[H] Cpu Scaling Review

Brent_Justice said:
All reviews will deal with retail available cards.
Click to expand...

... and will use 939 motherboards that are not yet available :-P

j/k!

Now my $500 3500+ seems somewhat worth it and why I bought it in the first place. Nice reading there [H].
 
Brent_Justice said:
Don't know what to tell ya, the results shown are from real gameplay in those games on the cards tested. No timedemos were used, only real gameplay evaluation. The performance was repeatable.
Click to expand...


In all honesty, half the review was CRAP. You are lucky you even put a apples-apples comparison. I was about to fire off a email so quickly on how stupid your review was.

When you are doing CPU scaling benchmarks, I think you should stick to time demo things that people can repeat. C'mon, this is about scaling. Traditional benchmarks are SO much better to understand.

And wtf is with this format? Have to click a link to get anywhere? As your settings etc etc change.

Oh well, until someone tells me that picture quality means ANYTHING AT ALL, I do not care. I'll continue kicking your ass with my ti4200 while you look at the ground in HQ.

Edit: While we are on the subject, Where is my memory bandwidth benchmark with these cpu's? I want to see CPU, Memory, Video Card all together. Which REALLY makes a difference? Scale up some FSB, show us what it really is.

This review is incomplete, half of it is waste.
Try again.
 
If it comes to Hardocp reviews versus all the other website's reviews, I am inclined to believe Hardocp. How many other websites show detailed graphs of fps over time? How many other websites take the time to playtest each card in each game and come up with the optimal settings that real users will benefit the most from? Seriously, I don't trust the other sites honesty and dedication 1/4 as much as I trust ocp.
 
agar said:
Found this review very interesting. Is it me or is the X800pro holding its own against the 6800U? I have to go over and look at it again, I just can't believe what I just saw. Yep, I guess those are the results they found. Another interesting detail was that the XT was put at a higher res and it would still either beat the 6800U or be withing 3-5fps. Is it me or are these results quite different from the recent reviews using the 61.34 drivers (ex. Guru3d)?

http://www.hardocp.com/article.html?art=NjMy
Click to expand...


I own one and with almost 4 ghz at my disposal I can beat the scores of any 6800u review i have seen..easily
 
Brent and Kyle, I can definitely see your point of view in only running varied tests to establish performance indicative of "real world" game play, but there is a segment of the gaming population out there that buys video cards for the numbers (whether they admit it to themselves or not). Regardless of what you may think about people who run benchmarks for the pure score and score alone, I feel you are really missing out on a niche of the video card market when you aren't testing using "Established" benchmarks. Indeed, many around these grounds just want to focus on the trees.

Your review has made me reconsider my position on my purchase of the 6800GT. I have cancelled my order with Outpost.com.


phas3d
 
Brent_Justice said:
I'm not seeing the same thing you are. He didn't test the same way we did, how could you possibly compare?
Click to expand...

Enlighten me. No seriously, I have no idea how you did it and how you percieve he did it so you're going to have to elaborate a bit more than that.

But MORE importantly, since we're on the subject of REAL GAMEPLAY...would it even matter? His experiance was that at 1600x1200 8xAA 16xAF max settings no ops it was "somewhat playable.. At some parts with action it dips into the teens" [his words]. I think knowing what you know about Nvidia's 8xAA mode you'd allow us to extrapolate the assumption that at 1600x1200 4xAA 16xAF max settings ops on Far Cry would be playable by your standards.

So this question is begged, if your article shows the max playable setting to be 1280x1024 2xAA 8xAF...what's going on?
 
Brent_Justice said:
We are DEFINITELY going to add that genre of games to our mix.

In your opinion which is the most taxing?

City of Heroes pretty good?
Click to expand...

Try going to an Anarchy Online tower battle (requires Notum Wars expansion pack). Now that is nuts :) Most of the time you have to reduce the visibility to 10% just to get acceptable framerates, as there are 100+ players onscreen with different armors, weapons, effects, etc all going off at the same time.

