Radeon 9800 se -> 9800

[incurable]

I will change my point of view when provided with a coherent case for doing so. You have not provided one. By what criteria are you exempt from this step?

I thought I'd understood you, but apparently my understanding was in error. You aren't "agreeing to disagree", you're "agreeing not to pay attention to the reasons for an opposing viewpoint". As I would be if I responded to your assertion about "it not making sense" by saying "conversation with you was futile" instead of providing the reasons why "it", contrary to your statement, does make sense.

I read your opinion. I read your reasoning. I don't agree with either. Still I accept that what you stated is your opinion. All I'm asking you for is to do the same.

OK, this is a "non-rhetorical" question you are proposing. Here is me being puzzled because I'm repeating something I just stated: The commentary in question was part of exactly what you were replying to in your post. In fact, it is the exact same commentary that is present verbatim in your post. To me, it is puzzling to have to point this out explicitly. Can you understand why?

Actually that question was rhetorical. Actually the GF4 MX reference was neither directly in a post I replied to, nor did I comment on it. I left the quote of Chalnoth's post in my reply to BRiT purely as a reference.

Eh? It is a 4 pipe card, with the 9800 name, and the suffix has been used to indicate higher performance contrary to its significantly lower performance. These are not subjective evaluations, nor is the performance, nor is the featureset, these are things objectively verifiable. The other cards and names have such factors as well.
There are an abundant amount of ways they can be ranked based on these things, but there are a limited amount of ways that make sense. You can ask a person and they can say that the 9600 could be called a 9800 and the 9800 a 9600. Does this make sense? Perhaps, but it would require a considerable explanation.

Ok, you seem to misunderstand me here. What I was trying to say was that it doesn't make much sense to quantify the 'level of badness' a certain naming decision entails and then rank multiple decisions by their 'level of badness', because obviously, different people would argue different rankings though in the end, all bad naming decisions, no matter how low or high their 'level of badness', ought to be rectified.

Well, it was first digit being technology generation (notwithstanding that this is discussion you are "agreeing to ignore") and numbers after indicating relative performance within the generation. With ATI's explanation of the Pro addendum (Pro full performance, non Pro somewhat reduced), the 9000, 9200, 9600, 9700, and 9800 (non SE) fit this by objective criteria. The 9100, 9500 Pro, and 9800 SE do not. You are free to interpret this differently, but you are not free to state that this evaluation does not make sense because "everything is subjective anyways".

It's interesting that you list the 9600 series with those that fit while delegating the 9500 Pro to the 'non-compliant' cards, though the latter came to the market much earlier and therefore set the yardstick the 9600 series should be measured against. That doesn't make sense. As you know, I don't think R200-, RV250- or RV280-based cards deserve the 9xxx denomination as they do not belong to the same technology generation as R300, R350 and RV350.

So for me, through my own subjective analysis, I concluded that while the 9500, 9700 and 9800 NP/Pro do fit ATi's old naming conventions, 9000, 9100, 9200, 9600 NP/Pro/TX and 9800 SE don't. See, based on the same facts, we come to widely differing conclusions, hence the "everything is subjective anyways" comment.

cu

incurable

 

[Myrmecophagavir]

Myrm,

Um, ATI once claimed that their naming scheme was aligned with DX versions. So, all 7xxx's should have been "DX7" cards. All 8xxx's DX8, and all 9xxx's DX9's. See something wrong with this picture? I hope so. ATI is at least as guilty as nVidia was with the GF4 MX.

If JC called the GF4 MX the greatest lie in the IT marketing history , why is ATi trying so hard to catch up ? First there was the 9000 a tad slower than the 8500 , then there was the 9600 a tad slower than the 9500 Pro and now this .... Could they at least use other numbering scheme or overclock the chips a bit to leave up to their name ?

OK , GF4MX had nothing to do with the original GF4Ti core while these 9800SE solution are using exactly the same core and they really ARE DX9 capable while GF4MX was mearely DX8 compatible .

Do you interpret these differently than I do, or did you just not notice them? Hopefully it answers your question.

Interpret differently Chalnoth said "ATI is as least as guilty as nVidia was [guilty]...", so he doesn't excuse them. David's post doesn't excuse them either, in fact I thought he was doing the opposite? "GF4MX had nothing to do with the original GF4Ti core... [and therefore the naming was very bad]". Well that's how I read it anyway but I still can't read these any other way.

Pardon, you seem to have some misunderstanding of my statements. Excusing means "making seems less bad than it was", not (necessarily) "making seem not bad" in my usage.

Personally I use "excuse" only to mean completely forgive, and "mitigate" for lessen-the-impact. So I certainly read your 1st post a bit differently than you intended. That said, I can't see any post before yours even making Nvidia's naming seem less bad anyway Chalnoth implicitly said Nvidia were "guilty", seems pretty damning to me.

Personally I think that if "technology generation" means the design processes and components made to produce the thing rather than exposed API features (in order to excuse the 9000), then the 4 MX would also qualify by the same rule due to LMA II, Accuview and the VPE, and introduction time as you mentioned (and first two characters of codename, were you serious with that :?) But I'm mighty suspicious of the fact that ATI's chosen generation identifiers happen to match up so closely with the DirectX versions of the time... if it was unintentional and they really did have in mind the possibility of releasing a shader-1 level part under the 9 generation all along they should have stopped and made that explicit to people from the start instead of hoping they'd understand what sort of "generation" they were talking about...

Oh and one other random thing, I don't believe the GF2 supported environment bump mapping? So it wouldn't have been dropped if the GF4 MX didn't support it either.

 

[demalion]

I will change my point of view when provided with a coherent case for doing so. You have not provided one. By what criteria are you exempt from this step?

I thought I'd understood you, but apparently my understanding was in error. You aren't "agreeing to disagree", you're "agreeing not to pay attention to the reasons for an opposing viewpoint". As I would be if I responded to your assertion about "it not making sense" by saying "conversation with you was futile" instead of providing the reasons why "it", contrary to your statement, does make sense.

