Talk:X86-64: Difference between revisions
Line 237: | Line 237: | ||
: See [[x86 virtualization]]. If it's not mention there, perhaps it should be, but not here (IMHO).[[User:War|War]] ([[User talk:War|talk]]) 20:25, 10 June 2017 (UTC) |
: See [[x86 virtualization]]. If it's not mention there, perhaps it should be, but not here (IMHO).[[User:War|War]] ([[User talk:War|talk]]) 20:25, 10 June 2017 (UTC) |
||
:Oh, great, but before you made any comment, please understand what you read! [[Special:Contributions/175.19.66.47|175.19.66.47]] ([[User talk:175.19.66.47|talk]]) 01:31, 18 June 2017 (UTC) |
::Oh, great, but before you made any comment, please understand what you read! [[Special:Contributions/175.19.66.47|175.19.66.47]] ([[User talk:175.19.66.47|talk]]) 01:31, 18 June 2017 (UTC) |
Revision as of 01:31, 18 June 2017
This is the talk page for discussing improvements to the X86-64 article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL |
Computing C‑class High‑importance | |||||||||||||
|
To-do list for X86-64:
|
This page has archives. Sections older than 30 days may be automatically archived by Lowercase sigmabot III when more than 3 sections are present. |
Protected edit request on 16 May 2017
This edit request has been answered. Set the |answered= or |ans= parameter to no to reactivate your request. |
In the video game consoles section, the paragraph states:
Both PlayStation 4 and Xbox One and their successors incorporate AMD x86-64 processors, based on Jaguar microarchitecture[78][79]. Firmware and games are written in the x86-64 codes, no legacy x86 codes are involved with them.
I would like it changed to:
Both the PlayStation 4 and Xbox One incorporate x86-64 processors from AMD's Jaguar microarchitecture.[78][79] Firmware and games are written specifically for the architecture, without the need for code written for the legacy x86 component.
Because it is not very well written and I would like it corrected and clarified. Here's the markup.
Both the [[PlayStation 4]] and [[Xbox One]] incorporate x86-64 processors from AMD's [[Jaguar (microarchitecture)|Jaguar]] [[microarchitecture]].<ref name=XboxOneMay2013Anandtechcomparison>{{cite news |title= The Xbox One: Hardware Analysis & Comparison to PlayStation 4 |author= Anand Lal Shimpi |publisher= Anandtech |url= http://www.anandtech.com/show/6972/xbox-one-hardware-compared-to-playstation-4/2 |date= 2013-05-21 |accessdate= 2013-05-22}}</ref><ref name=XboxOneMay2013SpecGameinformer>{{cite web |url= http://www.gameinformer.com/b/news/archive/2013/05/21/the-tech-spec-test-xbox-one-vs-playstation-4.aspx |title= The Tech Spec Test: Xbox One Vs. PlayStation 4 |publisher= Game Informer |date= 2013-05-21 |accessdate= 2013-05-22}}</ref> Firmware and games are written specifically for the architecture, without the need for code written for the legacy x86 component.
Thank you. FosterHaven (talk) 16:08, 16 May 2017 (UTC)
- @FosterHaven: Not done: The page's protection level has changed since this request was placed. You should now be able to edit the page yourself. If you still seem to be unable to, please reopen the request with further details. Murph9000 (talk) 08:43, 20 May 2017 (UTC)
- @Murph9000: Thank you! FosterHaven (talk) 15:21, 20 May 2017 (UTC)
"without the need for code written for the legacy x86 component"
- Your request to edit it is appreciated, but you change the meaning of that sentence. Because there arise another question, are the codes written in x86 able to work on those game consoles? Many years passed after they both introduced, Microsoft provides an virtual zone for Xbox One to run their customised Windows 10 on it, but there is no voice about the backward capabilities with x86 software. As to the PS4, there is nobody hacking it and enable it to work as a traditional PC. If you read my words on the talk page of x86, you would find the differences between x64 and x86-64 processor. One could make sure that both game consoles use x86-64 processors, but unsure that they are x64 processors or not. If they are not x64 processors, in other words, they may be incapable to run x86 application if without support from software. So this necessary would turn to possibility.119.53.119.155 (talk) 22:31, 16 May 2017 (UTC)
Both the PlayStation 4 and Xbox One incorporate x86-64 processors from AMD's Jaguar microarchitecture.[78][79] Firmware and games are written specifically for the architecture, without the need for code written for the legacy x86 component.
- There are some problems after you changed,
- 1. You censored the "successors", you might possibly treat those successors are the following versions of PS4 or Xbox One, or treat them as the same things.
- 2. Processors from microarchitecture, yeah, processors are the physical realisation of specific microarchitecture, but such expression is quite rare to find.
- 3. What is the legacy x86 component? Your expression might embrace the content of your own opinion or guess, such as "the need for". Why the designers and developers of game consoles put too much effort to the things they do not really care about?