Might want to try Lineage 2 or Star Wars Galaxies, those seem to be the two most popular ones these days. CoH doesn't look too graphically intensive.
 
wilkinru said:
In all honesty, half the review was CRAP. You are lucky you even put a apples-apples comparison. I was about to fire off a email so quickly on how stupid your review was.

When you are doing CPU scaling benchmarks, I think you should stick to time demo things that people can repeat. C'mon, this is about scaling. Traditional benchmarks are SO much better to understand.

And wtf is with this format? Have to click a link to get anywhere? As your settings etc etc change.

Oh well, until someone tells me that picture quality means ANYTHING AT ALL, I do not care. I'll continue kicking your ass with my ti4200 while you look at the ground in HQ.

Edit: While we are on the subject, Where is my memory bandwidth benchmark with these cpu's? I want to see CPU, Memory, Video Card all together. Which REALLY makes a difference? Scale up some FSB, show us what it really is.

This review is incomplete, half of it is waste.
Try again.
Click to expand...


you still game with a ti4200?

you are so far out of the game you might as well play solitaire
 
phas3d said:
Brent and Kyle, I can definitely see your point of view in only running varied tests to establish performance indicative of "real world" game play, but there is a segment of the gaming population out there that buys video cards for the numbers (whether they admit it to themselves or not). Regardless of what you may think about people who run benchmarks for the pure score and score alone, I feel you are really missing out on a niche of the video card market when you aren't testing using "Established" benchmarks. Indeed, many around these grounds just want to focus on the trees.
Click to expand...

Oh well, I guess they'll just miss out then. We'll continue concentrating on presenting real world gaming experience that video cards are able to provide to the gamer.
 
wilkinru said:
In all honesty, half the review was CRAP. You are lucky you even put a apples-apples comparison. I was about to fire off a email so quickly on how stupid your review was.

When you are doing CPU scaling benchmarks, I think you should stick to time demo things that people can repeat. C'mon, this is about scaling. Traditional benchmarks are SO much better to understand.

And wtf is with this format? Have to click a link to get anywhere? As your settings etc etc change.

Oh well, until someone tells me that picture quality means ANYTHING AT ALL, I do not care. I'll continue kicking your ass with my ti4200 while you look at the ground in HQ.

Edit: While we are on the subject, Where is my memory bandwidth benchmark with these cpu's? I want to see CPU, Memory, Video Card all together. Which REALLY makes a difference? Scale up some FSB, show us what it really is.

This review is incomplete, half of it is waste.
Try again.
Click to expand...

Sorry to waste your time. You can save your clicks and not read our reviews from here out as they will stay focused on the real gaming experience delivered by the hardware, not a "benchmark" that we have found many times to mean nothing.

There will be tons of other review sites that used canned apples to apples benchmarks that will not waste your time.
 
DocSavage said:
If it comes to Hardocp reviews versus all the other website's reviews, I am inclined to believe Hardocp. How many other websites show detailed graphs of fps over time? How many other websites take the time to playtest each card in each game and come up with the optimal settings that real users will benefit the most from? Seriously, I don't trust the other sites honesty and dedication 1/4 as much as I trust ocp.
Click to expand...


[h] owns cuz of steeveeeeeeeeee.

He keeps my asp.net coding going. Without steve I would try to kill myself with a Learn ASP.net in 21 days book.

Still waiting for my daoc review, Lets see how that game works with TONS of people in a big rvr matchup. I suspect it all has to do with CPU speed, as each "person" is a object. Memory speed would help too. GPU? nah.

I actually do not play that game any more, but people keep thinking a faster video card meant a lot in a dx7 game (originally)
 
Brent-

Yes Ati's web site does mention the Platinum Edition & Nvidia site does not mention the EE.

