Radeon 9800 se -> 9800

[demalion]

How many times will we have this discussion before people will pay more attention to provided information than what they simply like to believe?

Sorry, nVidia is not absolved of doing something worse than ATI based on repeating things demonstrated to be false. Why does that seem so unacceptable? There are significant problems with the 9800SE to discuss, but please give up on excusing the GF 4 MX and nVidia by saying others are doing the same thing while ignoring inconvenient information to the contrary.

 

[Dave Baumann]

On a point of technicality - I think the reason the 9800 SE is named as such is because ATI initially stated that for this generation the F-Buffer would only exist on their high end boards, and this being a cut down 9800 the F-Buffer would still exist on it.

 

[demalion]

incurable,

The RV250 was designed after the R200. That's what the 250 probably means.

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? Perhaps you equate the phrase with "feature level" or "performance level", instead of the design principles ATI has stated they used to reduce transitor count signifcantly but maintain the greater part of performance ("technology", to me) and time period targetted and release schedule ("generation"). Or maybe you're just stuck on "technology" being API features offered rather than the technology used to implement them?

The naming of the 9000-9200 cards also has clear problems, it is just that those problems are not the ones that people have decided to propose in association with equating it with the GF 4 MX (i.e., ones contrary to provided facts).

 

[incurable]

incurable,

The RV250 was designed after the R200. That's what the 250 probably means.

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? Perhaps you equate the phrase with "feature level" or "performance level", instead of the design principles ATI has stated they used to reduce transitor count signifcantly but maintain the greater part of performance ("technology", to me) and time period targetted and release schedule ("generation"). Or maybe you're just stuck on "technology" being API features offered rather than the technology used to implement them?

The naming of the 9000-9200 cards also has clear problems, it is just that those problems are not the ones that people have decided to propose in association with equating it with the GF 4 MX (i.e., ones contrary to provided facts).

I have the feeling that you're trying to start yet another debate based purely on semantics and quite frankly, I don't wish to engage in that type of conversation. Therefore, I'll only ask one question:

How does the 9100 fit in your argument. (I'm sure you agree that R200 can't be a generation ahead of R200. )

cu

incurable

 

[demalion]

incurable,

The RV250 was designed after the R200. That's what the 250 probably means.

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? Perhaps you equate the phrase with "feature level" or "performance level", instead of the design principles ATI has stated they used to reduce transitor count signifcantly but maintain the greater part of performance ("technology", to me) and time period targetted and release schedule ("generation"). Or maybe you're just stuck on "technology" being API features offered rather than the technology used to implement them?

The naming of the 9000-9200 cards also has clear problems, it is just that those problems are not the ones that people have decided to propose in association with equating it with the GF 4 MX (i.e., ones contrary to provided facts).

I have the feeling that you're trying to start yet another debate based purely on semantics and quite frankly, I don't wish to engage in that type of conversation.

incurable, what I was trying to say exactly what I stated. Note the series of statements and questions. A conversation is where you respond to them, not assign me a motive and dismiss what I've said based on it.

I'll progress the conversation on your terms, since you seem to be caught up in a preconception of my stance that I can maybe dispell right now:

Therefore, I'll only ask one question:

How does the 9100 fit in your argument. (I'm sure you agree that R200 can't be a generation ahead of R200. )

The 9100 was a relabelling based on an ill thought out approach to resolving exactly the problems I already referred to with regard to the 9000 (plus the problem of excessive 8500 stock, apparently). Namely, that performance level for the 9000 was inferior to the card that was released before it that had a "lower number" associated with it. As I said already, this is indeed a problem, it's just isn't the problem that you described. If you choose to continue a line of criticism on the 9100, my only concern would you keep it on topic...I've already chipped in my own criticisms against the name in a more suitable thread.

Please take a gander at the thread I linked above, as I've specified the reasoning behind this all in quite a bit of detail already within it.