- There are some problems after you changed,
- In my own opinion, your version is a little bit worse than the former. But I do appreciate your request. 119.53.110.4 (talk) 07:30, 17 May 2017 (UTC)
- I'm very willing to believe you don't actually appreciate my contribution. I just read up the sections in the x86 talk page you mentioned that actually wasted Guy Harris' time. Let me counter-argue in full support of his word salad claim:
are the codes written in x86 able to work on those game consoles?
- If by "codes" you mean the raw instructions, then I would like to point out that if it doesn't, then it's not an x86 processor. x86-64 is an extension of x86; it extends the capabilities of the x86 architecture to support 64-bit instructions. And if you couldn't tell what that means, x86-64 supports x86 instructions. At least, that's what the article should explain. If it didn't do that, then by all means, create an account and edit it so that it does.
- And I can tell you're inexperienced because absolutely nobody refers to programs as "codes".
As to the PS4, there is nobody hacking it and enable it to work as a traditional PC.
- I raise this blog post by fail0verflow detailing how they ported Linux to the PS4. And don't bring up "but it's Linux, not Windows," that just means you can do way more with the system than Microsoft allows with Windows. It's always been like that for Linux.
- I'm very willing to believe you don't actually appreciate my contribution. I just read up the sections in the x86 talk page you mentioned that actually wasted Guy Harris' time. Let me counter-argue in full support of his word salad claim:
There are some problems after you changed, [...] [and] your version is a little bit worse than the former. But I do appreciate your request.
- Subtle. But here's why those problems don't actually apply:
You censored the "successors", you might possibly treat those successors are the following versions of PS4 or Xbox One
- You might. But that's not what censorship is. I've removed it because we don't know if the successor consoles will have the same processors, and it's certainly not Wikipedia's job to speculate on that. A reliable source would have to confirm leaked or revealed information, and that obviously hasn't happened in the years that the consoles have been out. (As per the fact that "a schedule of future events may be appropriate if it can be verified.")
yeah, processors are the physical realisation of specific microarchitecture, but such expression is quite rare to find.
- It's actually not rare, so what's the "problem" you speak of? Is my expression not good enough for Wikipedia?
What is the legacy x86 component? Your expression might embrace the content of your own opinion or guess, such as "the need for".
- It's a copyedit. My expression embraces Wikipedia's perspective, which I didn't know was actually yours before I arrived. If a copyedit contains bias somehow, then the editor is not doing a very good job.
Why the designers and developers of game consoles put too much effort to the things they do not really care about?
- That's not for us to explain, certainly not on this article. And that's probably why that second sentence isn't even needed.
- So with your points countered, I would like explain what it is you're trying to do, because I know it very well; you are setting a trap, and so far it's worked. You show persistence, and a pure lack of involvement in the project's efforts. The first thing you tell us is that you try to find the claims we make, and apparently you can't? And so we're wrong as a result? And after failing to take in the opposing side, you insist that your conclusion is the correct one, even when Guy Harris stated the proof you brought up had went against your claims. So when he's proved all he's needed to, you go out of your way to get the last word as if it actually matters. And then, while I wander onto the page and see what I believe is an error, and suggest it accordingly because it's restricted, you dissect my copyedit for "changing the meaning of the sentence" which is basically that it goes against your perspective.
- So let me clarify. x86 is a set of architectures, x86-64 is the 64-bit architecture that adds more capabilities and is backwards compatible with x86. You cannot argue otherwise when scientific journals are using the jargon without your involvement. I sincerely apologize that you're not important to this article, but we rely on them for our facts, not you. Don't go around this place acting like you own it, because you don't even have an account, and I assume if you made one, you would get your editing privileges revoked immediately.