I read your opinion. I read your reasoning. I don't agree with either. Still I accept that what you stated is your opinion. All I'm asking you for is to do the same.

We're not arguing for to the same end with regard to each other's opinions. You are maintaining my opinion and reasoning do not make sense. I am not arguing that your differing reasoning about what technology generation means does not make sense, I am stating that it isn't the only one, and that you don't dictate which one applies. This is not based on mine being the only one, nor do I claim your reasoning for yours doesn't make sense...I am maintaining that the "ATI naming convention" is an objective standard we can both evaluate but neither of us sets
It is your insistence that my reasoning does not make sense, regardless of addressing my statements having shown that it does, that remains an issue.

When you said "we agree to disagree", I thought we'd resolved that problem. When you said "I haven't changed my viewpoint" and "I just gave up on convincing you that you don't make sense, but you still don't", it showed me that we had not. Your validation for this continues to be "everything is subjective" alone. That is not a sound argument, reasoning already provided, and provide one more time below.

OK, this is a "non-rhetorical" question you are proposing. Here is me being puzzled because I'm repeating something I just stated: The commentary in question was part of exactly what you were replying to in your post. In fact, it is the exact same commentary that is present verbatim in your post. To me, it is puzzling to have to point this out explicitly. Can you understand why?

Actually that question was rhetorical.

Well it appears as rhetorical as the first question you posed, as they both seem to have answers that seem obviously already stated by me, and both seem geared towards asserting that I am incorrect rather than considering my response. I.e., you propose them both as things to confound my reasoning, yet it seems to me I have clearly illustrated that they do not, even before you asked them.

Hence my quotes around "non-rhetorical", though you are free to explain what your actual intent was to me and I hope you'll understand why they seem similar from my perspective.

Actually the GF4 MX reference was neither directly in a post I replied to,

I think the person you were replying to might disagree, as it seems reasonable that they had it specifically in mind...when they quoted it. I know I had it specifically in mind when I saw the text in your post.

nor did I comment on it. I left the quote of Chalnoth's post in my reply to BRiT purely as a reference.

This is a logical contradiction. It is exactly it being there "for reference" that relates your commentary to it.

Let's take a person who agrees the the 9000 name is worse than the GF 4 MX name, but doesn't think the "technology generation" explanation makes sense. How is your post different than how they would have successfully communicated that?

Should they be required to say "I agree with the quote you are responding to" before continuing, or should someone who does not agree with the 9000 name being worse than the GF 4 MX name be required to state "While I don't agree with what you they said"?

From my perspective: either you believe the first, and I don't understand why, or it didn't occur to you to consider that I might have valid reasons for my commentary and that maybe you should have done the second.
You've indicated to me in the prior post (where I said "OK") that you don't think the GF 4 MX is "exactly the same as" the 9000 naming, and we can move on at any time now that you understand what about your text left the question....or we could not move on if you have some reasoning illustrating why my interpretation is unreasonable. Whichever.

Eh? It is a 4 pipe card, with the 9800 name, and the suffix has been used to indicate higher performance contrary to its significantly lower performance. These are not subjective evaluations, nor is the performance, nor is the featureset, these are things objectively verifiable. The other cards and names have such factors as well.
There are an abundant amount of ways they can be ranked based on these things, but there are a limited amount of ways that make sense. You can ask a person and they can say that the 9600 could be called a 9800 and the 9800 a 9600. Does this make sense? Perhaps, but it would require a considerable explanation.

Ok, you seem to misunderstand me here. What I was trying to say was that it doesn't make much sense to quantify the 'level of badness' a certain naming decision entails and then rank multiple decisions by their 'level of badness', because obviously, different people would argue different rankings though in the end, all bad naming decisions, no matter how low or high their 'level of badness', ought to be rectified.

Replace "it doesn't make much sense to" with "I don't care to" and we have nothing to disagree over. Please note the bolded statements where I think I've explained what I'm about to restate (for the last time, hopefully?) again, and your commentary seems to fail to recognize:

As I've explained (including why), it does make sense to do this, because what the consumer saw advertised and what they get from the cards can be evaluated objectively.
Yes, there are people who bought a GF 4 MX card, expecting "DX 8 effects" like shiny water, etc, that developers list for games. Those people did not get them.
No, there are not people who bought a 9000, and didn't get the capabilities for the "DX 8 effects" it lists, with their purchase. TR: AOD, HL2, Doom 3...all of these games continue to support this distinction as being real and definite, if my assurances are insufficient.

What doesn't make sense about this evaluation? There is no subjective doubt in these contrasting statements, and it would help if you would finally point out where there is instead of simply repeating that it is so in the face of discussion supporting the contrary.

It's interesting that you list the 9600 series with those that fit while delegating the 9500 Pro to the 'non-compliant' cards, though the latter came to the market much earlier and therefore set the yardstick the 9600 series should be measured against.

Thanks.

That doesn't make sense.

Well, I'll treat this as a rude form of "Why do you say that?".
I say that because: the RV350 was only left with the number 9600 between 9700 and 9500.
It seems to make sense to me that the 9500 Pro is valid outside of the context of the RV350, because it is faster than the 9500 non Pro, and slower than the 9700 Pro. It also seems to me that it would have made sense as a "9700 LE" (for example) because of the 8 pipes and relationship in performance to the 9700.
Why do I say the 9500 Pro doesn't fit? Because it left no room between itself and the 9500 non Pro to accurately name the 9600 and 9600 Pro, and there were other names that could have been chosen that would have left room.
How does this not make sense?

Now, you can show me how there was indeed a number between the numbers of the 9500 and 9500 Pro to fit the RV350 performance and featureset, or you can provide an argument that provides support for what you think the RV350 should have been named instead (which would leave us agreeing to disagree), but what you cannot do is simply state that I'm "not making sense", without failing to make sense yourself until you provide support of some form for that.

