Talk:AMD: Difference between revisions

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Browse history interactively
← Previous editNext edit →
Content deleted Content added
VisualWikitext
Line 166: Line 166:
:::As for the "near zero" quote, then it needs to be clearly stated that this was only for Spectre 2. Otherwise this is unclear and misleading.
:::As for the "near zero" quote, then it needs to be clearly stated that this was only for Spectre 2. Otherwise this is unclear and misleading.
:::And why bring up Intel? That is irrelevant here. ("Pepsi is a popular cola" cannot be included excluded simply because Coke is popular too.) ([[User talk:Dbsseven|talk]]) 21:11, 12 January 2018 (UTC)
:::And why bring up Intel? That is irrelevant here. ("Pepsi is a popular cola" cannot be included excluded simply because Coke is popular too.) ([[User talk:Dbsseven|talk]]) 21:11, 12 January 2018 (UTC)
::::AGAIN, I provided six sources that say AMD changed its position. And you were the one that brought up Intel. This is boring. The absurd, incorrect, 30%, scary as Hell number was forced into the Intel article and you look like you're trying to whitewash AMD's problems and responses in this article. Intel and AMD BOTH have major flaws with the same dangerous effect. Let's not dump on one and act as apologists for the other. [[User:Objective3000|O3000]] ([[User talk:Objective3000|talk]]) 21:23, 12 January 2018 (UTC)

Revision as of 21:23, 12 January 2018

See comprehensive edits below from Michael Silverman

Please provide sourcing. --JustBerry (talk) 22:17, 26 May 2014 (UTC)[reply]

Key staff changes

Shouldn't there be a section about staff changes/resignations, when it seems like they are quite major (long time AMD, ATI staff replaced by ex-IBM people?)

2002 Sanders -> Hector Ruiz

"Ruiz endorsed the decision to buy ATI, which led to a period of financial reverses. Ruiz survived rumors of his ouster in late 2007.[3][4] However, he resigned as CEO on July 18, 2008, after AMD reported its seventh consecutive quarterly loss" [5] (ref in Hector Ruiz

2007 EVP of AMD/ex-CEO of ATI David Orton resigned "It is with mixed feelings that I am leaving AMD,"

January 2011 CEO Dirk Meyer, father of Athlon "resigned in January under pressure from AMD's board" http://eetimes.com/electronics-news/4219307/AMD-appoints-former-Lenovo-exec-CEO

Fall 2011 Rick Bergman, head of Products Group, ex-ATI

Dec 2011 Nigel Dessau, marketing chief

Feb 2012 Eric Demers (Corporate VP & CTO, Graphics Division)


This is from comments on another site: "month after they fired Meyer I can see why they fired/pushed out his VP & COO/CIO/whatnot Rivet, along SVP Strategy head Seyer"

"now everything is headed by Papermaster, a guy who took the SVP & CTO of AMD as his 4th job within 4 years at 4 companies (pre-AMD was Cisco, Apple and IBM, in reverse order, all since 2008) and keep hiring his ex-IBM buddies. Of course, he was hired by CEO Rory Read who is a 20+ years IBM veteran"

Trademark of Numbers

The article asserts "U.S. Trademark and Patent Office had ruled that mere numbers could not be trademarked.[109]" The link is to an article in the New Yorker which states: "after finding that it was impossible to trademark a number..." I think this is kind of weak as a source, and I believe it to be a stretch of the truth, perpetuating a myth that's been around AMD processors for decades now. Here's a list of numbers registered as trademarks, with citations. Also, here's another that came up in a quick google search. Further, the mark 586 is trademarked by Intel!

Thanks! — Preceding unsigned comment added by 2601:880:c000:3062:19fd:52de:c139:d7e7 (talk)

CPU Cleanup Needed

Some CPU lines replaced others. Further, some CPU lines have been discontinued. Please point these out accordingly & have the template box up to date. — Preceding unsigned comment added by 73.140.208.59 (talk) 04:01, 30 June 2017 (UTC)[reply]

Organization and cleanup

Hi all, I'd like to clean up this article and have a few ideas. Any input before I get started would be greatly appreciated. I will try to incorporate the other notes here on the talk page not already addressed.