- I had been off to do other things for two days whilst not even being aware of this whole discussion taking place. I did not notice what happened at all but now I'm very, very glad I did, thanks to Murph notifying me above. FosterHaven (talk) 15:21, 20 May 2017 (UTC)
- Ignore the IP. The IP's edit warring and attempts to edit against consensus is the reason the page has been protected. If you read the IP's words on talk page of x86, and responses thereto, you will find that her/his assumption that x86-64 and x64 are, or could in any way be considered, two different things is completely unsupported. Jeh (talk) 02:37, 17 May 2017 (UTC)
- @Jeh: I figured that was the case after I skimmed it. I'll be removing the second sentence in the section, since I feel like it's a fact that shouldn't need to be stated already. FosterHaven (talk) 15:21, 20 May 2017 (UTC)
- I've you read your words, and that is enough! 221.9.20.168 (talk) 05:09, 21 May 2017 (UTC)
- I'm just being honest, even if a bit pushy. I know you want to help this article (and you should know that everyone here wants to, otherwise what would be the point), but not everyone is going to be on your side if you argue against the editors about how the article and the writers are wrong. Ultimately, Wikipedia being a collaborative effort means that even though we do value input, it's not a necessity to insert your perspective. I'm not saying we don't need your help at all, but you're positioning yourself in a bad spot so if you want to help, you'd have to follow the consensus. There's really not much more for me to say. FosterHaven (talk) 13:08, 21 May 2017 (UTC)
- Thank you for your reply. In 2008, you were 9, but I was 24. I read articles on Wikipedia.org since late 2004, I got in touch with x86-64 processor in that same year too. I experienced Windows XP on IA-32, Itanium and x64 platforms, I experienced the Xbox in your words too when in high school. So ..., ok, you are welcome. — Preceding unsigned comment added by 221.9.19.34 (talk) 14:53, 21 May 2017 (UTC)
- In 2008, I was 44, but you were only 24. It got "in touch" with x86 in 1986. You were 2. Looks like you are using a mix of the Appeal to accomplishment and Appeal to tradition. Fallacies aside, I've been reading this discussion with some interest. It appears to me to be a mix of trying to find and communicate the history of events correctly mixed with a misunderstanding that phrases like "x86-64" and "architecture" don't always maintain an exact meaning over time.War (talk) 18:59, 22 May 2017 (UTC)
x86 is a bunch of architectures, and x86-64 is only the 64-bit version instruction set?
From wiki article x86
"x86 is a family of backward-compatible instruction set architectures based on the Intel 8086 CPU and its Intel 8088 variant."
From wiki article x86-64
"x86-64 (also known as x64, x86_64, AMD64 and Intel 64 is the 64-bit version of the x86 instruction set."
Those two wiki articles say that the x86 architecture includes all the architecture(s) since Intel 8086 through Core i7 and AMD Ryzen; and x86-64 is only an instruction set of the former. I do not want to be accused as the human embodiment of the Dunning-Kruger effect by some a computer genius from Wikipedia.org, who experienced programming more than 40 years! Maybe writers and editors from AMD and Intel are fools because they deny them, by differentiating x86 and AMD64 or IA-32 and Intel 64. Most Linux distributions differentiate them both as x86 and x86-64 too, the very famous free productivity software, Libre Office differentiate them both in that way too. Maybe all of them are completely wrong, because they are not computer genius. Maybe Bill Gates is neither a computer genius, because his programming experiences began also more than 40 years ago, but today Microsoft also differentiate them both as x86 and x64.
A programmer with more than 40 years programming experiences would give the definitions like above about x86 and x86-64, I just realise the definition of computer genius maybe also need to be redefined! A person programmed for more than 40 years something like below,
"Repeating your assertions over and over again is not going to convince me. If you're also a software developer who's worked at that level, you can probably come up with arguments based on more than just arguments about the meaning of words; if you're just a fanboy who's read a bunch of articles and thinks that makes him an expert who can lecture people with more experience in the field, you're simply wasting your time and convincing us that you're a fool."
I understand I have to accept this person's words, "you're simply wasting your time"! For any penny, I have to say this guy is a computer genius! — Preceding unsigned comment added by 119.53.113.164 (talk • contribs) 08:00, 18 May 2017 UTC (UTC)
- That is WAY TOO MUCH editorializing. If you have a specific section or sentence you think should change. Quote it and then propose what it can be changed to. We can discuss it. I won't discuss who is or isn't a genius or how many years experience someone has...neither is relevant.War (talk) 15:24, 25 May 2017 (UTC)
Take an extent look of x86 or IA-32,
In the history of IBM PC and its compatible machines, the dominated Operating Systems are from Microsoft, such as MS-DOS, IBM PCDOS, Microsoft OS/2, IBM OS/2 Wrap, Windows 3.x, Windows 9x, Windows NT and Windows 2000/XP. Later IBM and Legend merge into Lenovo, the new IBM rather than the new Legend. IBM dropped their market on personal computing, and the new IBM, Lenovo lose all its dominated position in personal computing market. And that period was gone! IBM PC Compatible is gone too! But its platform rooted into today's personal computers.
What is x86 architecture? From another aspect of it, not constrained by the aspect from its inner world, another definition could be drawn,
The architecture (ISA) of IBM PC and all its compatibles is x86.
or
The x86 is the architecture of IBM PC and all its compatibles.
