![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Pepsiennis
[H]ard|Gawd
- Joined
- Feb 5, 2007
- Messages
- 1,290
To Dan_D:
Thanks again for taking the time to explain to me, Dan. I do sincerely appreciate it! You're not the best person to ask simply because you didn't author the ATI review, but one thing I've noticed with performance GPUs is that performance seems to scale in terms of resolution differently between grenerations and architectures. So sometimes we see a GPU actually perform better at a higher resolution in a given game than a lower one. In terms of data you've seen in your work, does this happen between CPU's sharing a common architecture and equal cache, such as the E6600 and the X6800?
As far as managing heat production, I know what you mean. Chongqing is a furnace in the summer. For a few weeks last summer we were hotter than Iraq, to give you an idea. I imagine, perhaps in error, most people with an SFF for LANs and such would take the performance hit and step down one rung to get a card that didn't output so much heat.
But I am digressing into the power consumption/heat argument which doesn't have anything to do with motherboards. Right? I mean, the 680i's great power needs compared to other chipsets isn't a factor in this conversation.
Thanks again, Mr. Dobrowolski! By the way, off topic. We're getting awfully close to the 1333 fsb Conroe's, aren't we? Three different vendors here have told me Intel has stopped shipping the higher-end processors (here in China high-end is E6600-E6700'ish), and that the 1333 procs will take their place in a few weeks. It'll be interesting to see you work your magic with those and a board supporting them!
phide:
As for spelling appreciated: new keyboard, stiff keys. At least it's obvious I intended to spell it correctly, and if the largest fault to be found in my post is a slip of the finger, then that's not so bad. But, "it is what it is", as I'm sure some would say. As for asking some not to post, it's because I would have liked to keep the observation (not hypothesis, as you and others repeatedly try to twist) and discussion clearer. The other less suitable option is using Hardforum's PM feature like a five dollar whore, which is probably not what the [H] is paying for. It's obvious that some folks don't want a clear discussion of some issues, and thus are out to 'pee in their neighbor's well'. I really don't want to endorse, or put myself in the position of endorsing, a particular review, as people will start arguing about that review instead of discussing what I might have noticed or might be in error about. As for your comments about power consumption: the [H] goes to great pains to avoid reviewing based on people's pre-conceived assumptions. And quite frankly, I also didn't see an "Assumption Satisfaction" entry in any of their graphs, either. It's wrong to blame ATI for not fulfilling someone else's preconceived notions. Now, if ATI had come out beforehand and unequivocably promised the consumer they'd deliver the performance at less Nvidia's heat/power usage, then it would be a fair issue. Now if the gameplay numbers in the review "are what they are", then power consumption doesn't matter. There is a practice of some to count the numbers when the GTS scores a bench, but then dismiss them with an irrelevant (to the numbers) mention of power consumption/heat when the 2900XT scores a bench, which comes off as seeming to imply the gameplay numbers 'aren't what they are'. It's wrong to confuse the two seperate issues.
Thanks again for taking the time to explain to me, Dan. I do sincerely appreciate it! You're not the best person to ask simply because you didn't author the ATI review, but one thing I've noticed with performance GPUs is that performance seems to scale in terms of resolution differently between grenerations and architectures. So sometimes we see a GPU actually perform better at a higher resolution in a given game than a lower one. In terms of data you've seen in your work, does this happen between CPU's sharing a common architecture and equal cache, such as the E6600 and the X6800?
As far as managing heat production, I know what you mean. Chongqing is a furnace in the summer. For a few weeks last summer we were hotter than Iraq, to give you an idea. I imagine, perhaps in error, most people with an SFF for LANs and such would take the performance hit and step down one rung to get a card that didn't output so much heat.
But I am digressing into the power consumption/heat argument which doesn't have anything to do with motherboards. Right? I mean, the 680i's great power needs compared to other chipsets isn't a factor in this conversation.
Thanks again, Mr. Dobrowolski! By the way, off topic. We're getting awfully close to the 1333 fsb Conroe's, aren't we? Three different vendors here have told me Intel has stopped shipping the higher-end processors (here in China high-end is E6600-E6700'ish), and that the 1333 procs will take their place in a few weeks. It'll be interesting to see you work your magic with those and a board supporting them!
phide:
As for spelling appreciated: new keyboard, stiff keys. At least it's obvious I intended to spell it correctly, and if the largest fault to be found in my post is a slip of the finger, then that's not so bad. But, "it is what it is", as I'm sure some would say. As for asking some not to post, it's because I would have liked to keep the observation (not hypothesis, as you and others repeatedly try to twist) and discussion clearer. The other less suitable option is using Hardforum's PM feature like a five dollar whore, which is probably not what the [H] is paying for. It's obvious that some folks don't want a clear discussion of some issues, and thus are out to 'pee in their neighbor's well'. I really don't want to endorse, or put myself in the position of endorsing, a particular review, as people will start arguing about that review instead of discussing what I might have noticed or might be in error about. As for your comments about power consumption: the [H] goes to great pains to avoid reviewing based on people's pre-conceived assumptions. And quite frankly, I also didn't see an "Assumption Satisfaction" entry in any of their graphs, either. It's wrong to blame ATI for not fulfilling someone else's preconceived notions. Now, if ATI had come out beforehand and unequivocably promised the consumer they'd deliver the performance at less Nvidia's heat/power usage, then it would be a fair issue. Now if the gameplay numbers in the review "are what they are", then power consumption doesn't matter. There is a practice of some to count the numbers when the GTS scores a bench, but then dismiss them with an irrelevant (to the numbers) mention of power consumption/heat when the 2900XT scores a bench, which comes off as seeming to imply the gameplay numbers 'aren't what they are'. It's wrong to confuse the two seperate issues.
