Misplaced Pages

This is an old revision of this page, as edited by Intgr (talk | contribs) at 16:06, 21 December 2006 (→AVIVO: comment on vector processing). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

I think that ati should redirect here whenever you type ATI or ati. I assume that over 90% of people typing those keywords are looking for this article. I just changed the logotype by uploading ATI_Logo_0903.png If I should rather have deleted existing ATI logo.png and replaced it with the new image, say so (or just fix it). As it is, there are two ATi logo images on Misplaced Pages, one of which no longer matches the official one (and it had that ugly placement of (r) anyway). Paranoid 20:25, 16 May 2004 (UTC)

I've corrected the acronym (it's NOT Allied Telesyn International), added some extra bio info (location, names of founders). There is much more that can be said for this company, but I'll leave it to the vidiots. - quanta, 7 July 2004 This needs coverage of Mobility Radeon, and an explanation of why I can't see the model number from Apple System Profiler... lysdexia 07:43, 23 Oct 2004 (UTC)

It seems to me that this page was written by the marketing department. Does it seem this way to anyone else? I hesitate to call a NPOV violation but the wordage seems better to be in a flyer than Misplaced Pages. Chotchki 04:29, 7 Feb 2005 (UTC)

I agree on this. It just reads horrible. Joachim.Kluge 00:12, 6 December 2005 (UTC)

What's with the "NOTE: If you are planning on using or using linux (OS) Its not worth the hassel to get an ATI graphic card since there poor drivers arent real good. If you plan to use linux I would recommend an Nvidia Graphic card." at the end? I recommend we delete it, it seems just a teeny bit unprofessional. --Easty 13:25, 21 Feb 2005 (UTC)

I am wondering if calling the Radeon Xpress 200 ATI's first chipset is really true. I seem to recall a few motherboards out there boasting ATI chipsets (such as the 320M for the regular Athlons and the 340M for Pentium 4's and more recently the 9100 IGP for the Pentium 4's) while I guess the 3x0 series might not qualify as chipsets in the truest sense of the word (as, I seem to recall the boards using southbridges from VIA or some other manufacturer) I'm pretty sure the 9100 series had southbridges manufactured by ATI, thus making it ATIs first true chipset. For now I'll just leave the article as-is because I may very well be wrong but someone who has more time on their hands may want to look into the issue a little more. - Kilroy1231 3.3.05

ATI used an ALI southbridge for the earliest chipsets, but that changed with the 9100. Since I wrote 30% of the present ATI page, excuse me if I have no hesitation in correcting this nonsense. Well spotted. Okay, I just split motherboards out to their own section. Its just a stub, but its a start. When I get the time /energy, I still want to fill back the current trends section to at least the Rage chips. Timharwoodx 20:36, 12 Mar 2005 (UTC)

Okay, I've now changed the layout, to fit with the one used on the nvidia page. ATI and nVidia are really twin companies these days, doing the same sort of things, so I think it makes sense to sync the layout for consistency of look and feel. Timharwoodx 14:02, 13 Mar 2005 (UTC)

I just put in a vote for deletion for a new page on the 9250. Jeesh, here is the rule guys. We do a page for each NOTABLE series of cards. The same rule for both ATI and nVidia. Copying out benchmark scores and spec sheets on every part nVidia or ATI ever made, really is not what an encyclopedia is about. Timharwoodx 20:26, 11 Jun 2005 (UTC)

Hi, I put a ATI makes crappy cards section. If anyone wants to challenge me in blowing over 500 dollars on defective cards go ahead. Its no coincedence that their cards glitch up...unless I have hidden mind powers that somehow break the things I buy. Marvelvsdc 01:26, 04 July 2005 (UTC)

This gen ATI seems to be faltering, which is why I've decided I'm going to upgrade to a 6800 ultra instead of sticking with ATI.Heraclius 01:30, 11 July 2005 (UTC)

Why does this page ignore ATI's past mistakes/failures/questionable actions yet it lists Nvidia's?

This article should be about ATI, why is every other sentence I see nVidia? Not just see nVidia - but I see over-blown publicity. This article seems a bit biased, is it really necessary to mention nVidia's name so much?

ATI also made a sound card called the Stereo F/X in 1992.

