Talk:I486SX

Free Online Dictionary of Computing article cited as source for this article makes numerous verifiably false statements

This article cites the free online dictionary of computing as the original source. Take a look at that article, dated 1997: http://foldoc.org/Intel+486SX. It makes numerous verifiably incorrect statements that demonstrates the author(s) of that article are hardly experts on microprocessors. Only a fool would rely on any material from that article. Let me point them out:

It states ALL 486SX chips were fabricated with FPUs. Take a look at Intel reference - http://www.intel.com/pressroom/kits/quickreffam.htm#i486 - it shows there were versions of the 486SX with 1.2 million transistors (with FPU) and 0.9 million transistors (without FPU).

It claims that 486SX systems would allow a "DX" to be plugged into an expansion socket. In fact it was a 80487SX not DX that was used to disable the 486SX. (http://upload.wikimedia.org/wikipedia/commons/2/22/KL_Intel_i487SX.jpg)

It states some SX chips only had a 16 bit wide external bus. All 486SX and 487SX chips had a 32 bit wide bus. The author is confusing the 486SX with the 386SX. Both the 386SX and the 486SX were meant to be lower performance as compared to their full functional counterparts, the 386DX and 486DX respectively. The 386SX was lower performance by using a 16 bit databus, but the 486SX was lower performance by disabling or removing the FPU.

It states the 80486 DX has a pin to select 16 or 32 bit bus, but the smaller sx had this hardwired to 16 inside the package. In fact, all 80486SX chips used a 32 bit wide databus had the same signal (BS16#) to control whether the 32 bit databus would perform a 16 or 32 bit transfer. See http://www.intel.com/design/intarch/datashts/27276904.pdf, pages 16 (databus d0-32) and page 20 (BS8#, BS16#). Again, the author is confusing the 486SX with the 386SX.

The author claims there was a 286SX, but there was no such thing as an Intel 286SX. Take a look at the reference above that lists the Intel processors before the 80486. The 386SX marketing campaign had a X through the 286 to suggest that users replace the 286 with a 386SX. http://blogs.intel.com/technology/2008/07/intel_chairman_craig_barrett_o.php

The ironic thing is that the only tiny bit of truth given in the article is ridiculed as a joke. the author ridicules the claim that the 486SX was a sort of "crippleware". In fact, this rumor was true. The idea was to offer a lower performance 486SX to target more price sensitive customers while not compromising the pricing of the high performance 486DX. This is a very common practice of price discrimination in business in general, and specifically in hardware and software projects. Think about Microsoft Windows 7 or Vista. Both are offered with various versions that have a subset of the functionality of their "Ultimate" version. Microsoft could just as easily give everyone Ultimate, but by offering partially "crippled" (to use his word) versions of the software they can squeeze more money out of less price sensitive customers that want all the performance they can get while not losing customers that are more price sensitive but willing to give up some functionality.

The joke about the drill through the precisely correct location is likely a joke made by someone who knew what they were talking about. It was likely not meant to be taken literally, but sort of a tease without explanation. The floating point unit was in fact never removed with a laser, it was disabled by tying a "disable floating point" input pad to power or ground using a bond wire in the package. This joke was probably a reference to the fact that if you disconnected that bond wire, the default state of that input pad was to enable the FPU. So I think the idea behind the joke was that you could disconnect that bond wire by drilling through the bond wire and thereby disconnecting it. The practical way to do this would be to flip off the lid on the PGA package to expose the die and bond wires. People did have a steady enough hand to straighten bond wires on occasion when that were shorting to adjacent bond wires. With a knowledge of which bond wire it was, a person with a steady hand could probably remove that bond wire with a tweezer-like tool. This was possible on the early units in a PGA package.

