chavvdarrr
Veteran
Well, that's his article. I'm talking about something that's been made by ixbt guys (UncleSam).Hanners said:
The script was available, I'm sure some guys from beyond3d have it for week now .....
My thread @ Futuremark(Re =Waite for Unwinder)
- Thread starter micron
- Start date
The script was available, I'm sure some guys from beyond3d have it for week now .....
I was pleasantly surprised by the digit-life article. Has Unwinder ever written an article there before? I was expecting more distortion and some gaping incompleteness, but about all relevant factors, within the choice of focus, were addressed...I'm thinking because of Unwinder's preference for technical discussion and deep familiarity.
The article does lend itself to equating ATI and nVidia, again, but a quite reasonable set of relevant factors are presented. The lack of emphasis of some of the factors that relate to ATI apparently not having trangressed as much, just seems to be the product of a bland English rendering and a focus on one single benchmark, which doesn't seem something that can fairly be held against them unless it is the extent of coverage on the issue. I do think it is the best article I recall seeing there, with a very clear and seemingly accurate level of detail presented on its focus.
That said, I look forward to ATI's response the clearly laid out issues. If you're wondering why I don't list nVidia: both nVidia and ATI could easily disappoint me with their responses, but with ATI atleast I can look forward to learning something new from that...maybe nVidia will surprise me with an excellent answer (by my criteria, which includes things like a modicum of honesty and factual correlation), but I can't reasonably look forward to something when I have such strong reason to expect the opposite.
The article does lend itself to equating ATI and nVidia, again, but a quite reasonable set of relevant factors are presented. The lack of emphasis of some of the factors that relate to ATI apparently not having trangressed as much, just seems to be the product of a bland English rendering and a focus on one single benchmark, which doesn't seem something that can fairly be held against them unless it is the extent of coverage on the issue. I do think it is the best article I recall seeing there, with a very clear and seemingly accurate level of detail presented on its focus.
That said, I look forward to ATI's response the clearly laid out issues. If you're wondering why I don't list nVidia: both nVidia and ATI could easily disappoint me with their responses, but with ATI atleast I can look forward to learning something new from that...maybe nVidia will surprise me with an excellent answer (by my criteria, which includes things like a modicum of honesty and factual correlation), but I can't reasonably look forward to something when I have such strong reason to expect the opposite.
So surprised by the solid facts, I neglected the conclusion
Hmm...the conclusion does seem to be based rather strongly on the idea of absolute equivalence for both ATI and nVidia, when outside of the context of 3dmark 2001, on which the article focuses completely, that isn't in fact established at all. That is indeed a rather significant distortion when such far reaching expectations for equivalent behavior from both is promoted as the expectation, when the differentiation between the company outside of the selected focus of the article appears to be rather contradictory to that.
"In the scandal around Quake/Quack NVIDIA stated that they never used optimizations for certain applications and that they advocated only general approaches for optimizations." A pretty short, direct, and accurate statement, even understating that nVidia provided the tools to expose Quack and tried to hide the action.
"In the disputes around the 3DMark2003 ATI took a position of a hurt child and was inconsistent in its actions: first it repented and promised to remove its optimizations saying that their programmers never used any rendering tricks to get higher scores in benchmarks, then they promised to optimize the drivers to catch up with the impious competitor." By contrast, this statement is just rife with emotional weighting, and ends by criticizing the idea of "optimizations" as if optimizations and cheating are the same thing. Ack!
The concept of "tenfold" more cheats for nVidia observed earlier seems completely forgotten, and completely contradictory to the choice of representation of the companies when trying to leave an impression of them at the end.
I really should have checked for a disconnect between the excellent factual support for the main part, and the speculative factors in the conclusion, but the facts were discussed so well I mistakenly assumed the conclusion would have stuck to a reasonable correlation. The conclusion is more usual of what the site typically does wrong, and it isn't the first time solid factual data suddenly shifted to a definitive conclusion not supported by the data...I don't know why I'm surprised, except that I glanced down to the byline and noticed Unwinder's name.
Hmm...the conclusion does seem to be based rather strongly on the idea of absolute equivalence for both ATI and nVidia, when outside of the context of 3dmark 2001, on which the article focuses completely, that isn't in fact established at all. That is indeed a rather significant distortion when such far reaching expectations for equivalent behavior from both is promoted as the expectation, when the differentiation between the company outside of the selected focus of the article appears to be rather contradictory to that.
"In the scandal around Quake/Quack NVIDIA stated that they never used optimizations for certain applications and that they advocated only general approaches for optimizations." A pretty short, direct, and accurate statement, even understating that nVidia provided the tools to expose Quack and tried to hide the action.
"In the disputes around the 3DMark2003 ATI took a position of a hurt child and was inconsistent in its actions: first it repented and promised to remove its optimizations saying that their programmers never used any rendering tricks to get higher scores in benchmarks, then they promised to optimize the drivers to catch up with the impious competitor." By contrast, this statement is just rife with emotional weighting, and ends by criticizing the idea of "optimizations" as if optimizations and cheating are the same thing. Ack!