Now we could debate this as a minor technicality(the boards do exist), we won't. I know you must operate within specific, set guidelines. We are not looking for fault, just making suggestions. I'm glad that you are secure enough to at least consider them.

Looking forward to the next one.

Thanks!

davidj
 
davidj said:
I feel if you are going to use the Platinum Editon for the x800, then the 6800 Extreme Edition should be used as well.

Maybe they are impossible for you to get right now. But they do exist.
http://www.evga.com/6800ultra.asp
Click to expand...

It seems as this point there there is not true EE card from NVIDIA. If you ask them, one does not exist. The EVGA card is an "OC" card to my knowledge.
 
theelviscerator said:
I own one and with almost 4 ghz at my disposal I can beat the scores of any 6800u review i have seen..easily
Click to expand...


Well, we could easily go back and "beat" our own scores as well. That would not be hard to do, just change paths a bit so that it is not near as busy visually. You seem to be missing the point that we are coming the cards in "identical" gameplay situations. The ideas of "beating" numbers does not even impact this in any way shape or form.
 
Cancelling your order over this review is silly IMO. The 6800U is on par with the X800XT and the 6800GT is not far behind. The X800PRO isn't really in the same class as either. There might be a few exceptions, perhaps Nvidia is not as good at 16xAF as ATI, but is just as fast using 8xAF? Why would you cancel an order based on a single site's benchmark of 3 games using only one setting which did not even compare multiple AA/AF levels? Not to mention that in the game that the 6800 lost, a performance enhancing SM3.0 patch will soon be made available that will not be made available for the X800 series.

If you look at the review trends, which mean the most because they give you a wide spectrum of performance across different games, its like this with the new drivers:

Trends show:

Performance
6800U = X800XT > 6800GT >> X800PRO >>> 6800 Standard

Features
6800U = 6800GT = 6800 Standard >> X800XT = X800PRO

Trends are much more powerful than a single review, because it gives you a wide range of different measures of performance across a spectrum of games and settings. They have shown that the two top cards trade blows in games, that the 6800GT owns the $399 price point, and that the 6800 series has a more desirable featureset than the X800 series. Nothing has changed, and the 6800 series is still on top overall with the new drivers IMO. 3dmark03 scores actually reflect the above performance equation, which is pretty cool.

If you cancelled your order over this review, I'm not sure you've done enough research and probably passed up on a good opportunity. Then again, you might save some cash by waiting.
 
phas3d said:
Brent and Kyle, I can definitely see your point of view in only running varied tests to establish performance indicative of "real world" game play, but there is a segment of the gaming population out there that buys video cards for the numbers (whether they admit it to themselves or not). Regardless of what you may think about people who run benchmarks for the pure score and score alone, I feel you are really missing out on a niche of the video card market when you aren't testing using "Established" benchmarks. Indeed, many around these grounds just want to focus on the trees.

Your review has made me reconsider my position on my purchase of the 6800GT. I have cancelled my order with Outpost.com.


phas3d
Click to expand...

We have shifted our focus to what we think will be a bigger help to more people that want to play games, not the ones worrying about benchmarks. I don't see any reason to spend tons of resources to show you the exact same canned benchmark that another site has so that it make someone happy. We have better things to do with our time that we feel will have a broader impact on the entire market when it is said and done. :)

Thanks for having faith in us!!
 
davidj said:
Brent-

Yes Ati's web site does mention the Platinum Edition & Nvidia site does not mention the EE.

Now we could debate this as a minor technicality(the boards do exist), we won't. I know you must operate within specific, set guidelines. We are not looking for fault, just making suggestions. I'm glad that you are secure enough to at least consider them.

Looking forward to the next one.

Thanks!

davidj
Click to expand...

To my knowledge you will never see a non-PE retail boxed X800XT from ATI.
 
Brent_Justice said:
We are DEFINITELY going to add that genre of games to our mix.

In your opinion which is the most taxing?