This does not mean that you aren't entitled to think otherwise than I do on the matter, but it does mean that I might also have ample reason to say "that doesn't make sense" to you even though...it is you. It also means that I just might be right, for reasons other than I'm me and I repeated it enough times (i.e., objective ones), if you don't have support that holds together. I can't make sense. while saying you do not, while skipping over that last step, if you've provided support...and neither can you.

As you know, I don't think R200-, RV250- or RV280-based cards deserve the 9xxx denomination as they do not belong to the same technology generation as R300, R350 and RV350.

That's because you continue to maintain that an interpretation of "technology generation" that differs from yourse "doesn't make sense", without the bother of showing why an argument showing that it does is wrong in any particular.

So for me, through my own subjective analysis,

I have more than my own "subjective analysis" to support my viewpoint on the 9000 naming. I've already provided it, quite clearly itemized. Your offering something based on your own "subjective analysis" is not equivalent as far as the 9000 naming, it just means you prefer to consider your own naming convention standard that you dictate and not ATI's. If you didn't confuse one with the other, we wouldn't be in dispute.

I concluded that while the 9500, 9700 and 9800 NP/Pro do fit ATi's old naming conventions, 9000, 9100, 9200, 9600 NP/Pro/TX and 9800 SE don't.

Contrary to popular belief, I do not pursue every opportunity for dispute.
I have reasons to differ with your evaluation of the 9500 (Pro) and 9600 that are not purely subjective, but did not provide them fully until you told me my viewpoint didn't make sense. The 9600 TX you list fits into my reasoning as well.
I have reasons to dispute your statements concerning "fitting ATI's naming conventions", that are not purely subjective, with regards to the 9000 and 9200.
I agree with the 9100, and 4 pipe 9800 SE.

See, based on the same facts, we come to widely differing conclusions, hence the "everything is subjective anyways" comment.

If the argument was "I do not like these names" instead of "these names do not fit into ATI's naming scheme, and it doesn't make sense when you say they do", we would be "agreeing to disagree" (for some of the names) right now. "I do not like" is a subjective standard, "they do not fit into ATI's naming scheme" is an objective standard, with objective criteria to be evaluated. Again: "It does not make sense" to confuse them in the way that you are doing, and dismiss evaluating those objective criteria because "everything is subjective".

 

[demalion]

Myrm,

Um, ATI once claimed that their naming scheme was aligned with DX versions. So, all 7xxx's should have been "DX7" cards. All 8xxx's DX8, and all 9xxx's DX9's. See something wrong with this picture? I hope so. ATI is at least as guilty as nVidia was with the GF4 MX.

If JC called the GF4 MX the greatest lie in the IT marketing history , why is ATi trying so hard to catch up ? First there was the 9000 a tad slower than the 8500 , then there was the 9600 a tad slower than the 9500 Pro and now this .... Could they at least use other numbering scheme or overclock the chips a bit to leave up to their name ?

OK , GF4MX had nothing to do with the original GF4Ti core while these 9800SE solution are using exactly the same core and they really ARE DX9 capable while GF4MX was mearely DX8 compatible .

Do you interpret these differently than I do, or did you just not notice them? Hopefully it answers your question.

Interpret differently Chalnoth said "ATI is as least as guilty as nVidia was [guilty]...", so he doesn't excuse them.

Well, you can interpret "excuse" differently for yourself, but not for me, heh. Please note that "make allowance for" and "overlook", as one set of examples, are not remotely synonymous with "make blameless". That would be something like "exhonerate", which is not a word I used.

David's post doesn't excuse them either, in fact I thought he was doing the opposite? "GF4MX had nothing to do with the original GF4Ti core... [and therefore the naming was very bad]". Well that's how I read it anyway but I still can't read these any other way.

Hmm...well, you can substitute "overlook" and "make allowance for" where you read my use of excuse, as I tried to help Chalnoth with, or you can take a look at my link above and decide whether you want to broaden your consideration of this particular word beyond how you do so at present. In any case, I hope the difference in interpretation is resolved.

As for David G., his silence (AFAICS) to having the contradiction in his post pointed out to him as an answer to his own question, indicates to me at this time that he did not mean to equate the GF 4 MX with the 9000 withour regard to features, rather than that he simply didn't realize the contradiction to his initial query present in his post. JC's commentary on the GF 4 MX being "the greatest lie in the IT industry", as David proposes he stated (I don't know the exact quote), is not remotely reasonably associated at all with the 9000, as my discussion covers, because the reasons he gave were specifically due to the GF 4 MX lacking shader functionality.

Pardon, you seem to have some misunderstanding of my statements. Excusing means "making seems less bad than it was", not (necessarily) "making seem not bad" in my usage.

Personally I use "excuse" only to mean completely forgive, and "mitigate" for lessen-the-impact. So I certainly read your 1st post a bit differently than you intended. That said, I can't see any post before yours even making Nvidia's naming seem less bad anyway Chalnoth implicitly said Nvidia were "guilty", seems pretty damning to me.

Well, here is a bit of a logical incompatibility between us...I've already covered my issue with the "all bad is equally bad" conundrum elsewhere. "ATI is at least as guilty" explicitly states that the GF 4 MX naming is not worse than the 9000 naming (it even indicates that it might be better naming) There are plenty of objective criteria that establish that as untrue. What do you call it when you say something compares less negatively than objective criteria indicate it actually does compare? Replace my associated phrases with that word or phrase for your interpretation here, and our different interpretations will continue to be resolved.

Personally I think that if "technology generation" means the design processes and components made to produce the thing rather than exposed API features (in order to excuse the 9000), then the 4 MX would also qualify by the same rule due to LMA II, Accuview and the VPE, and introduction time as you mentioned (and first two characters of codename, were you serious with that :?)

That would leave the issue of the DX 8 specs listing and comparisons to the prior generation high end (GF 3 Ti and Radeon 8500) in performance and features, as I've covered. But it would also make the naming argument irrelevant in the first place, while that is what Chalnoth is focusing on to the exclusion of featureset comparison. That was the reason for my "proving yourself wrong" commentary.

