Kyle VS Anand on the 3870X2

Status
Not open for further replies.
Don't know man, doesn't seem to show any bias to me, they point out flaws with the 3870x2 in the review. Anyways all I know is I can't wait to get this card in my grubby little paws so i can compare for myself and end my own confusion :D
 
Chris_B said:
Driverheaven for a while i found very very pro ati, seen plenty of reviews of cards and their results on that site were totally different to other sites and always made the ati cards out to be better performers.
Click to expand...

So what, i can say the same about this site and tom's hardware being very pro Nvidia. So please lets not go there.
 
The DH review seems really comprehensive to me, and with all the controversy going on over here at [H] and the rest of the community I think maybe someone got their ego out on a limb here. I am not saying [H] is wrong but calling every review on the net "canned benchmarks" seems really biased to me and sometimes people can get caught up in their own opinions and causes and stop thinking about what is right and what is right for just that person.

It sounds crazy when you think about it, I believe everyone has their heart in the right place, although some places I do not trust their reviews because looking at how they test and reading each page really seems like they were rushed to get the hits on their site for ad revenue.

I already ordered one of these little beauties so time will tell how good they are, I am certain it is a good investment for future proofing my rig and I hope to buy another one later on when the drivers mature and things settle down more. I try not to get caught up in the ATI vs Nvidia trolls since I care about performance and not a brand name, I generally buy ATI although it really has just worked out that way when I purchase my new vid card ATI seems to have a good card available at a good price. Maybe it is not the performance king of the world per say, but I am happy with it (would have bought the 8800GT earlier if the prices did not jump through the roof after initial launch).
 
noseguard20 said:
So what, i can say the same about this site and tom's hardware being very pro Nvidia. So please lets not go there.
Click to expand...

I think toms and here have been pro everything at one point or another depending on what screaming fangirls you listen to.
 
noseguard20 said:
So what, i can say the same about this site and tom's hardware being very pro Nvidia. So please lets not go there.
Click to expand...

Still not seeing the pro nvidia stance here. Please point this out to me. Are you refering the gold and silver awards they gave the 38xx series, or maybe refering to the 3850 as the best card under 200, oh wait you must be refering to when they called the 2900 a flop and not up to par when all the other review sites except 2 disagreed with [H]. Oh wait they were right. Were you calling AT, TH, and others ATi accomplices and being bought out? I highly doubt it.

I'm sorry all I see here is a site calling it like they see it, and that is from a gamers point of view, which when the rubber meets the road is all that matters.
 
Mercyreaper said:
The DH review seems really comprehensive to me, and with all the controversy going on over here at [H] and the rest of the community I think maybe someone got their ego out on a limb here. I am not saying [H] is wrong but calling every review on the net "canned benchmarks" seems really biased to me and sometimes people can get caught up in their own opinions and causes and stop thinking about what is right and what is right for just that person.
Click to expand...

Steve, our news guy misstated that, I have personally corrected it.
 
As long as the review accurately states the methodology with which their data was acquired, there shouldn't be any problem. After all, data is just data, and isn't inherently "good" or "bad".

Personally, I like to see as much data as possible, and from there I can understand the general performance characteristics of the hardware under a wide variety of operating conditions. The H reviews are generally pretty sparse on raw data, and heavy on conclusions (normally quite valid ones, I just prefer coming to my own).

As a result I have to supplement their reviews with a whole bunch of other sources to satisfy my desire to understand how the hardware behaves.


disclaimer: I work for the company that makes the product which was reviewed.
 
Well I for one thought the article by Hardocp was pretty lame by the way they tested the cards. You can't get an accurate readings by doing a "run through" in a game and judging its fps. That has to be the most unaccurate way to compare hardware.

The ONLY way to have a fair comparison is by a preset benchmark in the game itself or by some external program that runs each graphic card through the exact same graphics.
 
jimmyb said:
disclaimer: I work for the company that makes the product which was reviewed.
Click to expand...