The article claims that 486SX units were 486DX units that had defective FPUs. There is no citation for this and this happens to be false. The current wikipedia article claims this was true only for early units. That is false too. It was never true. I think this is speculation to rationalize why there was a 80486SX, made by people who did not understand that the price discrimination rationale actually is a very well known and well accepted business strategy. Over time more and more of the material from this article was removed from the wikipedia article. Its time to remove the last bits about the defective FPU. —Preceding unsigned comment added by Awitko (talk • contribs) 19:27, 1 January 2010 (UTC)

You are approaching this in completely the wrong way. Citing your own personal experience is not acceptable in place of sources, even when you address yourself in the third person as your recent edits did. Indeed, if anything that introduces a conflict of interest and if you are unable to remain neutral then this guidance applies. If you are unable to provide sources it doesn't go in - you can't cry off saying public sources are not available. As for the accuracy of the free online dictionary of computing a lot of your criticism are either irrelevant or misplaced. Where does it refer specifically to an INTEL 286SX? I have fuzzy memories of way back then but there were a lot of chips on modules and the like. A quick google suggests it may have been an ICT product. Your assertion that a 486 could not be added to a 486SX board is inaccurate too - many boards allowed it. I could go on but I think you get the general idea - your only source is a potentially false appeal to authority but you show yourself to be far less than infallible. I'll drop crispmuncher a note asking if he has those sources though - I agree it has been too long. 87.112.2.151 (talk) 22:45, 3 January 2010 (UTC)

I am citing what I know to be true based on personal experience of having worked on the device as an intel engineer. As I mentioned, the scenario you include in the wikipedia article also does not have sources so it does not seem rational or fair to require me to provide sources when you have included this without sources since the creation of this article. I am sure you will not find sources for either version because this is the sort of thing that Intel did not include in the databooks. It is the sort of thing that PC Magazine and other consumer magazines speculated on, often incorrectly. In this case, I know for a fact that the description is incorrect - I was the test engineer that created the test programs for the first two versions of the 486SX and versions after I was involved had the FPU completely removed. As for the free online dictionary of computing, I provide a link to the article in my comments. Its a short article so if you spend a few seconds on it you will find the reference to 286SX. Its possible some other company made something they called a 286SX but this article was about Intel products. The author(s) pack so many mistakes on basic information about Intel products in such a short article that it would not be advisable to rely on that article for any information. That author claims the 486DX was the upgrade device for the 486SX when in fact the most common upgrade device by far was a 487SX which had the extra pin to automatically disable the 486SX. Its possible that some systems allowed for the 486DX to upgrade the 486SX (I don't recall) but this was not typical and would require a jumper switch on the board to disable the 486SX since the 486DX did not have the extra pin. Although I can see crispmucher updated the page again, it does still does not include references and it still does not resolve the inaccuracies. 23:12, 24 January 2010 68.50.210.67

As for your comment that my pointing to numerous verifiable, fundamental errors in the sole source are irrelevant and misplaced, I think that does nothing but demonstrate your bias and committment to what has been in this article since the beginning. If an article that states 2+2=5 and 6+3=10, someone would be a fool to use that article as a source for the result of a more complex mathematical expression that they could not compute independently. You could split hairs and say that the basic errors are irrelevant because they were addition and the expression in question is not addition, but a wise person would discount the credibility of the source based on the fundamental errors. The source article gets most (and I mean most) of the statements about verifiable fundamental chip details wrong, why would you trust that same article to provide information on things you cannot verify? Its foolish. How do you justify changing its verifiably false statement that ALL 486SX had defective FPUs (false since many had the FPU completely removed - see above) to the wikipedia statement that EARLY units had defective FPUs (without any citations)? Yes, I do now vaguely recall that at least one system did allow a 486DX to upgrade the 486SX with a jumper switch to send the disable signal to the 486SX. But I think schemes like this were the exception, not the rule. The source article states an exception, but neglects to mention the rule. I don't know how anyone even remotely concerned about integrity of information would even consider relying on that article for any information. Again, at the very least you should apply the same standard to the information currently in the wikipedia article and remove information that cannot be supported by citations. --Awitko (talk) 15:05, 5 February 2010 (UTC)

In part I have to agree with that IP editor. The article in question refers to a 286SX. It does not refer to an Intel 286SX. I agree I've never heard of this part either but to suggest it can't exist simply because Intel never made one is putting words into that source's mouth. Similarly the FOLDOC source does not make the assertion you seem to be trying to pin on it. To quote:

Some systems, e.g. Aopen 486SX, allowed a DX to be plugged into...

Now where does that say that all system had that upgrade path exactly?