What I think you were after is for me to simply say that ATI did something wrong, that the 9100 is not a good name, and seems to violate ATI's own naming policy, which I agree with. Please don't confuse "Don't criticize ATI for reasons that don't make sense" with "Don't criticize ATI", which is the only thing that comes to mind that would explain why you'd think I'd say otherwise.

I even think that the 9800SE is worse name than the 9100 (atleast for the 4 pipe one), which is why I don't know why people who dislike it insist on bringing up the GF 4 MX when there is plenty of on topic and accurate criticism to bring forth, especially if this is now an official naming policy (it not being official naming when originally brought up a while ago was about the only saving grace I found in the name for the 4 pipe part).

 

[BRiT]

That doesn't make any more sense. RV250 is in no way a generation ahead of R200, neither is RV280, neither is R200 (obviously), still all these chips are sold on 9x00 series cards.

The RV250/RV280 made use of improved and tuned PS/VS engines as well as other minor improvements derived from the technology that went into the R300.

These slides in question are captured in a discussion back in November 2002 here: http://www.beyond3d.com/forum/viewtopic.php?p=52310#52310

And yes, I do agree the naming of the R200/8500 as a 9100 does not follow their own guidelines. The decision to do so was driven from excess 8500 stock which created the need to market it as being faster than a 9000.

 

[KimB]

I never said what ATI did wasn't misleading. I merely pointed out your statement that they were "at least as guilty", which I don't agree with. At least they didn't brand it as a Radeon 9700 MX last fall and document it as possessing full DirectX 9 hardware support, which really would've put them in the same league as Nvidia.

No.

Naming on GF4's:
Ti 4600
Ti 4400
Ti 4200

MX 460
MX 440
MX 420

I might have gotten a number or two a little bit off...but the exact numbers aren't what count here. That is, notice that the MX's were followed by numbers in the 400's, while the Ti's in the 4000's. As far as I'm concerned, that alone implies a significant difference between the two different lines of video cards.

 

[KimB]

iIt 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".

Technology generation != product generation.

It's that simple. The R3xx is in an entirely new league in terms of technology than the R2xx. It's a different technology generation.

Notice the slides on the link posted a couple of times already on this thread. The products on that slide were:

Radeon 8500
Radeon 7500
Radeon 7200
Radeon 7000

These products were released (or re-released) all at about the same time. The 8500 was clearly a "technology generation" ahead of the 7x00 products. We had the exact same thing upon the release of the 9000 and (later) the 9100 and 9200. These were products based upon older technology, and the use of the "9" in the first digit was not only as misleading as what nVidia did by labelling the GeForce4 MX as a GeForce4, but I consider it worse because ATI previously called nVidia to task on the 4MX naming, stating their own naming scheme formally in that presentation.

Edit:
One final note. This isn't about excusing nVidia. They, however, have corrected their mistake with the GeForce FX line's naming scheme. One can only hope that both companies do as good a job as nVidia has done with the FX line for future products.

 

[Bjorn]

Or maybe you're just stuck on "technology" being API features offered rather than the technology used to implement them?

What's wrong with being stuck on that ? After all, does the consumer care about "the technology that's used to implement them" ?

The name's are for marketing and thus, they have nothing to do with the underlaying technology and everything to do with features/performance/price. And thus, (all imo of course) they're wrong for pretty much the same reason that the GF4 MX was wrong.

One can only hope that both companies do as good a job as nVidia has done with the FX line for future products.

I hope you're only talking about the naming schemes here

 

[demalion]

iIt 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".

Technology generation != product generation.

I already provided my reasoning with regard to "technology". Simply saying "that is not so" does not do anything to answer it. The 9000 delivers the same old features, using new technology and less transistors. Please don't tell me silly things like technology does not mean technology.

It's that simple.

By the expedient of your having simplified things to your liking.

The R3xx is in an entirely new league in terms of technology than the R2xx. It's a different technology generation.