You work for ATI? When are the new drivers going to be released? Looked on the AMD.ATI page and the 3870x2 is not even listed yet under drivers?
 
tc17 said:
Well I for one thought the article by Hardocp was pretty lame by the way they tested the cards. You can't get an accurate readings by doing a "run through" in a game and judging its fps. That has to be the most unaccurate way to compare hardware.

The ONLY way to have a fair comparison is by a preset benchmark in the game itself or by some external program that runs each graphic card through the exact same graphics.
Click to expand...

As has been noted multiple times in the review thread, we do not base our opinions on conclusions or the run-through. We base them on our gaming experiences. The run-through is simply there to "prove" our experiences.

If you do not however like the fact that we evaluate the gaming experience supplied by the card by using actual gameplay, I can accept that and wholeheartedly say that you will find nothing of value in terms of video cards at HardOCP.
 
I want the best of both worlds. I want to see what [H] thinks are the max playable settings with each card, but I also want to see benchmarks with the cards at the SAME SETTINGS so I can get a better idea of how the cards stack up against each other. The "max playable" is fantastic, but the benches at different settings are pointless.
 
xp3nd4bl3 said:
I want the best of both worlds. I want to see what [H] thinks are the max playable settings with each card, but I also want to see benchmarks with the cards at the SAME SETTINGS so I can get a better idea of how the cards stack up against each other. The "max playable" is fantastic, but the benches at different settings are pointless.
Click to expand...

Somehow you are missing that we don't "bench" anything. We just play the game.
 
xp3nd4bl3 said:
No, I totally got you on that. The most important thing is gameplay. But in addition to the "max playable settings" you provide framerates and framerate comparison charts. My point is that comparing the framerates of the GPUs with different in-game settings is pointless. It's nice to see that CardA can run with shaders set to HIGH while CardB requires them to be on MEDIUM to be playable. But that doesn't give me any real idea of how much faster CardA is. It doesn't say that CardA is 15% faster, for example. I have no idea how much more power CardB has that allows it run shaders on HIGH.
Click to expand...

Well I am glad you find gameplay which is the most important thing to you is better measured through timedemo means. Good luck with that, sorry to waste your time with our pointless evaluations of actual gameplay and what you can expect out of particular hardware.
 
Kyle, he is just saying one card with settings set to high and another card with settings set to medium and both shown on the same graph at equal FPS is confusing and shows nothing
 
cannondale06 said:
well maybe Im crazy but I thought the 2900 and 3800 series do much WORSE not better with AA than Nvidia. :confused:
Click to expand...

4xAA|16xAF


8xAA|16xAF


Source:
http://www.computerbase.de/artikel/hardware/grafikkarten/2008/test_ati_radeon_hd_3870_x2/24/

Those are performance percents not FPS. Unless I am misunderstanding something. I don't speak German.

So to me it says from all the benches done @ 2560x with 4xAA|16xAF the Ultra gives you 97% of the performance of the HD 3870 X2 on average. When you move to 8xAA|16xAF @ 2560x it can only deliver 49.

I've heard it said before the nVidia cards pass the ATi cards @ 4xAA and the ATi cards often "catch up" @ 8xAA. Notice how the single 3870 also redeems itself.

I would agree with the other fellow that suggested more sites should try 8xAA, particularly when you are testing mid to high end cards.
 
Tamale said:
Don't put up numbers and comparison graphs for different cards at different IQ settings. This confuses people.. make your subjective analysis and then make a separate (and less emphasized) apple-to-apples numbers only comparison.
Click to expand...

That's exactly what I've been trying to communicate, but I've obviously done so poorly.
 
I'd agree that [H] should add their own 'canned benchmarks' as well for comparison. As Kyle says, it doesn't take a lot of time to do canned benchmarks, so adding them into the reviews shouldn't be that much of a burden.

If it can't be done for each review, at the very least they should have a dedicated article addressing this issue. It would definitely get the page hits amongst enthusiast circles and get discussion going...
 