City of Heroes pretty good?
Click to expand...

I'm a HUGE fan of this genre and play everything....I highly suggest CoH's....excellent choice. It's not taxing but it is very popular and a great review "bed." Taxing....SWG, or WoW...this is my opinion of course on a 9500 and a Ti500. Will let you know how it performs on the 6800u when I install it. :p
 
Brent_Justice said:
Originally Posted by Brent_Justice
FYI: Recorded timedemos in FarCry do NOT playback the AI, there is no shooting, no damage being taken, nothing, it is like a flyby.

Of course performance is going to be better in a recorded timedemo. It does not accuretly reflect gameplay.

Our real world gameplay evaluations are manual run-throughs in God mode taking on all the gun firring, all the AI reacting, and the visual damage that occurs to the player.

Basically our run-throughs show you real world game performance that the cards deliver.
Click to expand...

Brent_Justice said:
quoted again just incase anyone missed it at the bottom of the first page
Click to expand...

You know it will go un-noticed purely by choice ;) People only read/hear what they want to read/hear. It is like being in denial
 
Blackwind said:
I'm a HUGE fan of this genre and play everything....I highly suggest CoH's....excellent choice. It's not taxing but it is very popular and a great review "bed." Taxing....SWG, or WoW...this is my opinion of course on a 9500 and a Ti500. Will let you know how it performs on the 6800u when I install it. :p
Click to expand...

How can you benchmark a MMORPG? Wouldn't you have to have a large group of people stand in the same place with the exact same spells on to ensure accurate repeatable results?
 
Blackwind said:
I'm a HUGE fan of this genre and play everything....I highly suggest CoH's....excellent choice. It's not taxing but it is very popular and a great review "bed." Taxing....SWG, or WoW...this is my opinion of course on a 9500 and a Ti500. Will let you know how it performs on the 6800u when I install it. :p
Click to expand...


City of heros requires a faster cpu.

I was running a xp1600, ran like crap. upgraded to a 2800, ran great. Same memory. Same shitty ti4200 (hah).

More objects = more shit to calculate.
 
Perhaps [H]'s reviews do not apply to everyone, but they definitely apply to me. I'm not buying a ridiculously expensive video card, or any decent video card for that matter, to play a benchmark at a general setting, only caring about maximum FPS or even average FPS. I like to have a playable minimum, average, and maximum FPS experience at high resolution and top settings. Knowing which cards are "playable" at what settings is most important, especially in combination with a CPU.

I want to know how they perform where I'll be using them. I play at 1280 or 1600 preferably, with AA/AF sustained, and in real games with real gameplay situations. [H]'s reviews sum that up for me nicely which few other reviews do...trust me I've wasted hours filtering through card reviews trying to find results that actually mean something at my gaming preferences. Basically, [H]'s reviews are based from the viewpoint of the gamer, rather than just a tester or benchmarker; I think that's a critical difference to note...and a positive trend.

I don't intend to take this thread off-topic, I just want to state my viewpoint. As many have said, there are tons of reviews out there that cater to the needs of benchmarking alone; this place is for more mature reviews. I found this particular review to be informative, as well.
 
theelviscerator said:
I understand, I like apples to apples comparos myself, I am talking about tests you can duplicate at least such as benchemall
Click to expand...
Once you standardize on a test that everyone can use, you are just asking ATI & NVIDIA to try and cheat/optimize for that test to the detriment of everything else.
 
I really like hardocp's review methodology. I can only hope other sites will follow suite. You guys work hard to show the readers what kind of performance to expect when actually playing the games. Keep at it.
 
MrHappyGoLucky said:
You know it will go un-noticed purely by choice ;) People only read/hear what they want to read/hear. It is like being in denial
Click to expand...