I do have issue with relying on FOLDOC too heavily. Large parts of it are sourced from the Jargon File which was more a humorous take in many areas than attempting to be a reliable source. However we can find errors in any source - you yourself had to concede ground (re CPUID) last time you bought this up. Should we completely ignore you because that is what you are arguing for?

However other sources are available. As you yourself indicated "The version you describe is plastered all over the internet..." so finding sources is clearly not a problem. I'm still open to altering this along the liens you wish but we need evidence. Repeatedly exclaiming "I was there" and "I did this, that and the other" are not acceptable - you could be anybody as far as this project is concerned. We need solid verifiable evidence and it does not matter how much you cite your personal memories: citing your recollection is no cite at all since it does not fulfil the requisite criteria. I am not saying that is right or wrong, simply that that 'is the threshold for inclusion. CrispMuncher (talk) 19:08, 5 February 2010 (UTC)

Crispmuncher, this is getting to be ridiculous. I already stated that I can accept that my personal knowledge may not be enough, but at least I am bringing attention to the inaccuracy (of which I have no doubt) which should cause you to look more carefully at the sourcing of inconsistent statements. I point out that there are many blatant errors in that particular article from that source and you want to defend it by ignoring all the false statements and focusing on statements that may only be misleading rather than false? It does say some systems used a 486DX upgrade, not all systems, but it neglects to mention the Intel specified upgrade path -- the 487SX. The reasonable inference is that most systems don't have any upgrade path and that in some cases the 486DX was the upgrade path. It is misleading at best. It does say ALL 486SX had defective FPUs but you chose to ignore that verifiably false statement and partially modify it to EARLY 486SX chips had defective FPUs without justification. It talks about the Intel 486SX and 486DX and then mentions a 286SX. The reasonable inference is that this is also an Intel part, but you want to claim that this part *may* have been made by another manufacturer, though you are not aware of it. I did not concede ground on CPUID. In a section below, I pointed out that one reason the 486DX could not have been switched to a 486SX based on FPU testing is that they had distinct hardwired CPUIDs. You claimed that CPUID was not available in that chip, apparently assuming that the only way to access CPUID was the CPUID instruction. I explained to you and provided evidence for the fact that you were wrong -- CPUID did exist for that chip and was accessible. I saw the statement about a defective FPU plastered all over the internet, but not from a reliable source. For all I know this has been repeated based on peoples reading of foldoc, wikipedia and whereever else this rumor started, perhaps speculation by a consumer magazine. Recognize that false information can propagate all over the internet too and some judgement needs to be applied in terms of whether an internet source is reliable. The foldoc article is dated 1997 and this false information was imported from that article into the original 486SX wikipedia article in 2002 so it has had a lot of time to propagate from these sources. If you find an article on the internet that doesn't cite a source it is not unreasonable to conclude that it probably came from your own article. You should find a source with citations that trace back to a reliable source like Intel. But I know this information did not come from Intel. More than a year ago, you promised to produce an Intel source and also claimed that Intel explicitly stated this. Where is the source? At least you should acknowledge in the article that your version of events is a rumor with unknown origin. And if you want to be true to your readers, acknowledge the alternative explanation that I provide even if you have to add the disclaimer that it too has no official source. Otherwise, you really should apply the same standard to your scenario and remove it. Stick to the verifiable databook facts about this processor. --Awitko (talk) 21:12, 6 February 2010 (UTC)

With that I see no point in continuing this discussion. Not content with putting words into sources mouths, you are now doing the same with me. We have discussed this at length. You may legitimately challenge source but only with source. Vague recollections are not acceptable and it is also not acceptable to state they are simply not available. I said this over a year ago and you have still to cite any source. CrispMuncher (talk) 15:35, 7 February 2010 (UTC)