But I'm mighty suspicious of the fact that ATI's chosen generation identifiers happen to match up so closely with the DirectX versions of the time... if it was unintentional and they really did have in mind the possibility of releasing a shader-1 level part under the 9 generation all along they should have stopped and made that explicit to people from the start instead of hoping they'd understand what sort of "generation" they were talking about...

Indeed, but the DX version listed for support is DX 8. You need not confuse this discussion with one maintaining there is no problem with the 9000 name, it merely proposes that there is a consistent reason to name it as was done.

Oh and one other random thing, I don't believe the GF2 supported environment bump mapping? So it wouldn't have been dropped if the GF4 MX didn't support it either.

Well, not in comparison to the GF 2 Ultra, which I was allowing to be stipulated when I asked the question. My "GF 2 ~ Radeon" thought that prompted that will hopefully not lead me astry again.

 

[Bjorn]

That would leave the issue of the DX 8 specs listing and comparisons to the prior generation high end (GF 3 Ti and Radeon 8500) in performance and features, as I've covered.
But it would also make the naming argument irrelevant in the first place, while that is what Chalnoth is focusing on to the exclusion of featureset comparison. That was the reason for my "proving yourself wrong" commentary.

Ok, let me see if i get this straight.

What i get from what you're saying, correct me if i'm wrong, is that there would be nothing wrong (at least not more wrong then what Ati does now with the 9000-9200 series) of Nvidia to call the GF4 MX a GF4 if they only hadn't marketed it as a DX8 part ?

Edit:

To me, this is really simple. The reason for the controversy with regards to the GF4 MX was about the name, not one line about API support in a PDF. That was also what JC complained about. The name. Because it could potentially lead people to believe that it was really a DX8 card since the GF4 Ti series was and they were both called GF4.

What Ati did then was to realize that Nvidia was in a bit of trouble and went out and said something like "we're going to stick to our technology generation naming scheme so that the consumers knows what they get" in direct response to make them look better then Nvidia, and they were a LOT better at the time also. And now, Ati is imo doing the exact same thing with the R9000-9200 series. Even if you think that the 9000 and 9200 somehow can be labelled as a new technology generation and that the consumers somehow is helped by Ati calling them that despite that they're not getting any more features then the 8500, it still leaves the 9100 which definitely is exactly the same thing as the GF4 MX problem if were going to stick to the "technology generation" statement.

It was released along with the R300, and has the same first digit. This does indeed make a great deal of clear and unambiguous sense with regards to "technology generation". How do you say it does not?

Which of course means that it makes a great deal of clear and unambigous sense to name the GF4 MX as a GF4 (again, with regards to the technology generation) so i guess i was right above then and you don't need to clarify it

 

[demalion]

That would leave the issue of the DX 8 specs listing and comparisons to the prior generation high end (GF 3 Ti and Radeon 8500) in performance and features, as I've covered.
But it would also make the naming argument irrelevant in the first place, while that is what Chalnoth is focusing on to the exclusion of featureset comparison. That was the reason for my "proving yourself wrong" commentary.

Ok, let me see if i get this straight.

What i get from what you're saying, correct me if i'm wrong, is that there would be nothing wrong (at least not more wrong then what Ati does now with the 9000-9200 series) of Nvidia to call the GF4 MX a GF4 if they only hadn't marketed it as a DX8 part ?

First, drop that "nothing wrong" completely, it doesn't belong anywhere near my sentiments of the 9000 name, or the GF 4 MX name, even in this hypothetical situation.
Second, I direct you to our prior discussion (linked a few pages back), because I told you what I would think was wrong in that circumstance in that thread already. To provide some aid, this is related to the DX7->DX8 versus DX8->DX9 discussion that I referred to earlier, things I repeated to you several times before, and things I've repeated several times in this thread already.

Edit:

To me, this is really simple. The reason for the controversy with regards to the GF4 MX was about the name, not one line about API support in a PDF.

When you say "one line in the PDF", you seem to indicate this is the only place it is found, and the difference betwen "DX 7" and "DX 8" is totally defined by the number "1", rather than the featureset difference between them. Neither of these things are true. Perhaps perusing specs listings in product advertisements at some online retailrs, or canvasing some GF 4 MX boxes from several manufacturers, will clarify this for you. The PDF is a convenience of reference, not the only specs listing for the GF 4 MX in existence.

You really should read the thread I linked to earlier again(?), as I think it would help you avoid asking me to repeat myself.

...

As far as your apparent address of my reference to John Carmack:

That was also what JC complained about. The name.

Well, what he actually said was:

...
Do not buy a GeForce4-MX for Doom.

Nvidia has really made a mess of the naming conventions here. I always
thought it was bad enough that GF2 was just a speed bumped GF1, while GF3 had
significant architectural improvements over GF2. I expected GF4 to be the
speed bumped GF3, but calling the NV17 GF4-MX really sucks.
...

Hmm...ayep, he says there is a problem with the name. Hey, so did I. If you snipped that part out and evaluated by itself, I guess I could prove myself wrong!

However, I know that I claimed there was more wrong than just the name. How about John Carmack?

...
GF4-MX will still run Doom properly, but it will be using the NV10 codepath
with only two texture units and no vertex shaders. A GF3 or 8500 will be
much better performers. The GF4-MX may still be the card of choice for many
people depending on pricing, especially considering that many games won't use
four textures and vertex programs, but damn, I wish they had named it
something else.

Does John Carmack just mention "the name", or does he mention that the problem is the name, because it lacks vertex shaders? Do I have to bold "vertex shaders"? Do I have to prove that his sentence structure indicates that there is causality between this feature it lacks and his complaint about the name, or can we assume he didn't type the words because he thought they looked nice there?

Quick question: does the 9000 have "vertex shaders"? So, if I say that what he complained about for the GF 4 MX naming isn't the same thing as the problem with the 9000 naming, on what basis are you making your disagreement? Ignoring that he said anything except that there was a "problem with the name"?