Well, how do you discuss ATI without reference to NVIDIA? Its like asking someone to write a history of AMD, without any mention of Intel. Thats an absurd point. Then you say its like a fan page? Well, on the one hand you say its all about NVIDIA, then in the next sentance you seem to be saying it goes on and on about ATI. A wholly inconsistent set of observations. ATI has been the world's largest discrete graphics card company for the last 2 years, the products must have been pretty wonderful in that period, to beat out a company as tough as NVIDIA. Are you going to argue the 9700 and 9800 were bad products? Because I think you'll get flamed like crazy if you edit the page to say that. Timharwoodx 14:19, 12 May 2006 (UTC)

XGI

I think the purchase of a lot of the assets from XGI should be mentioned somewhere in this article.

Hollywood

NOTHING has been confirmed about this, other than it is NOT based off of any previous GPU or GPU series. ~Captain

ATI lack of drivers for linux ?

I don't think this : http://www.petitiononline.com/atipet/petition.html is 24.470 times my own POV. 82.244.80.175 12:09, 14 June 2006 (UTC) and also user:Yug

http://www.petitiononline.com/mod_perl/signed.cgi?atipet

Just because a lot of people signed some petition does not make your assertions correct. "Linux + an ATI video card = a really slow computer"? No, I have a computer right here to disprove that. I rewrote it into a neutral point of view. Haakon 12:16, 14 June 2006 (UTC)

For me, 24.470 signing people is not "some", and many Linux/ATI users don't know about this petition, so that probably talking about 200.000 to 1.000.000 of users. Moreover, I have a Radeon 8500 mobility and Linux, and the fact is just in front of me for the last 2 years. Afterwhat went I see and read a petition like this, I think you are really nice with ATI ("ATI's inadequate drivers" : I understand that ATI did nothing, and that Linux programmer try to fix it with "inadequate" things.).

I leat this like this, and just put this discussion in the article discussion page. Yug

I do not use the proprietary drivers (by principle), and my computer does not feel slow (I don't play 3D games), so "Linux + ATI = really slow computer" is not categorically true. I do think 25000 users qualify as "some", considering the many millions who are Linux users. Speculation on how many could potentially sign the petition are just that; speculation. I think the section is now on a form we can all agree on. Haakon 13:15, 14 June 2006 (UTC)

Too much plain text

Is anyone else a bit annoyed by the large field of text that comprises the Market trends section? Is there some sort of "Wall of text" template that can be applied to the page? I'll try and get creative to write some titles for the section, but still.ModusOperandi 02:51, 18 June 2006 (UTC)

Yeah it does look a little plain. Maybe a picture of some kind could spice it up? PrettyMuchBryce 03:50, 21 June 2006 (UTC)

ATi is not the proper spelling of the company's name!

The old logo might have looked like ATi, but the formal name has always been ATI. ATI stands for Array Technologies Incorporated. Get it right! :P Alex 19:23, 24 July 2006 (UTC)

And here's proof: Bilky asko 16:36, 10 November 2006 (UTC)

Merger or AMD buying AMD

I don't know much about business or economics and the wiki article is not particularly clear. At one place it suggests it's a merger, at another place it suggests AMD is buying ATI. So is this a merger or is AMD buying ATI? According to the AMD page it's a merger but according to Yahoo AMD is buying ATI. Also the Yahoo page states it's subject to ATI shareholder approval but doesn't mention AMD shareholder approval so again it makes me think AMD is buying ATI (although maybe they just didn't mention that shareholder approval was needed). Nil Einne 20:50, 24 July 2006 (UTC)

Mergers are almost always one company buying another company, and in this case, I believe it's AMD buying ATi. So yes, it's a merger.

Not according to the Mergers article however which states the differences between mergers and acquisitions is in how they're financed... Nil Einne 15:54, 20 September 2006 (UTC)

Old logo

Is there any information on the old logo, such as why they switched to the current logo? If the old logo was shown on this article page, I don't think that removing it was a good idea. Rather than that, there should be a comparison, such as how there is a comparison in the Microsoft article, for example. --Geopgeop 10:26, 1 August 2006 (UTC)

Chipsets

Maybe it's time to start moving motherboard chipsets to codename series, much like the Radeon? There are going to be lot more products than just Xpress 200 and 3200 as outlined in {{ATI}}, and all sorts of fancy codenames like RD480 (Xpress 1600), RD500, RD600, RS500, RS600 etc. --Dmitry _{(talk •contibs )} 18:52, 2 August 2006 (UTC)

Big Picture.