The dominated or major operating system for IBM PC are DOS and Windows! Such as 16-bit Windows, Windows 1.x/2.x/3.x; 16-bit/32-bit mixed Windows, such as Windows 3.1 Workgroup, Windows 95/95 OSR2/98/98SE/ME; 32-bit Windows, Windows NT/2000/XP. The architecture those operating system are build on is just the x86 architecture. And those Windows 9x operating systems are the good examples to say that this x86 architecture is an architecture rather than many. Because Windows 9x span over almost the entire modes of IA-32 processors in their running time. Omitting any one, these OSes would fail to work. From the aspect of a computing system rather than the level on the architecture of processor itself, the x86 is an architecture, rather than many. Of course, system based on 16-bit processors have no chances to work on Windows 9x, that is another reason to say the 16-bit architecture introduced by those 16-bit processors has been eventually evolved the IA-32 or x86 architecture. In other words, the x86 architecture is not separable. Another important role of Windows 9x is that they are the native OS for those IBM PC Compatibles and x86 processors. But Windows NT is not, they are not designed exclusively for x86 processors, in fact they are developed on an Intel RISC processors. The support for then mainstream RISC processors was their initial design purpose. On the port of x86, this OS just utilise the pure 32-bit mode, rather than saying that that 32-bit environment is the IA-32, or 32-bit version of x86. And what IA-32 refers to, in Intel official documents, has already stats the entire architecture of 32-bit processors, or x86, exactly.
Different from that Windows 95 is the native operating system for x86 architecture, Windows NT is a portable operating system. In order to make an OS to be portable, the designers must be in consideration of the differences among different architectures. They would design it carefully and avoid to get in touch with the unique features of each supported architecture, obviously, codes on those things are hard to be portable. In order to achieve this goal, the OS itself must be stand above a little bit higher than the native one. Like moving a building from a place to another, the most easiest way is put the live onto the bottom and contact part. Because moving that part, the remains above on it would follow its movement, without worrying it to be crushed. Windows NT, like so many other portable OS, such as Linux, BSD and VMS, it is built above on one portable abstract layer rather than get in touch with the hardware directly. For the port of x86, they just utilise the partial features found on the 32-bit architecture introduced by 80386, and denote such architecture as "i386" on its installation media. Of course, Windows NT is not designed exclusively and natively for x86 architecture. Even though the needs for backward support for the legacy 16-bit applications were important, they realise such features like in other ports, in the form of software emulation, rather than trap processors back to 16-bit real mode (like Windows 9x), or utilise the Virtual x86 mode. Like Windows NT, Linux is also a product of 80386 age. Even though Linux is not UNIX, but its ideas are brought from UNIX, so the process to make it realise on the 80386 processor is also a process of porting rather than design for x86 processor natively.
Evaluating the modern operating system and applications, AMD designed a 64-bit architecture, which is extended from x86 architecture, catering for those who are eager to port their current OS, rather than re-design from the ground, with enough support for the backward supports, and this 64-bit architecture is named as AMD64 or x86-64. Comparing with that Windows 9x is designed natively for x86 architecture, and Windows NT is actively ported to x86 architecture; this x86-64 architecture is designed intentionally for those modern OSes. Comparing to saying that Windows 9x is the native OS of x86; x86-64 is the native architecture for Windows, Linux, FreeBSD and so many other modern OSes. From the viewpoint of this, x86-64 is an architecture, rather than merely an instruction set. For its independence from legacy x86 architecture, it is an standalone architecture, rather than an 64-bit extension to the IA-32, like SSE, MMX and so forth.
Apple once gave an illusion that it were an 64-bit extension to the IA-32, like SSE, MMX and so forth since Mac OS X 10.4.6 though Mac OS X 10.7. But that is only an illusion, this special 32-bit kernel completely run within IA-32e mode rather than span the entire modes of Intel 64 processors. This link between IA-32 and Intel 64 is a fake. This once again says that the relationship between x86 and x86-64 is not the latter is part of the former, but just in the opposite, x8-64 succeeds the history of x86. Legacy in software development implies that those software would not be modified or updated. On x86-64 architecture, the shadows of x86 present itself in the legacy form no matter in Legacy mode or Compatibility Mode. Or in other words, the fundamental architecture of x86 has not been changed. In the words of AMD, it is called "backward compatible". This also stats that x86-64 is not the 64-bit version of x86, but for their similar natures, it is the 64-bit extension of x86 architecture. Or in other words, the x86-64 is designed mostly based on extending the x86 architecture, rather than some a version of it.
I do appreciate differentiate the architecturer(s) among every generation of x86 processor and x86-64 processor. But, that is not only my imagination, it is widely accepted that x86 is a 32-bit architecture introduced with 80386 processor, and also denoted as i386, Intel refer it as IA-32. This 32-bit architecture does not only point to the architecture within the 32-bit protected mode, but the entire architectures of Intel 80386 processor. Because no matter in protected mode or real mode, 32-bit registers are visible and could be utilised, and the values keep consistent across operating modes. This is not my interpretation of IA-32, such similar contents could be found on both Intel and AMD official documents.