  • It seems like AMD's discrete GPU business deserves its own second level section rather than falling under 'Other products and technologies' as this is a major part of their business.
  • Platforms should be organized into their own sub-level as this is a reasonably specific topic.
  • A specific discussion of the 'semi-custom' business that is a major revenue stream for the company.
  • The 'Embedded graphics' table should be moved to a more appropriate list/table article. This is pretty detailed data on a single product line for an article about the business as a whole.
  • Technologies are scattered within the article and should be organized into a single section.
  • More detailed sub-levels in history.
  • Updates as appropriate.

I have no intention of deleting any content, mostly moving content within the article (and perhaps moving specific detailed information to other articles). Dbsseven (talk) 20:32, 4 July 2017 (UTC)[reply]

Relatedly, is there any consensus on how to determine if a product is discontinued? It seems to me we can be reasonably confident the K5, K6 (and others) are discontinued. But companies don't generally announce when a product is no longer produced, only when production begins. Any thoughts? Dbsseven (talk) 20:33, 4 July 2017 (UTC)[reply]
Okay since no-one piped up I am giving this a try. Some notes, I have tried to keep everything that was already in the article there only clean up. Except, I have removed the platforms section (very out of date and a constantly moving target) and the embedded graphics table (moving to a more appropriate page). I have tried to keep the text as prose rather than bullet points where it made sense. I have also added detail and cites to much of the article.
I realize this is a major change. Please leave your comments here. I am sure there is still work to be done but I believe/hope this is an improvement over the previous version. Best. Dbsseven (talk) 19:23, 11 July 2017 (UTC)[reply]

External links modified

Hello fellow Wikipedians,

I have just modified 2 external links on Advanced Micro Devices. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 16:35, 1 December 2017 (UTC)[reply]

External links modified

Hello fellow Wikipedians,

I have just modified one external link on Advanced Micro Devices. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 08:13, 21 December 2017 (UTC)[reply]

security vulnerabilities

@Objective3000: I see you have reverted my recent edits about the 2018 security vulnerabilities. I do not see why. You state "revisionist history" but I do not see how that is the case. Dbsseven (talk) 15:39, 12 January 2018 (UTC)[reply]