I def would like to see him bench 1680x1050 more often.

When I bought my new monitor 2 holidays ago it was fast becoming the sweet res for 20-22inch(sometimes 24inch) monitors and personaly a 22 inch monitor is good enough for me for awhile .

I'd also like to see results at 1080p . I have a 42 inch tv and might want to play on that once in awhile. I know you like to tell us what you feel is playable but its nice to see these other resolutions also . If as you claimed your way is the best way to judge what a card is capable of shouldn't you give us more information so that those of us not using insane resolutions have some idea of how they play at your resoultions.

I used to like seeing games tested in 3-4 reslutions . It gives you an idea of options . Now I understand these things take time but you don't have to go as in depth as the higher resolutions .
 
there may be more to this possibily. Not only do Kyle and Anand have different performance reviews, but they are on opposite ends of how loud the fan is on the card. Is it possible that that might have something to do with it, that the ATI card wasn't running as hard for some reason in [H]'s setup and the fan wasn't blowing as fast for soem reason? No theories, but I find it interesting that the anand review says it's a very loud fan and Kyle says it's quiet. I guess it could be a difference in perception, but it could be other things as well, no?
 
conscript said:
there may be more to this possibily. Not only do Kyle and Anand have different performance reviews, but they are on opposite ends of how loud the fan is on the card. Is it possible that that might have something to do with it, that the ATI card wasn't running as hard for some reason in [H]'s setup and the fan wasn't blowing as fast for soem reason? No theories, but I find it interesting that the anand review says it's a very loud fan and Kyle says it's quiet. I guess it could be a difference in perception, but it could be other things as well, no?
Click to expand...

tech report confirms it's much louder than competing cards too. this is a good question... maybe [H] really did get a bum card.
 
Doesn't seem right that they would get a card that is just "slightly" bummed.
 
Mehen said:
Doesn't seem right that they would get a card that is just "slightly" bummed.
Click to expand...

no, but i've seen weirder things in my time. the steam processors might be smart enough to ignore problematic units.. if they were maybe only a few on the sample [H] received were bad.. *shrug*
 
noseguard20 said:
So what, i can say the same about this site and tom's hardware being very pro Nvidia. So please lets not go there.
Click to expand...



really? Their GPU charts show that the 2900XT is RIGHT BEHIND the 8800GT 512! (NOT GTS), yes and right behind meaning withing a 5% margin.

so please, do not go there


90% of the reviews i've read of the 8800GT (i use those because at that point Catalyst 7.11 drivers were out for 2900XT's, and the 2900XT was basically right behind the 8800GT, certain exceptions were there of course) show the 2900XT right behind the 8800GT. the 3870 right behind the 8800GTS 512. then obviously the 3870 x2 is probably better than a GTX as GTX's are not much better than GTS 512's.


however hard's testing really dissappointed me. it seemed like they wanted the 8800GTX to be better. I checked like 3 other review sites and they all said that the 3870X2 was better than a GTX.

I choose the majority. sorry

and yes, ATi is WAY ahead in drivers. I watch as each driver is released. I didn't go for a 2900XT because the few reviews i read showed like catalyst 7.8 or 7.5 drivers. 7.11 drivers really boost the performance. thats technology at work right there.
Secondly, Nvidia drivers SUCK! BADLY. You think ATi's are bad? I ahd to go thru like 4 different driver versions (169.xx) before I found one that was stable, didn't freeze games up, and didn;t give me blue screens when I woke windows up. thats no technology work right there.

and I know for a FACT that ATi image quality is MUCH MUCH better than Nvidias. Crysis doesn't look stunning at times with my 8800GT. some of it looks bland, and thats lame.
 
Tamale said:
tech report confirms it's much louder than competing cards too. this is a good question... maybe [H] really did get a bum card.
Click to expand...
well if you read the rest of what Tech Report then theres not really a discrepancy.