Because it could potentially lead people to believe that it was really a DX8 card since the GF4 Ti series was and they were both called GF4.

Well, he said his concern is with lack of vertex shaders, and he later went on to establish that pixel and vertex shaders is the dividing line for successfully delivering the features of the game.
The GF 4 MX does not have shaders. The 9000 does.

Which detail of my representation of John Carmack's statements was in error such that you brought his commentary up? Why did you say he said the reason was "the name" when he actually said "the name because it lacks shaders", other than to avoid tge inconvenience of his commentary indicating the difference you choose to maintain is not there?

Pardon, I've had the "ATI attacked nVidia and did the 'exact same thing'" discussion several times already, and you are demonstrating you are simply not interested in reading a word of my part of it. I don't see a point in repeating the conversation again simply because you managed to not read someone else's words as well.

 

[Bjorn]

Does John Carmack just mention "the name", or does he mention that the problem is the name, because it lacks vertex shaders?

Which detail of my representation of John Carmack's statements was in error such that you brought his commentary up? Why did you say he said the reason was "the name" when he actually said "the name because it lacks shaders", other than to avoid tge inconvenience of his commentary indicating the difference you choose to maintain is not there?

This is actually getting funny. Yes, he mentions other things but the name but all those things are connected to the name and wouldn't have been anything to complain about if it wasn't for the name. You seem to disagree with this though so let's have a look at the second JC quote again:

GF4-MX will ..........., but damn, I wish they had named it
something else.

Again, the complaints are about the name. Sure it lacks stuff which he mentions, but, why is that a problem ?, yep, because of the name.

Pardon, I've had the "ATI attacked nVidia and did the 'exact same thing'" discussion several times already, and you are demonstrating you are simply not interested in reading a word of my part of it.

I have no problems reading it. Reading is not always equal to agreeing though.

Quick question: does the 9000 have "vertex shaders"? So, if I say that what he complained about for the GF 4 MX naming isn't the same thing as the problem with the 9000 naming, on what basis are you making your disagreement? Ignoring that he said anything except that there was a "problem with the name"?

Ok, as have been mentioned before, let's compare the 9000 series with the supposed same "technology generation" that it belongs to, the 9500+ series. And then make the same statement as you did:

Does the 9000-9200 have DX9 support which the other cards in the same "technology generation" have ?

So if i say that the thing he complained about with the GF4 MX is exactly the same problem as with the 9000-9200, on what basis are you making your disagreement ?

 

[KimB]

...
GF4-MX will still run Doom properly, but it will be using the NV10 codepath
with only two texture units and no vertex shaders. A GF3 or 8500 will be
much better performers. The GF4-MX may still be the card of choice for many
people depending on pricing, especially considering that many games won't use
four textures and vertex programs, but damn, I wish they had named it
something else.

Does John Carmack just mention "the name", or does he mention that the problem is the name, because it lacks vertex shaders?

Lack of support for vertex programs isn't the main issue here. Notice what he states first: the NV10 codepath and two texture units (which implies limited programmability).

 

[KimB]

I think I'd like to sum up my arguments on this issue, finally.

1. GeForce4 MX. This product was a bad move for nVidia in two different ways. On one hand, it didn't support 2nd-gen programmability (NV2x, R2xx-level programmability). On the other, the product's name implies that it does. So, this product both held back the production of games using 2nd-gen features, and, quite possibly, fooled a number of people into believing that they had purchased a video card that would be capable of such things.

2. ATI introduces the Radeon 7000, 7200, 7500, and 8500. The new naming scheme was marketted as being unambiguous. "7" denoted "DirectX 7 parts," while "8" denoted "DirectX 8 parts."

3. ATI introduces the Radeon 9000. This new chip is a derivative of the Radeon 8500 core, and has the same programming interface. Yet, it is clearly marketted as belonging to the same family as the Radeon 9700 Pro video card, and thus it is implied that it is also a 3rd-gen part. This is in direct contrast to ATI's previously-marketted naming scheme, and is, in my mind, no different from the naming scheme of the GeForce4 MX. Furthermore, it was marketted as somehow being better than the Radeon 8500, which was certainly not the case. Of course, one cannot fault it for one of the Geforce4 MX's faults: that it held back technology. The 9000 was the first low-end 2nd-gen part, so it clearly wasn't holding back anything.

4. ATI further confuses the product line with the 9100, 9200, 9600, and 9800 SE parts. Not only are there now more 2nd-gen parts being marketted as belonging to a 3rd-gen product line, but there isn't even consitency in performance between the various products in this line.

5. nVidia introduces FX product line, which is 3rd-gen from top to bottom, and each part is clearly-lableled as to which is higher-performing than the last.

The way I see it, it is ATI who is now marketting confusion and misinformation, nVidia who has a very self-consistent product line.

 

[Joe DeFuria]

The way I see it, it is ATI who is now marketting confusion and misinformation, nVidia who has a very self-consistent product line.

Because the GeForce FX 5200 beats the hell out of the GeForce4 4600?

Both naming schemes are FUBARd at this point. Time to go back to the drawing board...both ATI and nVidia.

 

[KimB]

Because the GeForce FX 5200 beats the hell out of the GeForce4 4600?

Both naming schemes are FUBARd at this point. Time to go back to the drawing board...both ATI and nVidia.

Well, I think it'd be silly to buy a 4600 right now. There are better alternatives. Anyway, of course there are always going to be generational differences, but you can't very well label everything solely based upon performance, can you? The low end of the next generation is always going to be slower than the high end of the previous. I don't think you can expect anything better than what nVidia has now with the FX line's naming scheme. I just hope that they stick to it, though it is likely to be complicated by any products coming out this fall.

 

[Dio]

Because the GeForce FX 5200 beats the hell out of the GeForce4 4600?

Both naming schemes are FUBARd at this point. Time to go back to the drawing board...both ATI and nVidia.

Well, I think it'd be silly to buy a 4600 right now.

