put COD4 aheadof Hellgate.

Anywhoo, I think you'll find agreement from most of us here on the VGA charts, but I don't think they're aimed at us, I think they're aimed at the Best Buy "Tell me what to buy" buyer, and while updating the VGA charts would be nice, I'd simply prefer they concentrate on the full reviews instead because charts are for people who don't really look in depth anyways.

 

[TheGreatGrapeApe]

(because some steam games use the doom 3 engine)

D3 engine is still pretty limited compared to the value of the UT3 engine, which will likely power more games than HL2, D3 and Cryengine combined, although it may be close.

 

[LAAkuma]

HI guys, thanks for the encouragement of the basis of my argument.

GreatApe. I hear you on the driver update thing. Some sites do it. But to say no other sites do it is something I know Tom's doesn't care about, because Tom's is a LEADER.. Leaders don't copy what other people do, they make up their own standard and others follow them.

I think if Tom's wants to be the leader, they have to do what I said. And like I said, if they don;t have the right people to do it, I will do it for them.

I mean Tom's has the capacity to be the worlds best hardware review site. But they can;t do it with the weak hardware reviews they have been doing. I REALLY think there has to be a new rule in effect or something at Tom's studios. Rule #1. No playing any games during working hours whatsoever, or your fired, unless your specifically reviewing that game. It looks like to me a bunch of non-professional gamers are doing these reviews and works.

LAAkuma

 

[TheGreatGrapeApe]

GreatApe. I hear you on the driver update thing. Some sites do it. But to say no other sites do it is something I know Tom's doesn't care about, because Tom's is a LEADER.. Leaders don't copy what other people do, they make up their own standard and others follow them.

I understand what you're saying, but as someone who did the Catalyst forum sticky for just over a year reviewing each release on 1 card in a bunch of games/apps, I can tell you that keeping a list of cards from both IHVs current would be a full time job, which translate to a pretty heafty cost. And you still wouldn't please everyone, because by the time you got the latest batch finished someone would launch Beta driver XXX.xx that negates your work in some people's minds.

I mean Tom's has the capacity to be the worlds best hardware review site. But they can;t do it with the weak hardware reviews they have been doing. I REALLY think there has to be a new rule in effect or something at Tom's studios. Rule #1. No playing any games during working hours whatsoever, or your fired, unless your specifically reviewing that game. It looks like to me a bunch of non-professional gamers are doing these reviews and works.

The thing is there aren't dedicated studios for this, the reviwers are a collection of people with other jobs, wich is the same for every other site, and true of Thomas Pabst when he started the site. Unfortunately the resources available aren't such that it makes it easy to allow for that kind of depth and detail. Pretty much all sites do what they can and then they do something different from others to diffferentiate themselves.

I'd love to see the VGA charts be more up to date, but I think in the grand scheme of things to me, they're less of a concern than seeing more detailed specifics into the games themselves and into their use in things like HTPCs and non-traditional areas. I prefer the details rather than the charts, but I do agree that if they do exist they must be kept current on at least a quarterly basis.

 

[LAAkuma]

Does anyone here Know Tom so they can tell him about this growing concern?

LAAkuma

 

[buzzlightbeer]

I wanna know why they used a over clocked 8800gt ,was it to make the 3870 look bad? i think they should either use normal ones or use a over clocked 3870

 

[caamsa]

Ahhh I am not sure about this but I think they are gonna bench some new games and add them to the chart. Some of the newer cards were already benched on some of the newer titles so I imagine they could just incorporate that information into the charts and then bench some of the older cards on the newer games....etc........

 

[babybudha]

I agree. TH needs to pull up their socks.

 

[mrmez]

All valid points.
However i do realise that updating these benches with new games is a LOT of work. Like a LOT.

I use them more for comparing which card is better, not really the EXACT fps i can get from what h/ware.

 

[aznstriker92]

What I don't understand is why Tomshardware used those old games in the first place. I mean the chart was done over the summer I beleve and there were good games out then too. Whats with doom 3 and magic might w/e or war hammer?
And what kind of crap drivers where they using for the 3800s? They had the 3850 on par with a 7950GT and X1900 series. If you look at other reviews, the 3850 destroys both of those series.
Also if people didnt notice, the 3870 was also Overclocked. 825/2400. But its still had crappy results.

 

[justinmcg67]

...I think they're aimed at the Best Buy "Tell me what to buy" buyer, and while updating the VGA charts would be nice, I'd simply prefer they concentrate on the full reviews instead because charts are for people who don't really look in depth anyways.