And I'm not talking metaphorically. There's a huge, obnoxious picture of a Radeon X1900XT in the middle of the page! Someone should format it to Misplaced Pages styling (I would, but I don't know how yet) or remove it. (edit: forgot to sign) --CCFreak2K 09:03, 4 August 2006 (UTC)

Done'd. -AthlonBoy 13:57, 9 August 2006 (UTC)

Reduntant acronymn - ATI Technologies Inc?

Reading that first line is a bit disorienting. Array Technologies Incorporated Technologies Incorporated? It seems this article is suffering from RAS syndrome.

Yes, I know that ATI technologies isn't redundant, as that implies a technology of ATI. But ATI Technologies (with the capital letter) implies it is part of the name. Maybe the article ought to be moved to, lets say, ATI (Company), or the like. -AthlonBoy 12:30, 9 August 2006 (UTC)

x800 series?

Article jumps directly from the 9xxx to the x1yyy series. I don't know much about those, so could anyone add them? Pmbarros 21:46, 15 August 2006 (UTC)

IIRC, the x series (pre-x1000) is an extension of the 9xxx series. For example, a Radeon X800 would be comparable to a Radeon 9800, but the X800 would have...more, I guess. I won't add anything to the article because I lack sufficient information. --CCFreak2K 11:59, 28 August 2006 (UTC)

The X800 was nothing more than a 9800 with double the pipelines and a bit of extra MHz. There were some non-specific enhancements, but nothing really new. A more detailed explanation is on the Radeon R420 page. -AthlonBoy 16:38, 28 August 2006 (UTC)

'a 3D accelerator for free'?

Regarding the Mach64, this sentence is difficult to understand. Is the article saying that the video card was so inexpensive, or is it referring to some 3d acceleration features that neither of the Mach64s I owned had?

Stallman incident should be removed

The paragraph about the April 2006 incident involving RMS should be removed. If nothing else it should be moved to either the article on RMS or the FSF, but I would argue it is not encyclopedia material. -66.226.105.98 07:49, 17 December 2006 (UTC)

I agree. I'm the guy who wrote most of the ATI, NVIDIA, AMD, and other key IT pages. Its just too much detail for a summary page. Sadly, we have someone called User:MureninC going round the IT section vandalising pages at the moment, and none of the editors seem willing to do anything about it. He routinely breaks guidelines, and does not adhere to / respect consensus. He just reverts when I try and fix things, and correct the factual errors he has inserted. HELP! I'll try deleting the incident (AGAIN) and see what happens. Its the duty of whoever reverts to post to the discussion page, if not, I'll have to take this up with a senior editor, with a view to having the User:MureninC account suspended. Timharwoodx 15:13, 17 December 2006 (UTC)

Example of User:MureninC at work.

• http://en.wikipedia.org/Talk:VIA_CoreFusion

Another mess I'm trying to sort out! Timharwoodx 15:33, 17 December 2006 (UTC)

Sadly, we have Timharwoodx trying to use personal attacks in order to push for ATI and AMD — no consensus was ever reached in removing the Stallman incident. Everyone, please take a look at Misplaced Pages:Articles for deletion/EPIA and see for yourself. MureninC 16:33, 17 December 2006 (UTC)

Presently we have two registered votes for deletion of the Stallman incident, and no registered against. Yes, by all means look at the EPIA page. See how MureninC called my VIA C3 page 'the worst page in the WIKI' - slandering my hard work, then view the page history, and see how he inserted factual errors into the article, got core types utterly confused, mixed up max TDP with average TDP, and then how he called my attempts to edit his ungrammatical and factually inaccurate VIA_CoreFusion page 'vandalism.' Something that remains unresolved i.e. the WIKI is presently peddling information I know to be wrong on the VIA_CoreFusion page. This guy is so clueless as to be positively dangerous. What gets me even more, is that having been proven to be wrong time after time on technical issues, he keeps coming back, challenging me to reply to him. I try top walk away, and he keeps posting, as if craving attention from me. MureninC why do you keep posting? What contribution to the MIT debate did your last post make? Answer - none at all. As usual, you're just trying to get attention. Either debate the facts, or go away. You slandered my work, then added factual errors not present in my text. Thats where it started. Timharwoodx 23:19, 17 December 2006 (UTC)