Not as silly as it would be to by a 5200.

 

[demalion]

Does John Carmack just mention "the name", or does he mention that the problem is the name, because it lacks vertex shaders?

Which detail of my representation of John Carmack's statements was in error such that you brought his commentary up? Why did you say he said the reason was "the name" when he actually said "the name because it lacks shaders", other than to avoid tge inconvenience of his commentary indicating the difference you choose to maintain is not there?

This is actually getting funny.

In a sad sort of way.

Yes, he mentions other things but the name but all those things are connected to the name and wouldn't have been anything to complain about if it wasn't for the name.

Or, rather, the name wouldn't have been anything to complain about if it had shader features. One of those statements depends on Carmack's mentioning of shaders not being clearly indicated. One is based on him mentioning them specifically. Carmack mentioned them specifically.

Explain the problem with the statement that depends on him mentioning them explicitly, such that you represent his statements as having them absent. I'm not seeing it.

One of the sad-funny things about your commentary is how they answer some of the questions I put to you:

Quick question: does the 9000 have "vertex shaders"? So, if I say that what he complained about for the GF 4 MX naming isn't the same thing as the problem with the 9000 naming, on what basis are you making your disagreement? Ignoring that he said anything except that there was a "problem with the name"?

Taking note of (EDIT)the first bold question above, read this excerpt of my statements about John Carmack just prior to your post:

JC's commentary on the GF 4 MX being "the greatest lie in the IT industry", as David proposes he stated (I don't know the exact quote), is not remotely reasonably associated at all with the 9000, as my discussion covers, because the reasons he gave were specifically due to the GF 4 MX lacking shader functionality.

Do you have an answer to that question that I highlighted for you? It would save time, perhaps

You seem to disagree with this though so let's have a look at the second JC quote again:

When you are done holding a discussion with whomever you are thinking of...

GF4-MX will ..........., but damn, I wish they had named it
something else.

Again, the complaints are about the name. Sure it lacks stuff which he mentions, but, why is that a problem ?, yep, because of the name.

...and making large sections of text disappear when inconvenient (like...all the rest of Carmack's statements on the GF 4 MX ), get back to me. Or not, if it you continue to find it impossible, I guess.

 

[KimB]

Because the GeForce FX 5200 beats the hell out of the GeForce4 4600?

Both naming schemes are FUBARd at this point. Time to go back to the drawing board...both ATI and nVidia.

Well, I think it'd be silly to buy a 4600 right now.

Not as silly as it would be to by a 5200.

It's just a low-cost card. Perhaps nVidia made too many tradeoffs for performance in order to have a very low transistor count, but I don't think the 5200 is that bad of a product for those purchasing at the low end.

 

[Myrmecophagavir]

David's post doesn't excuse them either, in fact I thought he was doing the opposite? "GF4MX had nothing to do with the original GF4Ti core... [and therefore the naming was very bad]". Well that's how I read it anyway but I still can't read these any other way.

As for David G., his silence (AFAICS) to having the contradiction in his post pointed out to him as an answer to his own question, indicates to me at this time that he did not mean to equate the GF 4 MX with the 9000 withour regard to features, rather than that he simply didn't realize the contradiction to his initial query present in his post. JC's commentary on the GF 4 MX being "the greatest lie in the IT industry", as David proposes he stated (I don't know the exact quote), is not remotely reasonably associated at all with the 9000, as my discussion covers, because the reasons he gave were specifically due to the GF 4 MX lacking shader functionality.

I can't see a contradiction in David's post though - he's not even talking about the 9000, but the 9800 SE. First he asks (rhetorically?) why ATI seems to be copying Nvidia's bad behaviour in naming cards, when Carmack said that was misleading, and backs it up by listing their recent decisions to give cards higher marketing numbers than they ought to have. Then he observes that this case is not as bad as Nvidia's, because this really is a 9800 series (just with lower clocks) whereas there are all sorts of reasons for not wanting the MX to be called a GF4. But this has nothing to do with the 9000, nor does it seek to excuse (by anyone's definition of the term ) Nvidia's nomenclature.

"ATI is at least as guilty" explicitly states that the GF 4 MX naming is not worse than the 9000 naming (it even indicates that it might be better naming)

Yep agreed.

There are plenty of objective criteria that establish that as untrue.

Not agreed here though - "objective"! I think you'd be hard pushed to prove that these arguments are genuinely objective, it's a name after all! They can call it anything they want to at the end of the day, and I don't think you can produce an objective scale of "badness of naming"

Personally I think that if "technology generation" means the design processes and components made to produce the thing rather than exposed API features (in order to excuse the 9000), then the 4 MX would also qualify by the same rule due to LMA II, Accuview and the VPE, and introduction time as you mentioned (and first two characters of codename, were you serious with that :?)

That would leave the issue of the DX 8 specs listing and comparisons to the prior generation high end (GF 3 Ti and Radeon 8500) in performance and features, as I've covered. But it would also make the naming argument irrelevant in the first place, while that is what Chalnoth is focusing on to the exclusion of featureset comparison. That was the reason for my "proving yourself wrong" commentary.

You mean where they claim the MX has "Complete DirectX® support, including DirectX 8.1"? "Complete" is misleading but as far as API support goes I'd say it's valid to claim DX 8.1 compliance. They correctly expose all the relevent caps through the API to indicate which parts of the spec are available. W.R.T. comparisons to previous high-end models, yes the GF4 MX has generally lower performance and features missing from the GF3, but also adds some like a new antialiasing mode and a video processor, and more efficient memory controller... whereas the 9000 has generally lower performance than the 8500, but also adds some features like improved triangle setup and Fullstream video processing. So you win some, you lose some, with both. But why does that make the naming argument irrelevent?

But I'm mighty suspicious of the fact that ATI's chosen generation identifiers happen to match up so closely with the DirectX versions of the time... if it was unintentional and they really did have in mind the possibility of releasing a shader-1 level part under the 9 generation all along they should have stopped and made that explicit to people from the start instead of hoping they'd understand what sort of "generation" they were talking about...

