HardOCP Conroe Gaming Benchmarks

[vfrex]

http://enthusiast.hardocp.com/article.html?art=MTEwOCwxLCxobmV3cw==

It looks like I was correct :twisted:

[XJnine]

Looks like they jumped the gun a little bit. The article is dated tomorrow and even here on the right coast it's still well before midnight.

Interesting so far, I'll have to read a bit more thoroughly.

[Kougar]

Very interesting article. They/you do have a point in that the CPU doesn't make any difference when solidly GPU bound, except for Oblivion anyway. However I'd seriously wonder if they had used a 7800GTX SLI or x1900XT Crossfire test system instead how much of a difference they would have seen, I suspect there will be more than a few hundreths of a decimal place in difference. Or even a single x1900XT card would have been interesting to compare to these numbers...

In particular how the E6700 showed only a neglible difference from the x6800, while being half the price, and still "topping" the FX62... When the Intel system was using the midrange 965 chipset against the high-end nForce 590SLI chipset. I guess this article would be a very good example to why you should use SLI/Crossfire if you're buying a $500 or $1k processor. :roll:

[werty316]

Heres the rest:

Intel Core 2 Music, Images, & Movie Performance (http://enthusiast.hardocp.com/article.html?art=MTExMCwxLCxoZW50aHVzaWFzdA==)

Intel Core 2 Duo & Extreme Power Consumption (http://enthusiast.hardocp.com/article.html?art=MTExMSwxLCxoZW50aHVzaWFzdA==)

Intel Core 2 Desktop Pricing and Specifications (http://enthusiast.hardocp.com/article.html?art=MTEwNiwxLCxoZW50aHVzaWFzdA==)

I think its safe it say the Core2Duo owns but then again we all knew this when the previews came out so no big surprise. Waiting for benches of the lower end Core2Duo chips.

Here is another review from TrustedReviews (http://www.trustedreviews.com/article.aspx?art=3161)

[Kougar]

I think it's only fair to balance those GPU limited tests with some CPU limited benchmarks... http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795&p=14 (http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795&p=14) ;)

[vfrex]

Right, CPU limited benchmarks show that the Conroe is faster under those circumstances. So, I guess this just reiterates what we discussed before. You need to spend 600-1000 USD on a graphics system to avoid a graphics bottleneck today. In fact, even with a high end graphics card system, it has been said that conroe is still bottlenecked.

I'm happy with my current position of buying behind the cutting edge. By the time there is a reasonably priced graphics solution that can keep up with Conroe, Conroe's price will have dropped considerably.

[XJnine]

One thing that is driving me nuts is HardOCP's holier than thou attitude with their reviews. They act like their “real world gameplay” way of benchmarking is the only way to do it and if you do it any other way you are a complete idiot. I personally don’t like their style of reviews but I understand that some people do. I just wish they would stop acting like they are the only ones reviewing things the right way and everyone else is just f’ing up. All they're showing is how the hardware would run in a configuration they happen to like at a performance level they happen to be comfortable with. Their benchmarks do nothing to show how you may be future-proofing youself by buying one of these new cpu's because they don't factor the GPU limitations out of the equation. It may show how things could run today but I don't buy a CPU just for today, I buy it for years. I would much rather see benchmarks at varying resolutions, AA levels, AF levels, and other game options so I can see the performance difference between it all and I can decide for myself what combination of the above I would want to run at.

I better turn my rant mode off, I could keep on going if I dont....

[vfrex]

To be fair, how many sites are just posting the raw benchmarks you're talking about? And how many are giving you HardOCP's perspective? You can easily get the raw benchmarks anywhere.

[Scott]

Real World gaming systems my ass. They do not know what that is. How many of you would game on the net without Antivirus and a Firewall progam. And do you really want to shut down your IM client, InCD and other tools just to play a game?

By the way, we do all our reviews on real systems with all that stuff installed.