AMD clearly suggested that Spectre would not affect its processors and that it was an Intel problem, resulting in large moves in both Intel and AMD stock. Numerous articles appeared on how datacenters were thinking of switching from Intel and that this was a huge setback for Intel. Spectre wasn’t even mentioned in this article. Yesterday’s admission that Spectre affects virtually all of AMD’s chips for the last 15 years is a dramatic reversal. And the Intel article still makes the discredited performance claim of 30%, while this article makes no such ludicrous claims. O3000 (talk) 15:45, 12 January 2018 (UTC)[reply]
Objective3000 Your position of a reversal is in opposition to the cited source which states: "AMD's initial statement last week said, in part, that "AMD is not susceptible to all three variants." In other words, some of AMD's chips are vulnerable to two flaws -- just not all three." (And AMD's risk to Spectre is found in earlier sources too [1]) None of the cited sources state AMD "suggested" it was invulnerable to Spectre, or that this was a reversal of AMD's position. (Though it is a clarification of it's response.) As for datacenters switching to AMD, please provide a source because it is not in the cited sources. (And as we discussed before on the Intel talk page, stock pricing has nothing to do with WP or editorial policy.) Dbsseven (talk) 16:03, 12 January 2018 (UTC)[reply]
This is simply not true. On Tuesday the Register suggested that the flaw did not impact AMD's chips but instead primarily impacted chips from Intel. One of the many reasons I don’t like The Register as a source. But, this narrative was widely reported and AMD fell silent until yesterday when they reversed course. Numerous articles talked about the opportunity of the AMD CEO to take advantage of the "Intel flaw". Spectre comprises two of the three vulnerabilities, and is the most difficult to deal with as well as resulting in the largest performance hit. Spectre is also the only one that requires microcode changes. Until yesterday, AMD said nothing about a need for AMD microcode changes. I certainly don’t mean to cast aspersions – but an outsider reading these articles will come to the conclusion that we are downplaying the effect on AMD and exaggerating the effect on Intel. [2], [3]
Again, I do not see any source stating "reversal" (WP:PROVEIT). The Register is not cited here, nor has any credible source stated that The Register/CNBC speaks for AMD. AMD continues to state they are less at risk from Spectre: "AMD’s processor architectures make it difficult to exploit Variant 2" (the one where they're releasing microcode). Therefore, an assertion of a "reversal" is OR without a cite. That AMD "not speaking" was tacit approval of the narrative is without a source also (and may be mistaken). (And I would disagree that AMD didn't speak, sources clearly state they did, though the sources also state they should have been clearer [4]) Ptu another way: We need to be clear about what AMD said, and what was said about AMD. Dbsseven (talk) 16:38, 12 January 2018 (UTC)[reply]
If we want to include a discuss of market position/competition, that is reasonable, but that should be independent of AMD's risk to this vulnerability. Dbsseven (talk) 16:41, 12 January 2018 (UTC)[reply]
The market changes are due to the changes in AMD's position. AMD rebukes Intel, says flaw poses 'near-zero risk' to its chips.[5] The claim just wasn’t true. As for the claim that it might be more difficult exploit Variant 2, so what? It must still be fixed by both software and microcode updates and will still impact performance. Spectre affects Intel, AMD and others. AMD did not admit the scope of the problem until yesterday. [6] [7] [8] [9] [10] [11] O3000 (talk) 16:50, 12 January 2018 (UTC)[reply]
We must be clear about what AMD said versus what was said about AMD. Multiple of the source you cite there state "AMD said there was “no change” to its position on the susceptibility of its chips" and "...investors believed..." Again, clarification versus reversal. Dbsseven (talk) 16:59, 12 January 2018 (UTC)[reply]
Yes, AMD claimed there was no change in position. The sources I provided said there was a change in position. O3000 (talk) 17:06, 12 January 2018 (UTC)[reply]
I do not see a conflict between what AMD said then and now. Zero is not the same as near zero. Can we agree on that? (Similarly, I would assert that saying nothing is not the same as saying "we don't need to do anything".) Dbsseven (talk) 17:19, 12 January 2018 (UTC)[reply]
We need a consensus on language: Are we to discuss in this article about what AMD said, or what was said about AMD? (And if AMD "implied" something an explicit source is needed. I don't see how we can cite AMD not saying anything). Dbsseven (talk) 17:19, 12 January 2018 (UTC)[reply]
And this section is about security flaws, not market position. That is a fine discussion to have, but probably not in this section and runs the risk of WP:TOOSOON (just as discussing Intel losing market share would be too soon that page). Dbsseven (talk) 17:03, 12 January 2018 (UTC)[reply]
I have said nothing about adding any discussion of markets. O3000 (talk) 17:06, 12 January 2018 (UTC)[reply]
Perhaps I am mistaken, but if it's not relevant why is it brought up? "large moves in both Intel and AMD stock", "...opportunity of the AMD CEO to take advantage...", "Numerous articles appeared on how datacenters were thinking of switching from Intel and that this was a huge setback for Intel" Dbsseven (talk) 17:12, 12 January 2018 (UTC)[reply]
And if you'd like to discuss Intel's performance I am happy to discuss it on the Intel talk page as there are a number of new and reputable sources discussing performance impacts. Dbsseven (talk) 16:10, 12 January 2018 (UTC)[reply]
As for Intel, just dump the 30% number. The Register should not be used as a source. O3000 (talk) 16:13, 12 January 2018 (UTC)[reply]