Indeed, but the DX version listed for support is DX 8. You need not confuse this discussion with one maintaining there is no problem with the 9000 name, it merely proposes that there is a consistent reason to name it as was done.

Listed on ATI's website for the 9000 you mean? Yes that's better of them in terms of what the public seems to understand, but they still play the game with their 9700 description: "First to fully support DirectX® 9.0"... well obviously that conveniently forgets about things like PS/VS 3.0 and many other bits and pieces. Now fast-forward into the future when cards come out supporting this and in comparison the 9700 will simply not have the features yet still have been marketed like that.

I can see there is an argument for the naming of the 9000, I just believe the same argument could apply to the GF4 MX. So I say they're both out, and we all go round to their offices and give them a bonk on the noggin!

 

[demalion]

...
As for David G., his silence (AFAICS) to having the contradiction in his post pointed out to him as an answer to his own question, indicates to me at this time that he did not mean to equate the GF 4 MX with the 9000 withour regard to features, rather than that he simply didn't realize the contradiction to his initial query present in his post. JC's commentary on the GF 4 MX being "the greatest lie in the IT industry", as David proposes he stated (I don't know the exact quote), is not remotely reasonably associated at all with the 9000, as my discussion covers, because the reasons he gave were specifically due to the GF 4 MX lacking shader functionality.

I can't see a contradiction in David's post though - he's not even talking about the 9000, but the 9800 SE.

The cards mentioned concerning ATI "trying so hard to catch up" to Carmack's negative characterization of the GF 4 MX were: 9000, 9600, and 9800SE. Am I wrong in this somehow, or do you understand how your comment doesn't make sense to me?

First he asks (rhetorically?) why ATI seems to be copying Nvidia's bad behaviour in naming cards, when Carmack said that was misleading, and backs it up by listing their recent decisions to give cards higher marketing numbers than they ought to have.

Well, this relates to my discussion of Carmack's comments, and how he is specifically talking about the lack of shaders being the reason for an issue with the name. Hence, Carmack's commentary does not apply to any of the cards listed, as I've argued more than once already if you're wondering as to my reasoning.

Then he observes that this case is not as bad as Nvidia's, because this really is a 9800 series (just with lower clocks) whereas there are all sorts of reasons for not wanting the MX to be called a GF4. But this has nothing to do with the 9000, nor does it seek to excuse (by anyone's definition of the term ) Nvidia's nomenclature.

Yes, and this contradicts his comparison to Carmack's commentary on the GF 4 MX, which Carmack made for specific reasons I've already discussed. I'm sorry, but I just don't see a point anywhere in sight.

"ATI is at least as guilty" explicitly states that the GF 4 MX naming is not worse than the 9000 naming (it even indicates that it might be better naming)

Yep agreed.

There are plenty of objective criteria that establish that as untrue.

Not agreed here though - "objective"! I think you'd be hard pushed to prove that these arguments are genuinely objective, it's a name after all!

So, my name on this forum being "demalion" is not objectively verifiable? One number being greater than another is not objectively verifiable? You're not making sense...something being a name does not prevent it from being objectively evaluated, and "it's a name after all!" is your only counter-argument support!

They can call it anything they want to at the end of the day, and I don't think you can produce an objective scale of "badness of naming"

Err...did you look at the slides you thanked me for?

Reviewing the definition of objective: "Based on observable phenomena; presented factually", are those slides observable? Are they presented factually? Do they specify criteria for naming? How about DirectX features, and compliance and compatibility? DirectX is "a name after all!", does that mean there can't be criteria specified for these even if Microsoft says so?

Myrm, I still don't see anything you are stating here making sense, at all.

Personally I think that if "technology generation" means the design processes and components made to produce the thing rather than exposed API features (in order to excuse the 9000), then the 4 MX would also qualify by the same rule due to LMA II, Accuview and the VPE, and introduction time as you mentioned (and first two characters of codename, were you serious with that :?)

That would leave the issue of the DX 8 specs listing and comparisons to the prior generation high end (GF 3 Ti and Radeon 8500) in performance and features, as I've covered. But it would also make the naming argument irrelevant in the first place, while that is what Chalnoth is focusing on to the exclusion of featureset comparison. That was the reason for my "proving yourself wrong" commentary.

You mean where they claim the MX has "Complete DirectX® support, including DirectX 8.1"? "Complete" is misleading but as far as API support goes I'd say it's valid to claim DX 8.1 compliance.

Well, actually, I believe it is to avoid exactly this type of DX version listing abuse that "compliance" and "compatibility" took on specific meaning in the context of DX support. AFAIK, what it is valid to claim within that established terminology is "DX 8.1 compatibility", not "DX 8.1 compliance".

In any case, for this instance, the exact same phrase was used to define the GF 4 MX as having the same DX 8.1 support as the GF 4 Ti (exact same DX 8.1 featureset support listed), which has indeed specifically resulted in many consumers skipping over GF 3 cards for the GF 4 MX. Not something that corresponds to the 9000 compared to the 9700.

I'm glad we both recognize the deception in the DX version listing for the GF 4 MX.

They correctly expose all the relevent caps through the API to indicate which parts of the spec are available. W.R.T. comparisons to previous high-end models, yes the GF4 MX has generally lower performance and features missing from the GF3, but also adds some like a new antialiasing mode and a video processor, and more efficient memory controller... whereas the 9000 has generally lower performance than the 8500, but also adds some features like improved triangle setup and Fullstream video processing.

You don't lose features with the 9000. Isn't this a difference? What do you think my argument is?

So you win some, you lose some, with both.

You see, this phrase serves to attempt to dismiss the differences, but you haven't provided an argument for why that dismissal makes sense, other than to claim objective criteria doesn't apply because "it's a name after all!"

But why does that make the naming argument irrelevent?