ATi can't decide how to assign their own technology generations, even when there is logic and evidence to support their choice? Chalnoth, that doesn't make sense.

Notice the slides on the link posted a couple of times already on this thread. The products on that slide were:

Radeon 8500
Radeon 7500
Radeon 7200
Radeon 7000

These products were released (or re-released) all at about the same time.

Yes, because it makes sense to rename things when you are...introducing a new naming scheme and applying it to your entire product line up.

The 8500 was clearly a "technology generation" ahead of the 7x00 products.

Yes...is there something new that you're thinking these statements establish? If so, I'm missing it and I'd appreciate it if you clarify.

We had the exact same thing upon the release of the 9000 and (later) the 9100 and 9200.

Please read my prior comments on the 9100. It is a different issue than the 9000 and 9200. All 3, however, do have problems in their naming.

These were products based upon older technology,

Well, the 9100 was. The 9200 was a modified design, and perfectly consistent with the "9" technology generation already discussed, for the reasons already stated.

and the use of the "9" in the first digit was not only as misleading as what nVidia did by labelling the GeForce4 MX as a GeForce4, but I consider it worse because ATI previously called nVidia to task on the 4MX naming, stating their own naming scheme formally in that presentation.

In other words, you completely ignored the provided argument to the contrary and re-stated your viewpoint, as if I misunderstood it the first time? Responding "never" to my question at the top of the page would have been more brief, and introduced exactly as much new information and logic.

Edit:
One final note. This isn't about excusing nVidia.

And your evidence to counter the lack of apparent accuracy or coherence in your statements so far is...?

They, however, have corrected their mistake with the GeForce FX line's naming scheme.

Hmm...you seem to be dedicated to expanding the conversation to defend nVidia on new fronts at every turn. Wow.

The 5200 naming is another naming scheme I will defend (and have already defended) as not being fair to compare to the GF 4 MX. Please look at the linked thread for the reasoning...it doesn't just apply to ATI. You should search on my name if you want further details as to my reasoning .

One can only hope that both companies do as good a job as nVidia has done with the FX line for future products.

Well, the 5200 does have a problem with performance, especially acute when actually delivering "PS 2.0", and some people would say it "doesn't run DX 9 properly". However, that problem is not the same as the GF 4 MX problem, and should be discussed without being equated to it...like the 9000 should. Hence why the "DX 9 compliant" label for the 5200 is a label it is entitled to, but the "Complete DX 8 support" label the GF 4 MX shares with the GF 4 Ti cards (in nVidia's product reference PDFs, which is reflected in specs listings where the cards are sold) is a problem. Did I contradict this line of reasoning somewhere? Did you miss the other times this was discussed? If so, why do you insisting on continuing to miss them when links to them are provided?

I agree with your praise of the GF FX line naming within its technology generation...in particular the supposed "5900 Value" type label (if it makes it to market as I understand it will), which even fits some ideas I proposed as to how ATI should have labelled the "9500 Pro".

But what does this have to do with the GF 4 MX, aside from both being products from nVidia? What does any of this have to do with the 9800 SE except to rewrite history to favor nVidia's comparison to ATI at every new opportunity, without regard to little things like "facts" and "logic"?

 

[demalion]

Or maybe you're just stuck on "technology" being API features offered rather than the technology used to implement them?

What's wrong with being stuck on that ? After all, does the consumer care about "the technology that's used to implement them" ?

It's not wrong to be "stuck on that", it is wrong to say that ATI's "technology generation" doesn't make sense, as he did. Did you read the posts involved?

The name's are for marketing and thus, they have nothing to do with the underlaying technology and everything to do with features/performance/price.

Yes, marketing is to do with what is delivered to the consumer...in this case, the new technology wasn't focused on new features, it was for significantly reduced cost at the same time as retaining thefeatures and majority of performance of the previous high end card offered. Is there a mystery here? Does the new bottom card in the line up being comparable to the high end of the prior generation seem insignificant to you? Do you continue to propose that is the same thing as the GF 4 MX?