Funny you say that.

About 4 years ago when i really started to get into PCs I went into Best Buy to purchase a video card, I had no idea what to get, I just needed Command & Conquer generals to run on my PC at the time. So, I was going to purchase a Radeon 9600 (wasn't a bad card than) and the guy helping me at Best Buy took me to Tom's Hardware Guide and the VGA charts to show me that my product wasn't even being benched any more, and that I should proceed to purchase a 7800GTX for $400.

Ahhh, thanks for bringing back the memories.


Anywho, on a more serious note.


I don't think most of you understand how much time goes into updating every card from the Radon X1k series to the current NVIDIA 8 series and HD3k series of cards. That's this many cards...

Radeon X1300 128/256/512 (3)
Radeon X1300 Pro 256/512 (2)
Radeon X1600 Pro 128/256/512 (3)
Radeon X1600XT 256 (1)
Radeon X1650 256/512 (2)
Radeon X1650 Pro 256/512 (2)
Radeon X1650GT 256/512 (2)
Radeon X1650XT 256 (1)
Radeon X1800XTX (1)
Radeon X1800XT 256/512 (2)
Radeon X1800GTO 256 (1)
Radeon X1900XTX (1)
Radeon X1900XT 256/512 (2)
Radeon X1950XT 256/512 (2)
Radeon X1950GT 256/512 (2)
Radeon HD2400 Pro 256/512 (2)
Radeon HD2400XT 256/512 (2)
Radeon HD2600 Pro 256/512 (2)
Radeon HD2600XT 256/512 (2)
Radeon HD2900 Pro 512/1024 (2)
Radeon HD2900GT 256 (1)
Radeon HD2900XT 512/1024 (2)
Radeon HD3850 256/512 (2)
Radeon HD3870 512 (1)

* 40 total cards

GeForce 7200GS 128/512 (2)
GeForce 7300LE 128/256 (2)
GeForce 7300GS 256/512 (2)
GeForce 7300GT 256/512 (2)
GeForce 7600GS 256/512 (2)
GeForce 7600GT 256/512 (2)
GeForce 7800GT 256 (1)
GeForce 7800GTX 256 (1)
GeForce 7900GT 256 (1)
GeForce 7900GS 256 (1)
GeForce 7900GTX 256 (1)
GeForce 7950GT 256/512 (2)
GeForce 7950GX2 1024 (1)
GeForce 8400GS 256/512 (2)
GeForce 8600GT 256/512 (2)
GeForce 8600GTS 256 (1)
GeForce 8800GT 256/512 (2)
GeForce 8800GTS 320 (1)
GeForce 8800GTS 640 (1)
GeForce 8800GTX (1)
GeForce 8800GTX Ultra (1)

* 31 total cards

I may have missed some, most was from memory, the others were from a Newegg's help. But as you can see, it's a long list. It's a total of 71 cards combined from both companies.

Now, take 71 cards, benchmark them in Crysis, on Low, Medium, High, Very High, and with a mix of 2x/4x and beyond AA, than UT3 and it's settings, World in Conflict on Low, Medium, High, and Very High on with a mix of settings, FSX, Call of Duty 4, and you'll soon understand why these things take a VERY long time to update.

And if you want to do what some have suggested, which is to update the charts with every new driver, and you're talking about 16 hour days for weeks on end just to get the charts up to date...until the next driver comes out.

I hope that this gets across the point that you should be happy that they do these things for you, and that you can use this as a reference, because as far as I know, Tom's Hardware Guide is the ONLY site to have a full, comprehensive list of cards, as well as games to select from, on specific settings. Not only this, but Tom's Hardware benches at OC levels sometimes as well to give you a better representation of the cards.

Constructive criticism is great and all, but don't go over board and just be thankful that a site is offering tools to the public such as the VGA Charts.

 

[teh_boxzor]

maybe tom's is getting paid by nvidia to make their cards look good?

 

[rwinches]

Ok so I guess you are making the assumption that Tom's gets to keep every card they evaluate.
I don't think so, just look at the cards on the list then add up the cost of maintaining that inventory. How many hours would it take to run the tests as you keep adding cards.
I wish they did not test as if we all could afford huge flat screen monitors.
I preferred the static charts that just listed the spec for the cards.

I still think a 7600 GS 512 AGP is a good card, but what do I know?

 

[caamsa]

maybe tom's is getting paid by nvidia to make their cards look good?