Pardon? If you're saying their names are both completely justified (I'm taking you at your usage of "excuse") because of "technology generation", you're maintaining that the naming issue is irrelevant, which leaves you without a reason to criticize the 9000 (but still leaves you with issues to criticize about the GF 4 MX, as I've outlined). That's what I was referring to, and referencing Chalnoth's post about.
I don't think either name is completely justified, but I do think that ATI's naming of the 9000 is consistent with their stated naming policy. How many times do I have to say that the 9000-9200 all have naming problems?

My point has and continues to be that the GF 4 MX has significant and unique problems that the 9000 (and about any other graphics card product I can think of at the moment) do not have, and that equating the issues associated with why their names are a problem is not valid, nor is the description of these issues accurately described as "exactly the same issue", and also that it is invalid to say that "ATI is worse because they violated their naming policy" because we have objective criteria that it fits.

Are you losing track of the fact that these names are associated with products? If the GF 4 MX product offered vertex and pixel shader functionality, ATI's naming of the 9000 being "at least as bad" would no longer be an unreasonable statement. Do you think this doesn't make sense?
If you haven't read my link to where I discussed all of this already, please do so as soon as possible.

...
Indeed, but the DX version listed for support is DX 8. You need not confuse this discussion with one maintaining there is no problem with the 9000 name, it merely proposes that there is a consistent reason to name it as was done.

Listed on ATI's website for the 9000 you mean? Yes that's better of them in terms of what the public seems to understand, but they still play the game with their 9700 description: "First to fully support DirectX® 9.0"... well obviously that conveniently forgets about things like PS/VS 3.0 and many other bits and pieces.

If you want to go on record as asserting the 9700's listing of DX 9.0 support being equivalent to the GF 4 MX, be my guest.

Now fast-forward into the future when cards come out supporting this and in comparison the 9700 will simply not have the features yet still have been marketed like that.

3.0 is greater than 2.0, and 1.3 is greater than nothing. Therefore, the situations are exactly the same, nevermind any criteria associated with the things being named?

OK, you're not making constructing an argument worth my time with a post that holds together like this last one of yours seems to, but I will try, once again, to point out something you seem to have missed:
DX 8.1 offered significant new features, and the GF 4 MX did not offer them. DX 9.0 offers significant new features, and the 9700 does offer them.

This is a difference, yes? Saying the GF 4 MX's DX support listing compares unfavorably because of this is reasonable, correct? Does saying that the 9700 is just as bad, or worse, than the GF 4 MX in this regard make sense?

If you think not, then where in this entire post was there a point? If you think so, I'd really like you to go on record as clearly establishing why.

FYI: I do seem to recall specific discussion establishing that the 9700 is entitled to list "DX 9 compliance", related to Microsoft setting terms for such labels, which would coincide with "Complete DX 9 support". I'd even expect that the GF 4 MX specs listing had something to do with this becoming a focus. However, I simply don't see the point in my spending time to search for this for you with as little interest as you are showing in making sense, from my perspective.

I can see there is an argument for the naming of the 9000, I just believe the same argument could apply to the GF4 MX.

If you ignore enough things, sure.

So I say they're both out, and we all go round to their offices and give them a bonk on the noggin!

Of course. This thread was about that, once, for the IHV that apparently deserves it at the moment. It continues to be my suggestion that it return to negative commentary directed at them that actually makes sense, and that nonsensical re-hashing of nVidia's GF 4 MX naming being "exactly the same thing" or "no worse" without regard to facts related to the matter, would best serve by being absent.

 

[incurable]

demalion,

your posts in this thread show that you simply can't accept other people having a different opinion on a matter than you do. This only reinforces my opinion that I stated in my replies to you.

I regret that I actually asked a question in my first reply, even though I thought it was needed for clarification, it apparently gave you the opportunity to pull me into a debate that focussed more around your definition of certain words or terms than the issue at hand. Any further reply to your posts was a mistake. This whole thread, probably a few thousand words in total, has led nowhere and I'm not really surprised about that outcome.

Anyway, I think I made my points, so, I'm done with this thread.

cu

incurable

 

[Bjorn]

Are you losing track of the fact that these names are associated with products? If the GF 4 MX product offered vertex and pixel shader functionality, ATI's naming of the 9000 being "at least as bad" would no longer be an unreasonable statement. Do you think this doesn't make sense?
If you haven't read my link to where I discussed all of this already, please do so as soon as possible.

No, it doesn't make sense since you continue to compare the 9000 to the generation before (8000+, DX8) instead of the 9500+, DX9 that you should compare it according to it's name, "technology generation".

Sure, it didn't hold back technology the way maybe the GF4 MX did, and that might make the GF4 MX a "more bad" product. But we're talking about how bad/misleading it's name is and then we have the same difference, DX7 vs DX8 and DX8 vs DX9.

DX 8.1 offered significant new features, and the GF 4 MX did not offer them. DX 9.0 offers significant new features, and the 9700 does offer them.

Typo ?, maybe you meant to say, " DX 9.0 offers significant new features, and the 9200 didn't offer them"

 

[Bjorn]

Both naming schemes are FUBARd at this point. Time to go back to the drawing board...both ATI and nVidia.

I don't agree. I agree with Chalnoth here, i think most people know that the low end of a new generation might not necessarily be faster then the high end of the previous.

I think that the FX series are named just as they should be. The first digit denotes generation, the second performance within that generation. Now where have i heard that before ?

 

[Mariner]

I agree with Joe. Both ATi and NV are misleading in their naming.

Your average computer user knows absolutely nothing about graphics cards. If they see GeForce4 Ti 4600 and GeForceFX 5200 Ultra they will always think that the higher number with the 'Ultra' added is the faster chip, especially if they have heard that NV's latest and greatest is called the GeForceFX.

Similarly, a Radeon 8500 sounds less than a Radeon 9200SE (incidentally, why on earth does 'SE' indicate a cut down version? You'd think that 'Special Edition' would mean something better than the original :?).

Back to the drawing board for ATi & NV for the next generation of marketing names, I hope.