To the heart of the matter: Is the "DX 8" listed for specs for the 9000 equal to the GF 4 MX listing "DX 8"? If not, why do you insist on equating them; If so, please explain. Saying "technology generation" means "marketing" and has "nothing to do with underlying technology" does not cut it, strangely enough.

Really...how does this not have to do with the underlying technology? How exactly is it you propose the features/performance/price you are listing are being delivered?

And thus, (all imo of course) they're wrong

Agreed, if by "wrong" you mean "there are problems with the naming".

for pretty much the same reason that the GF4 MX was wrong.

Makes no sense whatsoever. We've discussed this...do you have some new reasoning to offer?

 

[Bjorn]

It's not wrong to be "stuck on that", it is wrong to say that ATI's "technology generation" doesn't make sense, as he did. Did you read the posts involved?

I happen to think that it's correct to say that ATI's "technology generation" thing isn't correct.

Yes, marketing is to do with what is delivered to the consumer...in this case, the new technology wasn't focused on new features, it was for significantly reduced cost at the same time as retaining the majority of performance and features of the previous high end card offered. Is there a mystery here?

Yep, there is. Because the naming convention is still just as wrong or isn't it ?

Ati mentioned "technology generation" specifically with regards to how wrong it was of Nvidia to name the GF4 MX as a GF4. But what is wrong with the GF4 MX name if we're only talking about the low level technology and not the actual features ? Wasn't it released at about the same time as the GF4 ? Doesn't it have some features (should say technology since they AFAIK, they took the men controller and MSAA unit from the GF 4) from the GF4 ? (MSAA f.e)

Nvidia focused on FSAA, Ati on cost reduction. And both named their products wrong.

 

[demalion]

Is there any chance we could get back to discussing the 9800SE? I think there are some useful things to say with how ATI should direct their naming in future and the problems with the name from the thread topic. Their response to the LE problems on the cards they themselves produces was positive, but I think they're failing to provide a clear path for their OEMs to avoid repeating their own history.

How about some confirmation on whether this is part of an "official" naming policy, since that seems more likely now?

How about some info on whether my impression as to their being two "9800 SE" cards is correct or not?

I actually think an 8 pipe R350 with a 128-bit bus makes sense as a "9800 SE"...at least, after "SE" is clearly established as an indication of reduced performance (which it is not, IMO, at the moment).

I think the 4 pipe part is a complete disaster, though...as much of a failure in naming as the 9500 non Pro seems a success to me, even counting the F-Buffer. Unfortunately, I'm getting the impression that the 4 pipe part is far more common. Perversely, I think it could have been called a "9500 Pro" far more consistently, if that name hadn't been taken.

 

[Bjorn]

I actually think an 8 pipe R350 with a 128-bit bus makes sense as a "9800 SE"...at least, after "SE" is clearly established as an indication of reduced performance (which it is not, IMO, at the moment).

I agree. Edit: Although i have some problem with the SE part. SE usually means special edition or something like that and i definitely don't like them to use that. Didn't they use LE for slower versions before ?

I think the 4 pipe part is a complete disaster, though...as much of a failure in naming as the 9500 non Pro seems a success to me, even counting the F-Buffer. Unfortunately, I'm getting the impression that the 4 pipe part is far more common. Perversely, I think it could have been called a "9500 Pro" far more consistently, if that name hadn't been taken.

How about 9500 Pro Special Edition

 

[demalion]

It's not wrong to be "stuck on that", it is wrong to say that ATI's "technology generation" doesn't make sense, as he did. Did you read the posts involved?

I happen to think that it's correct to say that ATI's "technology generation" thing isn't correct.

And you don't make sense when you say that. I ask you, as I asked Chalnoth, to please don't say something as silly as "technology doesn't mean technology" and think the statement makes sense as it stands.