Please stop separating one discussion into three. I bring up the market impact NOT because I want this included in the article; but because I do not want Wikipedia to be part of the incorrect messages that have caused market swings. And because it is caused by WP:RECENTISM in the press. I have warned about RECENTISM from the very start of this subject. We avoid recentism to avoid making false statements in WP. So, let's get the articles back on track. The changes you made to this article tend to downplay the effect on AMD chips by removing the useful list of chips and repeating their older claim that effects would be minimal. A claim made when they said there was a near-zero impact. Changing the microcode on all of their chips and requiring OS changes to numerous versions of Windows and Linux is not near-zero. OTOH, the Intel article still contains the ludicrous 30% claim that should never have been allowed. So, we state Intel is up to 30% impact and AMD near-zero. This is false and makes us look biased. What I added to this article is what happened. O3000 (talk) 17:32, 12 January 2018 (UTC)[reply]

Thank you for explaining why you keep bringing up the market. However, I believe the current version over-emphasizes a contradiction in what was said about AMD, rather than what AMD stated. (The "near zero" was only in reference to Spectre 2. Form all sources I have found AMD's Spectre 1 solution was always OS patches.) Therefore I propose the following text: "While AMD initially stated architectural differences would result in less risk to its chips, they later announced that software and microcode updates would be made available to further mitigate against these flaws." As for an impact of the AMD fixes, that will require a cite. (As for my writing style on talk pages, I believe it was in keeping with WP:INDENT) Dbsseven (talk) 18:25, 12 January 2018 (UTC)[reply]
We need to talk more to what RS say, not what AMD says. I gave six sources that say AMD changed its position. Of course, they said this in respect to AMD stock as they are financial sources. We shouldn’t talk about the stock prices. But, we should say they changed position. I don’t see any evidence that the microcode and software changes “further mitigate” the problem. They solve a problem that had to be solved. There is no reason to believe the problem is any more or less a problem than Intel’s. AMD is still mum about effects. Intel has discussed, to some degree, effects on various OS and chip families. Intel needs to do more along these lines as it would be difficult for a third party. AMD has said nothing about performance effects beyond its original, incorrect statement. O3000 (talk) 18:42, 12 January 2018 (UTC)[reply]
"There is no reason to believe the problem is any more or less a problem than Intel’s." this is original research without a cite and contradicts cited/citable sources. For example: If Intel had 100% risk and AMD had 1% risk, AMD and Intel releasing fixes to produce 0% risk does not make AMD's original position incorrect. (Even if AMD realized it originally had 5% risk.) We cannot assume equivalent risk without a cite, and technology publications/cites do not support this. [12][13][14] Dbsseven (talk) 18:57, 12 January 2018 (UTC)[reply]
And we cannot discuss impact without a cite (this screams our previous Intel discussion of WP:RECENTISM) Dbsseven (talk) 18:59, 12 January 2018 (UTC)[reply]
Proposed consensus language: While AMD initially stated architectural differences would result in less risk to its chips, they later announced that software and microcode updates would be necessary to further mitigate against these flaws. Dbsseven (talk) 19:01, 12 January 2018 (UTC)[reply]
Sorry, how can saying we have no reason to believe the problem is more or less different OR? It’s the opposite of OR. I’m saying we can’t make this claim. And of course we shouldn’t talk about impact quantitatively with AMD since they haven’t said anything. And there is no way we should assume that it is less than the effects on Intel chips. This is my problem with the Intel article. It contains a made up number that The Register attributed to an anonymous source. I am opposed to your changes. We have NO reason to believe there was any initial mitigation. That was a prior claim by AMD. AMD reversed its position and there exists no third party evidence that the effect on AMD and Intel chips is any different. Perhaps there is a different impact. But, we have zero AMD benchmarks from any source. We don’t even know if they know as they only say that they plan to make changes. It doesn’t appear that they have solved the problem as of yesterday’s announcement. O3000 (talk) 19:31, 12 January 2018 (UTC)[reply]
AMD is currently making the claim that it's architecture is low risk.[15][16][17] AMD is a citable source, even if one would prefer an external source to confirm. (And editor saying they have significant or equivalent risk is definitely NOT a citable source.) And a reversal would require them to have stated they have no risk, something they didn't do and reiterated in the latest sources.[18] Dbsseven (talk) 19:37, 12 January 2018 (UTC)[reply]
I did not say they are equivalent. All of the affected architectures are low risk as Spectre II is extremely difficult to hack.(No reports of hacks in 15 years and a billion chips.) I am not the one making any claims about effects on AMD chips. I don’t know the performance effect and am opposed to any such mention or suggestion that the effect on AMD chips is any more or less than Intel chips as no one is making this claim at this point. We must not make the mistake made in the Intel article of publishing assumptions that were bound to change. AMD’s initial statement suggested that no vulnerability had been demonstrated on AMD processors and they would probably not need to take any action. And, they promised no action. And that’s why RS focused on Intel, which had admitted the problem. Only eight days later did AMD admit vulnerability and say they were taking actions -- and RS reacted. I provided a half-dozen sources stating that AMD changed their position yesterday. O3000 (talk) 20:03, 12 January 2018 (UTC)[reply]
"I did not say they are equivalent." -> "There is no reason to believe the problem is any more or less a problem than Intel’s." ?! Dbsseven (talk) 20:19, 12 January 2018 (UTC)[reply]
Please stop bringing up performance. We agree, it should not be discussed yet. Dbsseven (talk) 20:24, 12 January 2018 (UTC)[reply]
So don't say "further mitigate". That indicates there is already mitigation. It suggests a performance impact difference betwixt Intel and AMD chips. The words have no other purpose. O3000 (talk) 20:27, 12 January 2018 (UTC)[reply]
It says nothing about performance, relative or otherwise. It addresses risk (And the already present mitigation is AMD's microarchitecture, according to them.) From AMD's site: "While we believe that AMD’s processor architectures make it difficult to exploit... further mitigate the threat" Dbsseven (talk) 20:32, 12 January 2018 (UTC)[reply]