[vfrex]

How many of you would game on the net without Antivirus and a Firewall progam. And do you really want to shut down your IM client, InCD and other tools just to play a game?

I'm not saying all of the above, but I certainly trim down if I'm playing a game that my computer has some trouble with.

[XJnine]

The thing about HardOCP's perspective is it's just one resolution and set of game settings that they happen to like. That's all they're giving. They're making my game visuals and performance decisions for me and telling me what I should think of the hardware based on their gaming preferences. Show me all the data and let me draw my own conclusions. Let me decide if I prefer the performance hit coming from AA vs AF. Maybe I want raw framerates instead and I want to turn them off. If I go up a higher resolution and drop some AA and AF what's the framerate impact? Maybe visually and performance wise that would be a benefit for me.

Their reviews are worth exactly what I paid for them.... ;-) It's still a ton of hard work and interesting reading. I just don't like the way they act like it's the way everyone should be benchmarking. It's only one way and no way is necessarily the right way.

Scott hit it on the head too with his comment also. They may be doing real world game play settings (debatable) but they are still using an optimized system and not running it like you or I would at home. Bjorn3D does their reviews on their actual personal use systems. That's real world gameplay.

[Kougar]

Well my beef with HardOCP really stems from those power consumption tests they ran. As far as I know, there is no reason behind the explanation they gave for intentionally skewing the results. While those powersaving functions would need to be disabled for overclocking, they were obviously not testing for overclocking... so why disable them? It reeks to much of them intentionally trying to skew the results towards AMD, as those are the worst figures I've seen anywhere at any stage in the process.

Going by their rationale, if the Intel system had been on a faster clock than the AMD system I bet they would have underclocked it to "balance" their results. :roll:

[XJnine]

They should have run the tests with the power saving features on and then again with them off to show you both levels of power consumption and to show you how well the power saving features work.

They should also have done that with both AMD and the Intel cpu's.

[Kougar]

Well, it be fair, I think they did at least disable AMD's CnQ feature.

But the values for their test are completely useless... since they are not a real life scenario, nor are they overclocking! I can see them running this test out of curiosity, but there is no grounding for them to use these numbers in place of having the power saving features enabled!! And people accuse THG of being biased... good grief. I have not even given HardOCP's third article a glance yet. :roll:

[vfrex]

One thing that the author/admin stated was that Intel cut their testing time in half from what they had originally anticipated. As HardOCP's testing seems to be more involved and time consuming, it is hard to blame them for not showing every possible combination.

[Kougar]

One thing that the author/admin stated was that Intel cut their testing time in half from what they had originally anticipated. As HardOCP's testing seems to be more involved and time consuming, it is hard to blame them for not showing every possible combination.

That is not the point here. The point is instead of testing for real life numbers, the instead tested for and listed completely superfluous numbers that any reader that was only skimming the article would get the completely wrong idea from. It would have taken them less time to test this correctly, so your arguement is baseless.

[vfrex]

...that any reader that was only skimming the article would get the completely wrong idea from

Why is that point any more valid than mine? Why should I care what people skimming the article think? The words aren't just there for show.

rephrased: why should I care if people misinterpret the article because they skip/ignore the context?

[Kougar]

Why is that point any more valid than mine? Why should I care what people skimming the article think? The words aren't just there for show.

Since you are only nitpicking details of my arguement, nitpick away then... Completely remove that sentence, and my point still stands. HardOCP appears to have skewed the results intentionally in AMD's favor or Intel's disfavor, no matter who reads the article. :???:

[vfrex]

Please, I'm not nitpicking your argument. I blew it away. You are suggesting that it is OK to get your news by looking at the pictures and captions.

[Kougar]

Please, I'm not nitpicking your argument. I blew it away. You are suggesting that it is OK to get your news by looking at the pictures and captions.

No, I wasn't suggesting that, but you completely making that up and trying to turn that into the main arguement/topic speaks volumes.