Yes, marketing is to do with what is delivered to the consumer...in this case, the new technology wasn't focused on new features, it was for significantly reduced cost at the same time as retaining the majority of performance and features of the previous high end card offered. Is there a mystery here?

Yep, there is. Because the naming convention is still just as wrong or isn't it ?

You keep saying "just as wrong", but I don't think you are really stopping to think about what the phrase means and that you are simply repeating something I just specifically addressed.

Ati mentioned "technology generation" specifically with regards to how wrong it was of Nvidia to name the GF4 MX as a GF4.

No, they mentioned "technology generation" specifically with regard to how they themselves were naming their parts. You keep ignoring my direct questions to you that would avoid us going in circles about simple things like this.

But what is wrong with the GF4 MX name if we're only talking about the low level technology and not the actual features ?

The features being labelled as "complete DX 8 support", the same as the GF 4 Ti, for one. The features not being the same as the GF 3 that was the previous generation's high end card, for another. Both things that differentiate the GF 4 MX from the comparison of the 9000 to the 8500, as already stated. All of this is pretty glaringly obvious, and I think for any hope of brevity I'll have to simply ignore the next time you ask me something I've already said.

Wasn't it released at about the same time as the GF4 ? Doesn't it have some features from the GF4 ? (MSAA f.e)

OK, this is going nowhere, please refer to my prior post and request for being on topic, and please excuse me if the "princess brideism" opportunity I tried to avoid above is still peeking through.

 

[demalion]

Bjorn: I think you're being too kind to the 9800 SE name in that latter post. That's why I said I don't think "SE" is good now. nVidia and ATI seem to be both working to use it in other product segments to communicate "reduced", so it might mean it in the future somewhere...but it still sucks right now.

(For the 4 pipe part) 9500 Pro SE still wouldn't cut it, I think. 9800 "some fancy marketing term that means 'half'" might, but SE is just way too ambiguous.

9800 Redux? That could make sense in a marketing way. 9800 Lite could even make sense, because that so strongly communicates that the consumer is getting something that is the 9800 but crippled, without stating "crippled". "LE" would also have worked much better, and ATI is already familiar with it...which is why I think OEM's demands are driving this and illustrating ATI isn't exercising enough naming protection at the moment. I do hope they learn before they compound their mistakes further.

 

[Bjorn]

No, they mentioned "technology generation" specifically with regard to how they themselves were naming their parts. You keep ignoring my direct questions to you that would avoid us going in circles about simple things like this.

Yes, and as i have stated already, that's wrong since the consumer won't care about the low level technology and only about the techonlogy they actually get in their hands. Which is API features, performance. That's why i don't accept the "technology generation " thing. You might say that technically, Ati didn't do anything wrong. But i don't agree. Especially since they mentioned this only to slam Nvidia.

The features being labelled as "complete DX 8 support", the same as the GF 4 Ti, for one. The features not being the same as the GF 3 that was the previous generation's high end card, for another. Both things that differentiate the GF 4 MX from the comparison of the 9000 to the 8500, as already stated.

Sure, there are degrees of evilness but the naming scheme is just as wrong.

And you're right, this isn't going anywhere maybe we should contine with the 9800 discussion instead

 

[Bjorn]

Bjorn: I think you're being too kind to the 9800 SE name in that latter post. That's why I said I don't think "SE" is good now. nVidia and ATI seem to be both working to use it in other product segments to communicate "reduced", so it might mean it in the future somewhere...but it still sucks right now.

Was i really that kind with regards to the "SE" thing ? I specifially said that i didn't like it and that "SE" to me sounds like it should be a "better then normal" part.

(For the 4 pipe part) 9500 Pro SE still wouldn't cut it, I think. 9800 "some fancy marketing term that means 'half'" might, but SE is just way too ambiguous.

I agree. The 9500 Pro SE was just a joke. But the biggest problem with all these extra letters isn't they can be somewhat misleading but rather they they're not always clearly visible on the box (as you mentioned below also).