People often use the term Chinese to refer to the entire Chinese ethnic from all over the world (they might not even know how to speak Chinese), but strictly, the term Chinese is only used to referred to the people native in China. In the similar way, people often use the term x86 to refer to all the processors whose architectures are similar or same as x86, but strictly x86 is only used to refer the architecture introduced by Intel 80386 processor, or IA-32. The fact that both x86 and x86-64 architectures co-exist on x86-64 processors, so the x86-64 is also an x86 processor. But an x86 processor might not be an x86-64 processor. Today there is no legacy x86 processor being manufactured anymore, all current x86 processor are actually the x86-64 processor. And it is the age of x86-64! — Preceding unsigned comment added by 221.9.20.9 (talk • contribs) 15:48, 18 May 2017 (UTC)
- Looks to me like you should have a blog or write a book. A talk page is not the place for essays. Do you have something specific you think should change? War (talk) 15:29, 25 May 2017 (UTC)
- Wikipedia.org should stand on a neutral position, but 64-bit version of x86 is a fake! It misleads readers, innocent readers from all over the world, believing that x86-64 is just x86, but it is not! Who are you? I am no interested in your age, your professional too, because you mislead them too! — Preceding unsigned comment added by 119.53.119.96 (talk) 22:43, 25 May 2017 (UTC)
- Nothing here says or even slightly implies that x86-64 is "just x86". What is "non-neutral" about "x86-64 (also known as x64, x86_64, AMD64 and Intel 64) is the 64-bit version of the x86 instruction set"? What "side" is not being fairly represented thereby? Jeh (talk) 07:30, 26 May 2017 (UTC)
"AMD64 might have been AMD's response to IA-64", is it real?
We need sources! — Preceding unsigned comment added by 221.9.19.34 (talk) 10:23, 21 May 2017 (UTC)
- 1) "might have been" and 2) it's not a claim in an article, it's part of an edit comment. It's a possibility - I might have seen it suggested in an issue of Microprocessor Report, given that Intel and HP had a lot of patents on IA-64 and were not offering to license them to anybody, including AMD, AMD would have to do something if 64-bit processors became important. What they did was to come up with a 64-bit version of x86; whether that was the motivation for developing x86-64 is another matter. If I had a source, I'd put it in an article; as I don't, I'll mention it in discussions as a possibility, but not claim it to be a fact. Guy Harris (talk) 18:19, 21 May 2017 (UTC)
- They are the words of many people, but that might not be the real story. Like SPARC or UltraSPARC, IA-64 was designed as an open architecture, in other words, the developers wished it could be implemented by lots of semi companies rather than only Intel. But for the complicated nature of EPIC IA-64 architecture and its unforeseen future, few semi-companies would be content to get that try. HP saw Intel had the largest number of consumers from the all over the world with their x86 based products. With the help of Intel, HP believed that rooting such processors onto the everyone's desktop would accelerate this architecture evolved to be mature and popular. With the efforts of Intel, this VLIW-based EPIC architecture embraced the farther relationship of x86 or IA-32, in those days when RISC processor played as an important role in server and professional market. Windows powered by Itanium would put a heavy beat towards Mac OS X on PowerPC. But, after all, x86 would never evolve into EPIC, and what Intel possessed all the time is the x86 rather than IA-64. But the future of x86 architecture might leave Intel an unclear future, with the disguise of Itanium, they swift from the focus on the architecture(s) onto the implementations, when x86 or IA-32 has already been mature. The earliest information on AMD64 from Microsoft is on the installation information of Windows XP in 2001. Maybe during all those days, no matter Intel or Microsoft consider a common question, why another x86 architecture? But for what reason why let that architecture gone if some a force is still working on? And the most important thing is that there were already so many out-of-box applications and developers who are familiar on such architecture. So they would be pleased with another x86 rather than like a fish flying on the sky, if transiting from x86 to EPIC IA-64.
- Comparing with x86 or IA-32 architecture, when so many others are dead nowadays, but x86 is still strong even though its nature is 32-bit and x86-64 dominated the professional and server market. The reason is that the unique architectural characteristics of x86 make it easily be expanded and/or extended, and the backward supports for codes written on previous generations of processors could be kept retained. AMD model such things onto the x86-64 architecture, without even breaking the x86 architecture at any penny, because the native x86-64 architecture eventually evolved lacks almost everything around that.
- When Intel released Itanium processor, there were no fears coming from AMD. Because that processor is only strong for its presumed architectural advantages but lacks the support base from everything what x86 processors already had, it did not even have a native OS to boot up. Since then to now, yup, there is not an O/S exclusively designed for Itanium, but only ported to it. When Microsoft released Windows XP 64-bit Edition (Version 2002) they did not even put it at the pace of Windows XP Professional for x86, without modern GUI, and without even needing an activation. But like AMD K6 processors, AMD put the things on Windows a second time to their processors, this time not the Windows Logo, but the 9x mirrored "xp" onto the name of their processors, Athlon xp. So AMD64 was not like AMD's response to IA-64, not then, not now!
- "Like SPARC or UltraSPARC, IA-64 was designed as an open architecture, in other words, the developers wished it could be implemented by lots of semi companies rather than only Intel." Prove it.
- "Windows powered by Itanium would put a heavy beat towards Mac OS X on PowerPC." Windows already had 90+% of the personal computer market, and it's not as if you needed more than x86 to beat PowerPC.