Naaa if they were getting paid they would have benched more games. I "personally" prefer Nvidia cards. To me they work no matter what....no need to really fool with them.

I think the....cpu and gpu charts are useful especially if you are looking to upgrade. I would rather have an idea about how a new GPU performs with a certain game before going out and getting the game or deciding I need to upgrade to a new video card.

 

[TheGreatGrapeApe]

Funny you say that.

About 4 years ago when i really started to get into PCs I went into Best Buy to purchase a video card, I had no idea what to get, I just needed Command & Conquer generals to run on my PC at the time. So, I was going to purchase a Radeon 9600 (wasn't a bad card than) and the guy helping me at Best Buy took me to Tom's Hardware Guide and the VGA charts to show me that my product wasn't even being benched any more, and that I should proceed to purchase a 7800GTX for $400.

Ahhh, thanks for bringing back the memories.

No problems, I remember when I e-mailed Lars/Borsti to tell him his 3Dmark list from a review was colour photocopied and pasted to the counter top of the local Future Shop (which was later bought out by BestBuy and now you find them side-by-side in malls because people think they're different and thus think they are 'shopping around' by visiting both - which are the same people I usually mean).

Funny thing everyone starts out somewhere like that, and then some take that limited knowledge and improve upon it. And when you learn more of course you learn just how limited any single set of tests can be, and the best thing to do is to read multiple sites (without which most people wouldn't know the subtle differences like the 8XAA differences, and TAA/CSAA/etc.)

You missed a few cards (I'm sure I noticed it more than most because you specifically mentioned it, so I actually re-checked the list), but it accurate in the relationship and easily converys the difficulty of the task.
I may have missed some, most was from memory, the others were from a Newegg's help. But as you can see, it's a long list. It's a total of 71 cards combined from both companies.

Also, instead of making another post (I'm going out) just an FYI to everyone, TOM aka Dr Thomas Pabst hasn't contributed or owned the site in a long time. He is the creator, but he's not directly involed anymore, and so trying to bribe him wouldn't make much sense.

 

[UTNemesis]

I have to agree about the point regarding the new DX10 charts. They benched DX10 cards, but with no DX10 titles. It just seems off base.

 

[LAAkuma]

I know exactly how to do the system if I worked for Tom. And I mean exactly. Of course, if he started such a system, he would have to start over with any cards he wanted to display, and use those for x time.

Its very simple really. Very.

Before the new year, you get a few boxes. They all have the fastest chips in them you can buy at the time.
You then swap out each card and drivers, and bench them all as you get them during the year. So your not sitting there doing 100 cards in a row, its more like 1-4 a week, all year long perhaps. very easy.

Then at the end of the year, you get new boxes reflecting the fastest at that time, that you keep another year, and go from there.

For each new year, you simply take say, the hottest 4 or 5 gpu's of the last year and add them to the database for the new year.

I mean I don't even do this for a living, but its literally so easy.

Right now, right now right now, at this time, there are only a few video cards people would even consider buying. So might as well start the new database right now. Start with.
8800 GT 512
8800 GT 256
8800 GTS G92
HD3750 512
HD3850 512
HD3850 256

Those are the only cards anyone should consider buying right now. and those will be real popular I would guess all the way through 2008.

That would start off 2008 on a clean slate.

Each one of the above cards should be in SLI and Crossfire as well, and displayed in the SAME table as the single cards, and each time the drivers are updated for NVIDIA and ATI, each card should be benched again and the tables updated.

Right now there is really only 3 markets.

Cheap - 3850
mid range -3870-8800 gt
and high end 8800GTS G92 or crossfire ATI

Thats it.

LAAkuma

 

[Pintail]

Hey all, I take the charts with a grain of salt, they gives me a general idea, and reviews are what generally make up my mind on a purchase.

But having said that, I WOULD like to see how my BFG 8800 GT OC2 compares to the card I upgraded from.

GeForce MX 440. :lol: :lol: :lol:

 

[pchoi04]

Amen... I think everyone is feeling this right now. As for the VGA charts with the amount of video cards there are... Not to sound to critical, but I dont know about you guys but I really dont care how a x1300 or a 7200gs performs on any benchmark other than knowing I can browse the internet. With that being said maybe they should just shrink down the number of video cards being tested that way they can afford to spend more time benchmarking video cards that the majority of the people are interested in. This way we can see how the cards are improving with newly release drivers.

 

[Rabidpeanut]

Tom's is really dropping the ball on this one, these charts suck, ESP for ATI who update their drivers every month that is why i dont use these charts anymore.