try again consensus

Okay Objective3000 I'd like to try again. Starting from the current text I have three larger and one minor thought. Larger:

  • I strongly disagree with "reversed" as it not AMD that was reversed, but what was said about AMD's products. Either we should replace the word or, state "AMD said... while commentors stated/believed/assumed...".
  • I think the “near zero” quote should be removed. This only ever applied to Spectre 2, and is an incomplete view of AMD's position.
  • I also believe it worth noting that AMD stated their products' microarchitecture makes this attack difficult.

Minor:

  • Listing the affected products is un-necessary and extraneous. A better descriptor, such as is used in the Intel article (by date range) would be better.


I will let you propose consensus language, as I have tried previously. Dbsseven (talk) 20:47, 12 January 2018 (UTC)[reply]
  • AMD changed their position. This is important as some readers may have only seen the original position
  • Spectre 2 is the difficult one to fix. Near-zero was what AMD said and RS repeated.
  • Intel’s microarchitecture is also extremely difficult to attack in this manner. 15 years and a billion chips without an attack. Also, AMD made the statement before they changed their position.
  • We don’t have a source for a date range, and might be dates of manufacture, not sale. We have a source for affected chips – which may be of interest to AMD users. O3000 (talk) 21:02, 12 January 2018 (UTC)[reply]
Please provide evidence AMD changed their position. I have searched all the cites provided, none say "reversed". One clearly states "AMD clarified that it never said its chips were not susceptible to variant 2"[19] Journalists/commentators/investors may have assumed that this but there is no cite for AMD taking this position. (A number of the cites note this fact.) AMD only said "near zero", not "zero".
As for the "near zero" quote, then it needs to be clearly stated that this was only for Spectre 2. Otherwise this is unclear and misleading.
And why bring up Intel? That is irrelevant here. ("Pepsi is a popular cola" cannot be included excluded simply because Coke is popular too.) (talk) 21:11, 12 January 2018 (UTC)[reply]
AGAIN, I provided six sources that say AMD changed its position. And you were the one that brought up Intel. This is boring. The absurd, incorrect, 30%, scary as Hell number was forced into the Intel article and you look like you're trying to whitewash AMD's problems and responses in this article. Intel and AMD BOTH have major flaws with the same dangerous effect. Let's not dump on one and act as apologists for the other. O3000 (talk) 21:23, 12 January 2018 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:AMD&oldid=820060190"