The concept of "tenfold" more cheats for nVidia observed earlier seems completely forgotten, and completely contradictory to the choice of representation of the companies when trying to leave an impression of them at the end.
I really should have checked for a disconnect between the excellent factual support for the main part, and the speculative factors in the conclusion, but the facts were discussed so well I mistakenly assumed the conclusion would have stuck to a reasonable correlation. The conclusion is more usual of what the site typically does wrong, and it isn't the first time solid factual data suddenly shifted to a definitive conclusion not supported by the data...I don't know why I'm surprised, except that I glanced down to the byline and noticed Unwinder's name.
demalion
Think that discussing r they equivalent or not is just waste of time.
This is the job of PR departments to prove that vendor X (where X=NV, ATI...) is "knight in the shining armor", isn't it?
There r several evidences that with antidetect script performance is dropping not only in 3dmarks.
It's time to think about industry.
imho
Think that discussing r they equivalent or not is just waste of time.
This is the job of PR departments to prove that vendor X (where X=NV, ATI...) is "knight in the shining armor", isn't it?
There r several evidences that with antidetect script performance is dropping not only in 3dmarks.
It's time to think about industry.
imho
Re: So surprised by the solid facts, I neglected the conclus
First, it's the second of my articles published on Digit-Life (the first was about investigating Intellisample AF optimization techniques in Det 43.45/OGL ). The rest investigation aricles are available on www.nvworld.ru. So there is no need to tell about wrong conclusions typical for DL. It's _my_ conclution, I can agrue it and nothing but NV/ATI comments will change it. Unfortunately ATI already had a chance to comment 3DMark2001 issues and they missed it. I'm afraid that only the ppl who can see "Bad ATI, good NVIDIA" or "Good NVIDIA, bad ATI" may say that the conclusion is wrong.
Second, the article is really rather emotional. I do feel a lot of frustration after seeng such shame in both drivers. I was not surprised to see a lot of bad things in NV drivers, but it's really sad that ATI benchmark results are also distorted by the driver. I _respected_ their reaction on 3DMark2003 and I really trusted them. My ATI related comment which you've quoted is based upon public CATALYST Maker's statement in this thread:
http://www.rage3d.com/board/showthread.php?s=&threadid=33689928&perpage=20&pagenumber=11
Quoting:
I would _really_ like to make different conclusions from such statement and investigation results, but I see nothing but lying in this statement. Nothing more, nothing less. Are you still trying to assure me that the conclusion is wrong? My position is plain and simple: if you're trying to play good guys, do it completely or don't do it at all.
Finally, it looks like you missed the main idea of this investigation. It's clearly stated in the last paragraph.
RuslanK
73?
demalion said:
First, it's the second of my articles published on Digit-Life (the first was about investigating Intellisample AF optimization techniques in Det 43.45/OGL ). The rest investigation aricles are available on www.nvworld.ru. So there is no need to tell about wrong conclusions typical for DL. It's _my_ conclution, I can agrue it and nothing but NV/ATI comments will change it. Unfortunately ATI already had a chance to comment 3DMark2001 issues and they missed it. I'm afraid that only the ppl who can see "Bad ATI, good NVIDIA" or "Good NVIDIA, bad ATI" may say that the conclusion is wrong.
Second, the article is really rather emotional. I do feel a lot of frustration after seeng such shame in both drivers. I was not surprised to see a lot of bad things in NV drivers, but it's really sad that ATI benchmark results are also distorted by the driver. I _respected_ their reaction on 3DMark2003 and I really trusted them. My ATI related comment which you've quoted is based upon public CATALYST Maker's statement in this thread:
http://www.rage3d.com/board/showthread.php?s=&threadid=33689928&perpage=20&pagenumber=11
Quoting:
I would _really_ like to make different conclusions from such statement and investigation results, but I see nothing but lying in this statement. Nothing more, nothing less. Are you still trying to assure me that the conclusion is wrong? My position is plain and simple: if you're trying to play good guys, do it completely or don't do it at all.
Finally, it looks like you missed the main idea of this investigation. It's clearly stated in the last paragraph.
RuslanK
73?
OpenGL guy
Veteran
Re: So surprised by the solid facts, I neglected the conclus
Look at it from ATI's point of view: A couple companies put in optimizations for 3D Mark 2003, some valid, some not. These optimizations are revealed to the public. ATI agrees to remove said optimizations. Other company says nothing but "They're trying to make our products look bad." Weeks later, ATI has lived up to it's promise, other company still says nothing, but instead releases a new driver will all the "optimizations" reenabled.