I am sort of getting that feeling as ANYONE who would read that should understand that a FarCry demo and FarCry real gameplay are not going to look anything alike statistically.
 
theelviscerator said:
I understand, I like apples to apples comparos myself, I am talking about tests you can duplicate at least such as benchemall

I wish you would have thrown in a oc'd system as well though.....since many of us are operating far past retail speeds....
with your settings 1280 and 2/8 AA AF I run 60 to 65 frames consistent in far cry.. with my cpu and oc'd card running at 540/540...
Click to expand...

We did have an OCed system operating far past its retail speed. We did not have one on the Intel side. If you are already running a CPU far beyond retail speeds, what are you going to do, slow it down? Obviously game play was moving forward with CPU scaling. I am sorry that we did not take the time to find out the top of the scale but quite frankly I dont think that is going to impact many folks or really change their purchasing decision.
 
Vagrant Zero said:
Enlighten me. No seriously, I have no idea how you did it and how you percieve he did it so you're going to have to elaborate a bit more than that.

But MORE importantly, since we're on the subject of REAL GAMEPLAY...would it even matter? His experiance was that at 1600x1200 8xAA 16xAF max settings no ops it was "somewhat playable.. At some parts with action it dips into the teens" [his words]. I think knowing what you know about Nvidia's 8xAA mode you'd allow us to extrapolate the assumption that at 1600x1200 4xAA 16xAF max settings ops on Far Cry would be playable by your standards.

So this question is begged, if your article shows the max playable setting to be 1280x1024 2xAA 8xAF...what's going on?
Click to expand...

I'm trying to figure out if you are mad because a website found different results than what is posted on one thread at nvnews forum or if it was the X800Pro keeping up or beating the 6800U.
 
Maxx said:
Perhaps [H]'s reviews do not apply to everyone, but they definitely apply to me. I'm not buying a ridiculously expensive video card, or any decent video card for that matter, to play a benchmark at a general setting, only caring about maximum FPS or even average FPS. I like to have a playable minimum, average, and maximum FPS experience at high resolution and top settings. Knowing which cards are "playable" at what settings is most important, especially in combination with a CPU.

I want to know how they perform where I'll be using them. I play at 1280 or 1600 preferably, with AA/AF sustained, and in real games with real gameplay situations. [H]'s reviews sum that up for me nicely which few other reviews do...trust me I've wasted hours filtering through card reviews trying to find results that actually mean something at my gaming preferences. Basically, [H]'s reviews are based from the viewpoint of the gamer, rather than just a tester or benchmarker; I think that's a critical difference to note...and a positive trend.

I don't intend to take this thread off-topic, I just want to state my viewpoint. As many have said, there are tons of reviews out there that cater to the needs of benchmarking alone; this place is for more mature reviews. I found this particular review to be informative, as well.
Click to expand...

Many thanks for letting us know we are hitting the target we are aiming at.
 
I don't think it silly to have cancelled my order for the BFG 6800GT after this review. I didn't say I have opted for one of the ATI cards. I was debating keeping my order active and this review pushed me over the top. Now it's time to sit back, take in ALL the reviews from ALL the most frequented sites and ponder my choice. I just want to make sure I get every dime of my money's worth out of my video card purchase.


phas3d
 
DocSavage said:
How can you benchmark a MMORPG? Wouldn't you have to have a large group of people stand in the same place with the exact same spells on to ensure accurate repeatable results?
Click to expand...

Obviously you don't play MMORPG's. :D Can you benchmark in the traditional sense? No. But you can however benchmark easily with [H] present methods of playability within actual game play.

Raids in EQ come to mind.
 
I just wish there was a wider range of frame rates. 30-40 fps average and dipping down into the 20's isn't acceptable to me.
 
Thanks for the replys Kyle and Brent.

Interesting-
Nvidia does not Officially recognize their fastest boards, but Ati does. Even if companies like BFG and e-VGA are licensed/allowed to change things.

I guess it does not really matter. If I have a problem with a vc, I will deal with BFG directly and they have a true/genuine life time waranty.