- "but x86 is still strong even though its nature is 32-bit and x86-64 dominated the professional and server market." x86 started out as a 16-bit architecture, added 32-bit capabilities with the 80386, and added 64-bit capabilities with Opteron. There's nothing "naturally" 32-bit about it.
- "The reason is that the unique architectural characteristics of x86 make it easily be expanded and/or extended, and the backward supports for codes written on previous generations of processors could be kept retained." The "unique architectural characteristics of x86" that made it dominant are the IBM Personal Computer, MS-DOS, and Windows, so processors implementing it were made in very high volume, and Intel could throw research and transistors at the problem of making it run fast. About the only thing about the x86 instruction set that could be viewed as making it more expandable than other instruction sets is the variable-length instruction encoding, but it's certainly not the only ISA that 1) went from 32 bits to 64 bits or 2) picked up a bunch of SIMD/vector extensions.
- "When Intel released Itanium processor, there were no fears coming from AMD. Because that processor is only strong for its presumed architectural advantages but lacks the support base from everything what x86 processors already had, it did not even have a native OS to boot up" One does not only fear immediate threats - if the future were to go 64-bit, and AMD had no 64-bit processors, they'd be in trouble. They were an x86 vendor, so they'd either have to 1) find some vendor willing to license their 64-bit ISA, 2) invent a brand new ISA, or 3) invent 64-bit x86.
- "Since then to now, yup, there is not an O/S exclusively designed for Itanium, but only ported to it." So what? Most of the OSes for the various RISC processors were ported to, or adapted for, those processors. Having a Shiny New OS for your Shiny New ISA isn't a requirement for its success.
- I've ordered a copy of the Microprocessor Report article on AMD64. We'll see what it says. I may also order articles on IA-64 to see what they say about how "open" it was intended to be. Guy Harris (talk) 02:15, 22 May 2017 (UTC)
- I think it will be hard to find a source that expresses AMD's "intent" vis-à-vis the Itanium. At the time there were many applications with memory issues and the x86 "extended memory" model just wasn't cutting it...especially with servers becomes more prominent. I bet you are more likely to find this as the driving force behind the AMD64 rather than a response to Itanium. War (talk) 19:16, 22 May 2017 (UTC)
- The Microprocessor Report article in question, AMD Drops 64-Bit Hammer on x86 (it's behind a USD 95 paywall), didn't say anything about that (although it sure used "x86" a lot to refer to the instruction set family, from 16 bits to 64 bits, emphasis on "to 64 bits"). The question is whether AMD, at any point, realized that they wouldn't be able to track Intel into the 64-bit space, due to patents etc. (the MPR article does note that neither Intel nor AMD signed on any other processor for their architecture, the claims by Aaron Janagewen that "IA-64 was designed as an open architecture, in other words, the developers wished it could be implemented by lots of semi companies rather than only Intel" notwithstanding), and that this meant that AMD wasn't going to be able to continue forever as a 32-bit x86 vendor, and wouldn't be able to become an IA-64 vendor, if 64-bit computing became significant, and whether that contributed to their decision to go with x86-64. If Intel had gone with 64-bit x86 from the beginning, and if AMD's licensing of Intel patents would have covered that 64-bit x86, they might not have invented their own 64-bit x86.
- However, perhaps they could have become, for example, a PowerPC vendor rather than becoming the first x86-64 vendor, so it's not as if not being able to implement IA-64 would necessarily have closed off all opportunities to go 64-bit other than "roll your own instruction set", although, at the time, IA-64 may have been viewed as a much bigger threat to other 64-bit architectures than it turned out to be, so they might have figured that the future would ultimately belong to IA-64 unless they could figure out a way to have a 64-bit ISA that would have a marketplace advantage over the other 64-bit processors - and "hey, it can run 32-bit x86 code, including OSes for 32-bit x86, at full speed, and it just adds things to x86 that an existing x86 compiler or operating system doesn't need a lot of work to deal with" seems to have been what they needed. Guy Harris (talk) 21:02, 22 May 2017 (UTC)
- Intel announced the processor based on HP IA-64, Itanium, in year 1999; released the very first Intel Itanium processor in 2001, the very first Pentium 4 with EM64T enabled in year 2004, transmit the entire product line to Intel 64 in 2006. 2006 - 2001 = 5 years, or 2006 - 1999 = 7 years. Is it too long for the five years, transitioned from IA-32 to IA-64? Maybe in the former British Hong Kong, if you could stay there for more than 7 years, you would become a British Hong Kong citizen naturally. Intel 80386 was released in 1985, but when in 1994, most applications were most written in 16-bit codes, 1994 - 1985 = 9 years! I believe Intel had very few real expectations transition from IA-32 to IA-64, but only because actual AMD64 processor was not on earth during that time. Only one year later, Intel cannot wait to releasing their implementations of x86-64 architecture. And only two years later, they refreshed their entire product line with Intel 64 processors. Within so few years, Itanium was kept out, and become another solo ISA among others. So AMD64 was only advertised to be the AMD's response to IA-64, but it was not since the very beginning! 119.53.114.43 (talk) 02:02, 26 May 2017 (UTC)
- I don't see how any of this discussion is going to lead to improvements in the article, which is what talk pages are for. We have no way to reliably source AMD's intentions or strategies of the late 90s. Jeh (talk) 07:27, 26 May 2017 (UTC)
"VIA Technologies introduced their version of the x86-64 architecture...", why there are so many versions?