Since having a discussion with you is quite pointless, you are welcome to make up all the fictitous crap in your threads that you want. :)

[vfrex]

Since having a discussion with you is quite pointless, you are welcome to make up all the fictitous crap in your threads that you want.

I really don't know what you are talking about.

No, I wasn't suggesting that, but you completely making that up and trying to turn that into the main arguement/topic speaks volumes.

Why don't you summarize the point you are trying to make? Are you saying that benchmarks which only use a single high end graphics card are not realistic gameplay for the majority?

[werty316]

I think you guys are thinking too far into this about talking how the review was done. The whole point was to test new hardware and how it performs compared to current hardware.

Speaking of Speedstep and Cool n' Quiet:
I have updated our Intel Core 2 Power Consumption (http://enthusiast.hardocp.com/article.html?art=MTExMSwxLCxoZW50aHVzaWFzdA==) article to include Intel Speedstep and AMD Cool n' Quiet technologies. Some of you suggested that we were hiding something, but as usual, the results speak for themselves and we still stand by our previous conclusions.

Intel has fixed their broken power consumption legacy and the Core 2 now fits inside a power envelope that will give it a very competitive performance per watt position. At the same time, however, I would highly suggest keeping your eyes on AMD’s low wattage / energy efficient processors for those projects that require a noiseless solution.

[Kougar]

Why don't you summarize the point you are trying to make? Are you saying that benchmarks which only use a single high end graphics card are not realistic gameplay for the majority?

I summarized my point twice and you completely ignored it, instead choosing to take something completely out of context from within the post and make wild assumptions from it. Just like you did again above. Keep on digging, vFrex. :mrgreen:


Werty, I probably am reading to much into it, yes. While the review is much better now, and the numbers aren't quite as off as I thought they were, they still seems AMD biased.

In regards to their graphics bound testing, I think the premise behind it is flawed as anyone that would be buying $1000 and $1200 processors would not be running a single $420 GPU. Most would be running a high-end Crossfire or SLI setup. As Anandtech's article demonstrate, it's a completely different ballgame once you take that into account.

[werty316]

Its like everyone says about reviewers, they are biased so if you want proper scores you'll have to run benchmarks yourself.

[vfrex]

In regards to their graphics bound testing, I think the premise behind it is flawed as anyone that would be buying $1000 and $1200 processors would not be running a single $420 GPU. Most would be running a high-end Crossfire or SLI setup. As Anandtech's article demonstrate, it's a completely different ballgame once you take that into account.

The E6700 is HALF of that. $530. You know that. It maxes out the GPU as well. The article works for me, because I wouldn't be buying SLI/Crossfire and I wouldn't be buying a $1000 processor. Every other article released what could be done with a $2500 budget. But how many people actually do have that?

I summarized my point twice and you completely ignored it, instead choosing to take something completely out of context from within the post and make wild assumptions from it. Just like you did again above. Keep on digging, vFrex.

I don't know what you are trying to say. Maybe if you could be more specific, I wouldn't have to try to determine what you are actually trying to say.

Going by their rationale, if the Intel system had been on a faster clock than the AMD system I bet they would have underclocked it to "balance" their results. <---- Flamebait

Well, it be fair, I think they did at least disable AMD's CnQ feature. <--- Flamebait

Since having a discussion with you is quite pointless, you are welcome to make up all the fictitous crap in your threads that you want.

How exactly do you want me to respond to that? I've had multiple discussions with you since joining this forum. Clearly, having a discussion with me isn't pointless, because you yourself have done it in the recent past. What have I made up? You stated that a person skimming the article would misunderstand it. I answered that statement, saying that a person just skimming the article deserves to misunderstand if he doesn't read at least the context behind the testing. I watched the thread on the HardOCP forum all morning, and watched person after person come to the thread and ask why the benchmark numbers were different from sites x, y, and z. Everything was included in the article writeup, but people didn't feel obligated to read it.