Yeah, I just checked the page history, MureninC has just put the MIT incident back in, despite the vote presently being 2:0 for it to be removed. This user ignores consensus, and arbitrarily edits pages. Its a pattern of vandalism across multiple IT pages. I'm going to remove the incident one more time, reflecting the 2:0 voting presently demonstrated in this matter, not to mention the non encyclopedia nature of the material anyways. If MureninC YET AGAIN ignores consensus, this will have to be escalated for attention from users with higher privilege levels. This is ridiculous. Timharwoodx 23:37, 17 December 2006 (UTC)

Yep, MureninC reverted (again) within minutes. He seems determined some incident on the MIT campus deserves pride of place in a summary of ATI's corporate history. He obviously has this page on his watch list, and is determined that University campus incidents should be a major part of corporate histories in the WIKIPEDIA. I'll wait for more opinions to be posted on this matter, then seek escalation, as and when appropiate. Timharwoodx 00:19, 18 December 2006 (UTC)

Timharwoodx, I'm tired of your disruptive behaviour. What consensus are you talking about? What 2:0? There are already two users wanting to keep the article. Me and 72.173.112.236 . The content is perfectly encyclopaedic. If you want it deleted, organise a proper vote for one week, mentioning the fact that the vote is going on the subsection page. And while we are at it, your article about VIA C3 is still very poor. MureninC 00:23, 18 December 2006 (UTC)

Timharwoodx, your edits are perfect as usual. You've put the Content template into the wrong subsection of the article! It should be inside the section that you want to have removed. Until then, the voting would be unfair, as people would not know what they have to vote for. MureninC 00:43, 18 December 2006 (UTC)

Also, Content doesn't seem to be the appropriate template for this request of yours. MureninC 00:46, 18 December 2006 (UTC)

Please stop discussing and challenging each other's competence and/or previous edits, and concentrate on resolving this problem. Timharwoodx's comments about MureninC's previous edits are clearly irrelevant, as the section is written in a good encyclopedic style, and does cite reliable sources on the issue. What is being contested by the initial comment is the relevance of this section, and not its style or factual correctness.

In my opinion, ATI's neglect for the free software community is very relevant to the article, but the mentioned section does not convey that very well. I think the section should be rewritten to explain how ATI is said to neglect the free software community, and mentioning Stallman's incident as an indicator that a conflict exists. -- intgr 05:04, 18 December 2006 (UTC)

Even Stallman wouldn't think it was important if a cop hadn't shown up, but ATI had nothing to do with that. The people in charge of the event tried to throw him out and he wouldn't go so they called the cops. He's trying to spin it into a "corporate fatcats abuse their power to oppress the little guy" case when it's really just a matter of, he was making a nuisance of himself and wouldn't leave so they called the police. Happens every day in Wal*Mart parking lots across the country. Does NPOV not apply when it's inconvenient for Misplaced Pages's pet causes? -66.226.105.98 14:16, 18 December 2006 (UTC)

You may consider the Stallman's incident alone unworthy of mention, but you cannot deny the conflict between the free software community and ATI. -- intgr 15:12, 18 December 2006 (UTC)

I find the whole open driver cause to be controversial. On one hand it allows people to write free drivers. On the other hand, it exposes the IP of ATI. It is a great threat if say NVIDIA can gain advantage if that open code shows them things they didn't know. It also exposes some of how the hardware functions. That risks jobs. Not just evil greedy manager people jobs, but the jobs of the hundreds of engineers who toiled to produce that code. That is not a worthwhile risk IMO. ATI does do Linux drivers (Linux being where open code is most demanded), and they are getting better.