Question-
Does Ati or Sapphire etc, offer a true/genuine life time waranty on their boards?
 
MrHappyGoLucky said:
I'm trying to figure out if you are mad because a website found different results than what is posted on one thread at nvnews forum or if it was the X800Pro keeping up or beating the 6800U.
Click to expand...

I'm not mad. You've seen me mad before don't you remember. I typically begin making snide comments about your intelligence when I'm mad. Since I have not done that [yet] I am not mad. I am merely curious. HOCPs 'numbers' are not matching up to other websites 'numbers' and their 'REAL GAMEPLAY' isn't matching up people who have these cards are saying is their 'REAL GAMEPLAY' [which is my main issue, if they're 'numbers' are not representive of actual gameplay experiance as they don't seem to be than they should be held forfeit].

So I'd like to know what's up, but so far my questions have been completely ignored which does nothing to assure me about the opppositions credibilty [though I have been given a few blanket statements such as 'but they're DIFFERENT...as in NOT 100.000000% the EXACT same pixel-by-pixel comparision to which I can say...no duh...but the comparisions are close enough to call the performance discrepancies into question]. It's quite simple really, if I am so wrong, it should be very easy for someone to come along and prove it so without much effort, yet no one has bothered to even make mention of my statements. As such, you'll have to forgive me for not blindly following the reports of a single entity. I like to keep my options open. In the future I shall endeaver to be as fanatical as yourself if that would please you.
 
Vagrant Zero said:
I'm not mad. You've seen me mad before don't you remember. I typically begin making snide comments about your intelligence when I'm mad. Since I have not done that [yet] I am not mad. I am merely curious. HOCPs 'numbers' are not matching up to other websites 'numbers' and their 'REAL GAMEPLAY' isn't matching up people who have these cards are saying is their 'REAL GAMEPLAY' [which is my main issue, if they're 'numbers' are not representive of actual gameplay experiance as they don't seem to be than they should be held forfeit].

So I'd like to know what's up, but so far my questions have been completely ignored which does nothing to assure me about the opppositions credibilty [though I have been given a few blanket statements such as 'but they're DIFFERENT...as in NOT 100.000000% the EXACT same pixel-by-pixel comparision to which I can say...no duh...but the comparisions are close enough to call the performance discrepancies into question]. It's quite simple really, if I am so wrong, it should be very easy for someone to come along and prove it so without much effort, yet no one has bothered to even make mention of my statements. As such, you'll have to forgive me for not blindly following the reports of a single entity. I like to keep my options open. In the future I shall endeaver to be as fanatical as yourself if that would please you.
Click to expand...

I feel very confident that our numbers are dead on where they should be. As we have explained many times over we pick sections of the games that we feel show to be a "worst case possible scenario" when it comes to performance. In other words we don't pick the easiest level for the video card, we try to pick the hardest. Our thoughts there are that in this we will not be giving anyone unrealistic goals for their games and home and by definition of what we are doing SHOULD get even or better frame rates than us on ANY card tested in a similar setup.

So it is my thought that you are seeing exactly what you should see. :)
 
Really Kyle, it all makes sense to me. Some people just aren't happy with a present, no matter how pretty and wrapped it may be if it is missing a bow. :D

phas3d
 
Good review...although I would have liked to see a barton running 2.4-2.6GHz with like 220-230 FSB. It's safe to say a lot of us have rigs like that.
 
DaveX said:
Good review...although I would have liked to see a barton running 2.4-2.6GHz with like 220-230 FSB. It's safe to say a lot of us have rigs like that.
Click to expand...

Only so many hours in a day. This article took a HUGE amount of work as it is. With that said, I don't think I have ever written a "complete" article in everyones eyes. I did not think we NEEDED those numbers in there as the folks that are running those machines are very aware of how they scale and the XP numbers there should be a base in which the XP OCer can project from.
 
You must log in or register to reply here.
Back
Top