"64-bit version of x86...", "enhanced version of PAE...", this title referenced and many more, is that the versionstar invented by version genius? No, the x86-64 architecture in VIA Nano processors are the strict clone of AMD64, only with additional unrelated additional instruction sets from Centaur Technology. — Preceding unsigned comment added by 221.9.22.167 (talk) 10:14, 25 May 2017 (UTC)
- If you have a specific section or sentence you think should change. Quote it and then propose what it can be changed to. We can discuss it. I won't discuss who is or isn't a genius.War (talk) 15:24, 25 May 2017 (UTC)
- Your people discuss what? Only that Guy Harris stood on his/her position without being shocked, sticking to those falsies all the time, this guy and his crowds were not trying to find proofs to improve the wiki article, but find all the things to kick me out! If Wikipedia.org needs only minor editors, and prevents others from contributing or correcting, I suggest Wikipedia.org just close the door! I also found a very strange wikipedian who use Wikipedia.org as the thing to advertise this person itself and his/her training company. I found so shame about that! You know who I am really talking about! Maybe they trained too many students with such things incorrect, so they refused to change or correct them, let those lies look like true for ever! — Preceding unsigned comment added by 119.53.119.96 (talk) 22:50, 25 May 2017 (UTC)
- "Only that Guy Harris stood on his/her position without being shocked, sticking to those falsies all the time" I don't wear a bra.
- "I also found a very strange wikipedian who use Wikipedia.org as the thing to advertise this person itself and his/her training company." They also write drivers, as anybody who actually read their web site and understood it would know (yes, this means you either didn't read it or didn't understand it). Guy Harris (talk) 02:58, 26 May 2017 (UTC)
- I don't see why the word "version" there couldn't be changed to "implementation"; as far as the x86-64 ISA is concerned, that's what Nano is. Jeh (talk) 04:25, 26 May 2017 (UTC)
- Sounds good. I've done that. (Maybe they've done things differently from both Intel and AMD, in the ways AMD64 and Intel 64 differ, in which case you could perhaps call it "VIA 64" or something such as that, but, without a source for that, let's just call it an implementation. Note, BTW, that VIA repeatedly speak of it as an x86 chip....) Guy Harris (talk) 05:43, 26 May 2017 (UTC)
"Note, BTW, that VIA repeatedly speak of it as an x86 chip...."
- I thank you make that correction, and I comprehend why you emphasise the "x86 chip" in the above sentence. Remember that VIA Technologies is also a vendor of ARM processors and devices based on them. They use x86 to make differentiation. Like what I explained an x86-64 processor is also an x86 processor, but an x86 processor might not always be an x86-64 processor, ..., anyway, well done, thank you! 221.9.12.193 (talk) 09:13, 26 May 2017 (UTC)
Under the official openSUSE wiki we are said that AMD is also a donor to openSUSE, however why openSUSE refers 64-bit into x86_64 (like Fedora, macOS, Arch Linux, Slackware), instead of amd64 (like Debian and his derivatives and Gentoo)?