9800 Lite could even make sense, because that so strongly communicates that the consumer is getting something that is the 9800 but crippled, without stating "crippled". "LE" would also have worked much better, and ATI is already familiar with it...which is why I think OEM's demands are driving this and illustrating ATI isn't exercising enough naming protection at the moment. I do hope they learn before they compound their mistakes further.

Lite or LE is definitely what i would prefer also. Ati's marketing team probably don't agree though but i hope they prove me wrong. They've done everything right when it comes to hardware and drivers lately (well, of course not everything but you know what i mean ) so hopefully they continue this trend in other areas as well.

 

[incurable]

incurable, what I was trying to say exactly what I stated. Note the series of statements and questions. A conversation is where you respond to them, not assign me a motive and dismiss what I've said based on it.

I'll progress the conversation on your terms, since you seem to be caught up in a preconception of my stance that I can maybe dispell right now:

It's clear from your previous post that your definition of technology generation doesn't overlap very much with my own, but that's ok, we'll just have to agree to disagree there.

Therefore, I'll only ask one question:

How does the 9100 fit in your argument. (I'm sure you agree that R200 can't be a generation ahead of R200. )

The 9100 was a relabelling based on an ill thought out approach to resolving exactly the problems I already referred to with regard to the 9000 (plus the problem of excessive 8500 stock, apparently). Namely, that performance level for the 9000 was inferior to the card that was released before it that had a "lower number" associated with it. As I said already, this is indeed a problem, it's just isn't the problem that you described. If you choose to continue a line of criticism on the 9100, my only concern would you keep it on topic...I've already chipped in my own criticisms against the name in a more suitable thread.

Please take a gander at the thread I linked above, as I've specified the reasoning behind this all in quite a bit of detail already within it.

What I think you were after is for me to simply say that ATI did something wrong, that the 9100 is not a good name, and seems to violate ATI's own naming policy, which I agree with. Please don't confuse "Don't criticize ATI for reasons that don't make sense" with "Don't criticize ATI", which is the only thing that comes to mind that would explain why you'd think I'd say otherwise.

So the 9100 doesn't fit with ATi's own naming conventions, great, we agree.

I even think that the 9800SE is worse name than the 9100 (atleast for the 4 pipe one), which is why I don't know why people who dislike it insist on bringing up the GF 4 MX when there is plenty of on topic and accurate criticism to bring forth, especially if this is now an official naming policy (it not being official naming when originally brought up a while ago was about the only saving grace I found in the name for the 4 pipe part).

Would you please leave out the GF4 MX references? I never brought that up nor did I compare ATi's apparent problems adhering to its own naming conventions with it. Thanks.

IMO the 9800 SE is just yet another product carrying the wrong name, just like all 9000, 9100, 9200 and 9600 series cards.

cu

incurable

 

[KimB]

I already provided my reasoning with regard to "technology". Simply saying "that is not so" does not do anything to answer it. The 9000 delivers the same old features, using new technology and less transistors. Please don't tell me silly things like technology does not mean technology.

And the GeForce4 MX offered new things like multisampling AA and anisotropic filtering, as well as a number of video features. Obviously you don't think that those qualify it for a GeForce4 name. Why would you think that a "streamlined" R2xx core would qualify the 9000 for a "9" as the first digit in its number? From what I've seen, the GeForce4 MX offered quite a bit more in terms of new features.

But that doesn't matter. Technology generation has always referred to the programming interface, nothing else.

ATi can't decide how to assign their own technology generations, even when there is logic and evidence to support their choice? Chalnoth, that doesn't make sense.

nVidia can't decide how to assign their own technology generations, even when there is logic and evidence to support their choice?

You're applying a double standard. I don't excuse the GeForce4 MX, and I don't excuse the Radeon 9000.

Yes, because it makes sense to rename things when you are...introducing a new naming scheme and applying it to your entire product line up.

No, it was convenient for ATI at the time. It was a way to lash out at nVidia.