What the general public thinks matters to me personally as I visit several forums and want people to take me seriously. But OEMs have much bigger influence with IHVs because they mean big $$, and $$ is what makes a business successful. If OEMs don't see what that other company is doing as wrong, then ATI must follow suit. That was Catalyst Maker's point. Either the playing field is level, or it's not, and it's my job to level that playing field as needed. We don't like taking shortcuts, as it goes against the very principle of high-end graphics, but if that's what it takes to keep OEMs happy, then so be it. So far, OEMs have been responsive to recent developments and haven't pushed us down a path I don't want to take. As long as things stay that way, then I am happy and you folks should be as well.
Now, on a happier note, I, and my coworkers, are pushing for some more AF options so the end-user has more control over quality and performance (no pun intended).
I'll get off my soap box now...
Your findings imply a single benchmark in ATI's case.Unwinder said:
And did we not do as we said we would? The optimizations for 3D Mark 2003 are gone.
Where is the lie? In fact, there is not one lie in the whole statement. If you don't like Catalyst Maker's comment, that's fine, you're entitled to your opinion, but it's a gross misstatement to call anything he said a lie.
It looks to me like you missed the main idea of Catalyst Maker's comments as well, so we're all entitled to make mistakes.
Look at it from ATI's point of view: A couple companies put in optimizations for 3D Mark 2003, some valid, some not. These optimizations are revealed to the public. ATI agrees to remove said optimizations. Other company says nothing but "They're trying to make our products look bad." Weeks later, ATI has lived up to it's promise, other company still says nothing, but instead releases a new driver will all the "optimizations" reenabled.
What the general public thinks matters to me personally as I visit several forums and want people to take me seriously. But OEMs have much bigger influence with IHVs because they mean big $$, and $$ is what makes a business successful. If OEMs don't see what that other company is doing as wrong, then ATI must follow suit. That was Catalyst Maker's point. Either the playing field is level, or it's not, and it's my job to level that playing field as needed. We don't like taking shortcuts, as it goes against the very principle of high-end graphics, but if that's what it takes to keep OEMs happy, then so be it. So far, OEMs have been responsive to recent developments and haven't pushed us down a path I don't want to take. As long as things stay that way, then I am happy and you folks should be as well.
Now, on a happier note, I, and my coworkers, are pushing for some more AF options so the end-user has more control over quality and performance (no pun intended
).I'll get off my soap box now...
Re: So surprised by the solid facts, I neglected the conclus
Are you suggesting that there will now be three options? With a Full Trilinear being the highest and the current 'Quality' being a middle setting?
OpenGL guy said:Now, on a happier note, I, and my coworkers, are pushing for some more AF options so the end-user has more control over quality and performance (no pun intended
Are you suggesting that there will now be three options? With a Full Trilinear being the highest and the current 'Quality' being a middle setting?
OpenGL guy
Veteran
Re: So surprised by the solid facts, I neglected the conclus
The big problem is balancing the competitive situation...
I can't comment on the details because it's not set in stone yet, but I will say that things will only get better in terms of quality if the changes are made.DaveBaumann said:Are you suggesting that there will now be three options? With a Full Trilinear being the highest and the current 'Quality' being a middle setting?OpenGL guy said:Now, on a happier note, I, and my coworkers, are pushing for some more AF options so the end-user has more control over quality and performance (no pun intended
The big problem is balancing the competitive situation...
Re: So surprised by the solid facts, I neglected the conclus
Where did he suggest that there will be 3 options?
Didn't he say "pushing for some more AF options"?
So I wander where did you get the number 3 from?
DaveBaumann said:OpenGL guy said:Now, on a happier note, I, and my coworkers, are pushing for some more AF options so the end-user has more control over quality and performance (no pun intended
Are you suggesting that there will now be three options? With a Full Trilinear being the highest and the current 'Quality' being a middle setting?
Where did he suggest that there will be 3 options?
Didn't he say "pushing for some more AF options"?
So I wander where did you get the number 3 from?
Yes, but 3 would make sense in the context of recent discussions. Presently there is "Performance" which relates to full Bilinear only and "Quality", which relates to mixed Bi and Triliear. A logical progression from that would be to add a further mode that enabled full Trilinear filtering (although there are certainly other prgressions as well).
OpenGL guy
Veteran
*shrug* I don't make the decision on what gets put into the driver and what doesn't. Plus, this has to be coordinated with several groups so is non-trivial. We've made a good case for it, now it's just a matter of waiting to see what the decision is.K.I.L.E.R said:
Re: So surprised by the solid facts, I neglected the conclus
OpenGL guy
I perfectly understand ATI and NV positions in benchmark 'optimization', but it doesn't mean that I like it. Money rule the world...
But _please_ read the last paragraph of the article once again.
I bet that you're talking about NV-styled per-stage AF control.
OpenGL guy
I perfectly understand ATI and NV positions in benchmark 'optimization', but it doesn't mean that I like it. Money rule the world...
But _please_ read the last paragraph of the article once again.
I bet that you're talking about NV-styled per-stage AF control.
Re: So surprised by the solid facts, I neglected the conclus
I'd be surprised if the hardware could do that.
Unwinder said:I bet that you're talking about NV-styled per-stage AF control.
I'd be surprised if the hardware could do that.