"And yet, I really can't complain too much about the X2's noise levels. The cooler can be loud when the system is first powered on, but that's it. During normal gaming, like the scenario we tested here, the card really isn't very loud. It's at least not into the "annoying" range for me. We're talking loud-whisper level here."
 
cannondale06 said:
well if you read the rest of what Tech Report then theres not really a discrepancy.

"And yet, I really can't complain too much about the X2's noise levels. The cooler can be loud when the system is first powered on, but that's it. During normal gaming, like the scenario we tested here, the card really isn't very loud. It's at least not into the "annoying" range for me. We're talking loud-whisper level here."
Click to expand...

I'm a little skeptical of that subjective analysis.. if the meter registered that large of a difference compared to the other cards minimal differences, I think it'd be very noticeable. I hate video card fan noise..
 
I'm really enjoying this debate.

*snip*

Time to suck it up Mr. Bennett, you either got a bum card or your testing method really really is not fair. I must say though that your "real world testing" method, while intuitive, is only good when you can perfectly reproduce the results.

It's like reading car reviews, ya the good reviews will give you the 0-60mph times, the quarter mile times, and the skid pad, the great reviews also take the car around a track and then give a "real world" score.

Only one type of testing is obviously flawed. What Mr. Bennett has succeeded in is showing how flawed his die-hard approach to just one type of test is. Time to pull up the terrible "canned" benchies and once again include them. Remember it's an easy way to see if you did end up with a bum card.

*snip*
 
^^^ you should read ALL the reviews out there before stating Brent not Kyle got a bum card....


Read PC Per, Hexus, Overclockers Club, Extremetech and then Hard Ocps and you will see not that big of differences....
 
btdvox said:
^^^ you should read ALL the reviews out there before stating Brent not Kyle got a bum card....


Read PC Per, Hexus, Overclockers Club, Extremetech and then Hard Ocps and you will see not that big of differences....
Click to expand...

Eh the real point of my post is to say. Look just using one form of testing will always be flawed. "Canned" tests take so very little time there's no point in not doing them. The "real world" method is also great. Now put them together and BAM you've got the best review possible.

[H]ard|OCP is all about giving the best info and the best reviews. I think that honestly doing just one form of testing is a disservice and should be rethought.

And you're right, this isn't a review done by Kyle, but Kyle has done much more to defend it than Brent. Anyways, when you're the boss you get the praise and the heat. Remember "The buck stops here" is the sign on the boss's desk, not anyone else's.
 
lol if you guys don't like or agree his reviews then don't read them, simple as that. Don't bash him for posting his review, he took his own time to do them, and Im sure people greatly appreciate him for doing so. I guarantee if you guys posted your own reviews some fanboy would bash you for something you did or did not do. If you don't like it, use your mouse to click the little x in the top of the corner.
 
Kyle and [H] have made it their stance and mission over the years to change how benchmarking is done and this is how they feel it is best for gamers.

The problem for most people though is that "highest playable settings" and the way they are interpreted is often very highly subjective. When gamers look for information on cards, they want hard numbers, they want empircal scientific method results, it's just (enlightened) human nature. Personally, I do prefer a restricting timedemo or bench that is the exact same sequence when comparing the performance of cards, especially because it is easily to detect where bottlenecks occur and it minimizes system differences or subjective reviewer results. it's more easily readable, understandable, and it allows the enduser to process the information themselves (rather than having the review do it for them). Also, face it, many people who buy these cards buy them for a large part in having the psychological boost knowing that theirs is the fastest. People overclock the hell out of their cards - for what may not be any noticible subjective increase in game playability! - Rather they do it because they like the numbers, it justifies their purchase, it stokes their egos like many other things people buy.