Which regards to this Stallman guy, some guy being a jackass at a conference is not worthy of its own paragraph. Sure, it could be referred to in the context of ATI's closed code practice. But right now it sticks out as if it's some huge issue and just casts this strange "ATI is evil" feeling over the whole article. And I don't think that such spotlight exposure is "relevant" or really proper for NPOV encyclopedic content. Honestly, a sentence saying ATI doesn't offer open source code is enough for me. --Swaaye 21:31, 18 December 2006 (UTC)

Obviously it's controversial, which is why it's relevant. For the record, they're not demanding open source drivers, they're demanding documentation. -- intgr 07:08, 19 December 2006 (UTC)

The documentation that is required in order to write a fully functional driver DOES NOT reveal any details on how the hardware itself is designed, it only tells you how to communicate with the hardware effectively. Are jobs in the car manufacturing industry lost because of the repair manuals that show how to repair a car? ATI releasing programming documentation does not provide any advantage to NVIDIA — in fact, it'll only be disadvantageous to NVIDIA, because it would mean that ATI is going to be fully supported by every FOSS distribution out-of-the-box without any user configuration being necessary. See Vendor reasons and excuses; also see Users defending vendors on the same presentation. MureninC 03:31, 21 December 2006 (UTC)

"Stallman wouldn't think it was important if a cop hadn't shown up"? That's a really interesting suggestion. Why was he organising the protest, then? MureninC 03:36, 21 December 2006 (UTC)

Perhaps I should have said "Noteworthy". Does Stallman update Misplaced Pages every time he organizes a protest? The FSF protests a lot of things without making notes on wikipedia about each one. -66.226.105.98 05:58, 21 December 2006 (UTC)

AVIVO

Look, I know I created the AVIVO page, but stream processing is the AMD brand for AVIVO. Its the same thing. Programming shaders to do CPU like tasks. This is a summary page, if AMD are branding CPU like functions as 'stream processing' then AVIVO is no longer relevant. Its an expired marketing term. ON a summary page, I don't see space for an expired marketing term from the (old) ATI company. Its just not relevant any longer. If you disagree, post reasons, and form a consensus before reverting the main page. As the primary author of this page, I think I deserve a voice. Timharwoodx 15:21, 17 December 2006 (UTC)

Quote: "but stream processing is the AMD brand for AVIVO"

NOT true. My two questions:
1. Did somebody mentioned AVIVO supports "Hardware-Assisted" MPEG2, H.264 and VC-1 (WMVHD?) decoding?
2. Did anybody mention the "stream processor" does the work as hardware assisted decoding of MPEG-2, H.264 and VC-1 videos? (Or should I say "use"?)

Or, you mean that "Stream processor" is a product for video decoding?

"Stream Processing" from "AMD" provides a "thin hardware interface" named as "Close To Metal", it allows programmers to write codes to use the instructions inside the GPU, and what does AVIVO has to do with THAT (other than decoding videos)!? (AVIVO though has a "converter"... A video converter)

It is known that AVIVO is an accelerator on video playback (yeah, the ATI saying), and "Stream processing" is simply using GPU to do (32-bit?) precision floating-point calculations, and I personally do not think they are the "same thing".

BTW, the AVIVO capabilities are provided using a external chip (named as "Theater"? I forgot), which I think "AMD Stream processor" do not need that in the FP calculations. And currently AVIVO is still being implemented on graphic cards that branded as "ATI", from X1900 series to X1300 series.

• And some references for you:
  

1. AMD "Stream Processor",
2. "Close to Metal" and
3. AVIVO press release

P.S. I know this is equal to flaming, but I want to point out that even though they are marketing terms (new or old, whatever), but I can surely say they are not the same, with reason that a "Stream processor" is not of the same use as a "Video accelerator" and vice versa. --202.71.240.18 13:56, 21 December 2006 (UTC)

Timharwoodx: Note that you're confusing "vector processor" and "CPU", too. While a CPU is capable of dealing with highly dynamic tasks, vector processors (such as the GPU) are only useful for applying (rather) simple operations on large datasets. So they're by no means "CPU-like".

The comment "That is to say the computation complexity of GPUs has moved so far from the fixed functions of Direct-X 7 requirements, that actual code programming is now possible, for specific application scenarios." is awkward at best – while GPUs have indeed became more complex, they're nowhere near the flexibility of a CPU. And a chip does not have to be complex simply to do vector processing, they just need a larger number of dedicated pipelines. It's just the fact that vendors haven't officially exposed the vector processing capability to third party programs running on the CPU until now, AFAIK. -- intgr 16:06, 21 December 2006 (UTC)