202.40.137.199 (talk) 01:49, 4 June 2017 (UTC)
- I don't understand the question. In particular, I can't tell if you are referring to something in the article you think should be changed, added, removed or what. War (talk) 03:19, 4 June 2017 (UTC)
- Agree with War (talk · contribs · deleted contribs · logs · filter log · block user · block log). This appears not to be a question about the subject of this article but rather about the history of openSUSE's 64-bit support. And the claim that AMD is a donor to openSUSE appears to be irrelevant to the issue - they also donate to other projects. Ask over at talk:openSUSE. Jeh (talk) 04:00, 4 June 2017 (UTC)
- The more interesting question might be why some Linux distributions don't call it x86_64, given that the Linux kernel mainly speaks of it as x86_64. openSuSE might just be saying "hey, that's what Linux calls it".... But that is, as Jeh notes, not particularly relevant to the article; it's not disputed that openSuSE calls it x86_64/x86-64 (a quick look at their Web site shows that). Guy Harris (talk) 04:45, 4 June 2017 (UTC)
- Debian and his a part of derivatives are obviously the barely only distro calling amd64 instead of x64 by major, here I have been said the AMD processors are FATALLY power biting making the LITTLE TO MINOR market shares for the desktops and laptops, hence just also making the Linux kernel and the majorities of Linux distro calling x64 (or x86_64 in long term) from Intel, which I need to wonder why barely only a part of Debian derivatives names it to amd64 (instead of x64).124.217.188.71 (talk) 05:33, 4 June 2017 (UTC)
- Again - not a concern of this article. And btw, one might also ask why Microsoft (and much of the press) picked x64. We don't know, and as far as verifiable, reliable sources go, we probably can't find out. So further discussion of this question here seems pointless. Again: maybe you should ask this over at talk:openSUSE? Or maybe at SuperUser.com ? Jeh (talk) 05:59, 4 June 2017 (UTC)
- Hey pal, most Linux distros refer the x86-64 architecture as AMD64 (x86_64). Similar as IA-64 architecture, before physical AMD64 chips were realised onto the market, Linux kernel developers had already been busy with porting Linux onto this then-future 64-bit architecture, AMD64, and Debian is also one of the early distros among them. At that moment they had few knowledge that whether it would be merged into the x86 family or just another architecture alongside with it. So they just referred it as to AMD64. x64 is a term mostly used in Microsoft products, it has similar not completely same meaning as x86-64, with emphasising the 64-bit extended rather than that architecture. Because AMD64 processor is rooted onto the traditional x86 platform, during initialisation and backward support, IA-32 codes are necessary involved into the kernel. So later the developers decided the merge both x86 and amd64 architectures onto the same kernel source tree for convenience. But the Linux distro still use term x86_64 to refer the product designed for 64-bit system based on AMD64 and Intel 64 processors. Term x86-64 is used to refer to such thing, the architecture and the instruction set belonging to it. But there still exist two viewpoints
- 1. Standing on side of the x86 processor, the x86-64 is treated as a 64-bit instruction set extension, emphasising this 64-bit instruction set, when IA-32 software still dominated the software market.
- 2. Standing on the side of x86-64 processors, the x86-64 is treated as a 64-bit architecture, and the support for the x86 software is just to retain the legacy. And this viewpoint satisfy today's real situation.
- So x86-64 is a 64-bit architecture alongside with 32-bit x86 architecture. 175.19.66.153 (talk) 08:57, 10 June 2017 (UTC)
- Again - not a concern of this article. And btw, one might also ask why Microsoft (and much of the press) picked x64. We don't know, and as far as verifiable, reliable sources go, we probably can't find out. So further discussion of this question here seems pointless. Again: maybe you should ask this over at talk:openSUSE? Or maybe at SuperUser.com ? Jeh (talk) 05:59, 4 June 2017 (UTC)
- Debian and his a part of derivatives are obviously the barely only distro calling amd64 instead of x64 by major, here I have been said the AMD processors are FATALLY power biting making the LITTLE TO MINOR market shares for the desktops and laptops, hence just also making the Linux kernel and the majorities of Linux distro calling x64 (or x86_64 in long term) from Intel, which I need to wonder why barely only a part of Debian derivatives names it to amd64 (instead of x64).124.217.188.71 (talk) 05:33, 4 June 2017 (UTC)
- First, I'd like to invite you to sign your posts. Second, please read WP:NOR. When I first started editing I made the same mistake you are making. I thought myself an expert on certain topics and edited content with the knowledge that I had. That's not good enough. You must provide references and reliable sources, regardless of your personal knowledge and experience. Wikipedia is not a blog. War (talk) 20:41, 10 June 2017 (UTC)
War, the IP to whom you responded (175.19.66.153 (talk · contribs · deleted contribs · logs · filter log · block user · block log)) is a confirmed sockpuppet of LTA Janagewen. I recommend WP:RBI, consistent with WP:DFTT. Jeh (talk) 01:14, 11 June 2017 (UTC)
- Thank you for the information. I will take your recommendation.War (talk) 04:10, 12 June 2017 (UTC)
Virtualisation should also be mentioned in the section Differences between AMD64 and Intel 64
Both AMD64 processor and Intel 64 processor have distinct differences when virtualising x86-64 guest OS under VMWARE.
1.Intel 64 processors needs VT-x enabled in both Legacy and IA-32e mode to virtualise x86-64 guest OS.
2.AMD64 processors does not need AMD-V enabled in either Legacy or Long Mode to virtualise x86-64 guest OS.
This distinct difference could be seen as the essential differences between both x64 processors. Today more and more applications needing 64-bit OSs to be virtualised, AMD64 processors could support such virtualisation under Legacy Mode, or in other words, there exist some differences not only one the physical processors, but also the AMD64 architecture. So I think it should be mentioned! 175.19.66.153 (talk) 10:19, 10 June 2017 (UTC)
- See x86 virtualization. If it's not mention there, perhaps it should be, but not here (IMHO).War (talk) 20:25, 10 June 2017 (UTC)
- Oh, great, but before you made any comment, please understand what you read! 175.19.66.47 (talk) 01:31, 18 June 2017 (UTC)