[H]ardOCP is great but I must admit, I generally don't really look at the reviews much anymore except just to get a feel for how well a card can play a game. When I make a buying decision, I always look to hard numbers on Anand and Toms's charts. When you buy a card, it's not usually to see how playable a handful of games are, but you want to know that it will be able to handle anything and the buyer wants to be able to extrapolate the lifetime and futureproofing that the card will provide. That's one of the most important things isn't it? To extrapolate how long this card will last you until the next upgrade cycle and if it is worth doing now. It's just much easier with traditional numbers and canned benchmarks/demos which allow people to work that out for themselves or compare cards via a VGA chart. Contrary to the added endnotes to the 3870X2 review in which Kyle says that the tide is turning and other sites will see the benefits of this method... I think [H]ardOCP will be alone in this arena and other sites will continue to bench and timedemo as they have always done and traffic will continue to flow to those sites with traditional benchmarking. But hey, that's a good thing and it makes [H]ardOCP unique and it gives gamers much more than just benchmark numbers to look at.
 
Darkstofknightz said:
lol if you guys don't like or agree his reviews then don't read them, simple as that. Don't bash him for posting his review, he took his own time to do them, and Im sure people greatly appreciate him for doing so. I guarantee if you guys posted your own reviews some fanboy would bash you for something you did or did not do. If you don't like it, use your mouse to click the little x in the top of the corner.
Click to expand...

Like I said, I've been there. Ya I've been bashed. People have even claimed I was employed by OCZ because I was willing to test their products when everyone else said they were evil and should never be supported.

The point is to have the best review, it's not a personal attack, it's fact. Just using any ONE method of testing is always only going to give part of a picture. I read Tom, Anand, Kyle, and even many small sites when something peeks my interest. It's a fairly good rule of thumb that if your review is markedly different from everyone else's something is up. So it was already stated that they'll be reevaluating results. I'm sure they'll be receiving cards from various manufacturers as well, we'll keep getting more of a picture.

Kyle's assault on a form of benchmarking is that it doesn't show the whole picture, well neither does just using his form of benchmarking.
 
Hardware is the backbone but software and drivers can make or break the card. Sure the new Dodge 1500 Hemi's make a ton of HP, but they are not as efficient at delivering it as the new Tundras or Titans. Unlike Car's though, software is easily updated for free, so indeed this new ATI card could be a sleeping monster, but I still have my doubts. I think it's good for ATI, and finaly Nvidia has some competition. Persoanly I like Nvidia software, but ATI hardware (over all, just not recently). My Nvidia cards seem to have a MUCH higher failure rate than my ATI cards. But ATI drivers never seem as flexable or solid. For the longest time you had to run 3rd party software to have a different background on each of your monitors with an ATI card, but Nvidia it was built into he drivers and worked well too (just an example I have not had a new ati card for a while).

Well either way its making me wish I had waited on my GTS that still sitting on my desk waiting for some removable ramsinks (unused for a month now :p ). I'm not sure which card is clocking better but I was reading over on another forum that the ATI card is locked at 999?

Time for Nvidia to release it's new line of cards! (not some stupid x2 card.)
 
Tamale said:
Your methodology and passion are both commendable, but listening to all the feedback this article has generated and judging by the fact that numerous 'review lists' are omitting the [H] review (the inq and dailytech come to mind) should be an indication that you might want to start listening too! Consider providing some baseline comparison (read: 'canned benchmarks') graphs in addition to your subjective analysis.. or at the very least some more insight into your methodology...
Click to expand...
I think Kyle has made it very clear over the years that this is what he thinks is best and if you dont agree than there are a bajillion other reviews available for people to read.

Personally, I agree with him. At least he gives people a choice, an alternative to the canned and meaningless benchmarks. Anybody making a buying discision based solely on one review is doing themselves a disservice (and being an idiot as well).

Who cares if the GeForce is 5% faster or 15% faster. For approximately the same framerate you can bunch the resolution. I dont personally change framerates to match the card, I change settings to meet allowable framerate goals.
 
After seeing this happen upon most major release on [H] I am pretty sure its just about clicks a money. Nothing sales better than controversy.
 
Status
Not open for further replies.
Back
Top