I did not put words in the foldoc article's "mouth" as you call it. One only has to read the reference to see that. I did not put words in your mouth either. One only has to read below to see that. Your post on July 18 2008 (see below) you stated: "Intel certainly said at the time that the FPUs were tested and disabled if not functional." Your short comment attached to your Nov 3 2008 reversion of my edits stated "Only the latest steppings even had a CPUID instruction, and you need one hell of a cite to justify contradicting Intel's statements regarding the defective FPU." When I challenged those assertions including the existence of an Intel source, your Nov 4 2008 reply stated you would find an Intel source: "I'll dig out a cite in the next day or two, but your edit itself admitted the existence of public statements that this was the case. I'm sure I can find a contemporary press release, datasheet or whatever but that means going up in my loft which I'm not doing tonight." That was 15 months ago and you have yet to provide that source. In the same posting you said "Regarding the CPUID instruction, your reference actually confirms my assertion - CPUID was introduced in 1992 whereas the 486SX was introduced a year earlier." I replied: "Crispmuncher. If you look on page 11 of your appnote (http://www.intel.com/Assets/PDF/appnote/241618.pdf), you will see it state: "Later, with the advent of the Intel386 processor, Intel implemented processor signature identification that provided the processor family, model, and stepping numbers to software, but only upon reset." So the CPUID was accessable for the 486SX without using the CPUID instruction." What words did I put in your mouth? Also, again this is internal implementation issue -- there are no Intel publications about it which is why false rumors like that in this article can propagate uncontested. You point out that you don't know who I was within Intel, but you also don't know who the foldoc source was either. It seems that foldoc is a moderated dictionary based on public contributions so this information could have come from anyone. You and others state that making assertions based on my personal experience is not acceptable support for inclusion in wikipedia, but you have no idea what supports the disputed claims included in foldoc. For all you know it is also based solely on personal experience. Or it may be based on a consumer magazine article written by someone who never worked a day in their life in the semiconductor industry. Or it may have been written by a teenager maliciously making up a lie for kicks. You have no idea. This "source" does not identify how they know or where they got this information and they include this information in a cryptic, unprofessionally written article full of verifiable inaccuracies. If this foldoc article is still considered a reliable enough source for wikipedia to use, that is pretty sad. You state that you have an issue with relying on foldoc too heavily (see above), and yet you use foldoc as the sole source for the contested assertions even though you apparently believe that Intel sources exist for them. These assertions have existed for 8 years on wikipedia -- it should have been removed a long time ago for lack of a reliable reference. --Awitko (talk) 16:33, 7 February 2010 (UTC)

And BTW, the person at 87.112.2.151 above (I think that is the IP editor you refer to) also acknowledged that from his point of view you owed the references promised from long ago: He wrote "I'll drop crispmuncher a note asking if he has those sources though - I agree it has been too long."

And I did not say this was a "vague recollection" as you state in your last post. I have said many times I have no doubt about the points I am contesting in the wikipedia article. This is a project i worked on for months after spending about two years working on the 486DX. There is no doubt about the fact that 486SX units were not 486DX units with a defective FPU. There is no doubt that the FPU was not removed by laser. My job was focused on manufacture and test of the 486DX and 486SX chips during that period. I am well aware of the Intel specified way of implementing the upgrade using a 487SX but I did not survey board designs. I spent a lot less time considering the various alternative configurations that some board manufacturers implemented but I do vaguely recall that something like the 486DX/jumper alternative existed. My point was that the article failed to mention the intel specified upgrade path - one of the many fundamental errors and omissions in that source article.

And you claim that you have been asking for a source from me for a year. Not true. See the section below and the explicit cite above. After claiming that Intel was the source of this information, *You* promise *me* on Nov 4 2008 that you will provide an Intel source even though I tell you that this is the sort of thing that Intel does not publish. I do not hear from you for more than a year and its only recently when I follow up on this matter that you try to turn the tables by falsely claiming that you asked me to provide a source (because you no doubt have since figured out that you can't find an Intel source). As I have now said many times, I am fine with not including the truth in the article because I do not have access to a public source supporting what I know to be true. What I now hope to at least accomplish is to have the false information from the article removed by getting you (or someone else at wikipedia) to apply the same standard to that false information. There is no reliable source for that information so it should be removed even in the absence of any competing theory and/or citation. --Awitko (talk) 19:56, 7 February 2010 (UTC)--Awitko (talk) 19:56, 7 February 2010 (UTC)

And as far as the 286SX that you and the IP manager speculate was made by another manufacturer, please show some evidence or acknowledge this is just another one of the many errors in the source article. I did a google search for 286SX and looked through the sites. The first cite for example mentions a 286SX in a chat and a follow up response is that a 286SX never existed. Several of the top responses is for a Powerflex 286SX but there are only 5 references to that on google and 790 references to a powerflex 386sx so those other references are probably typos for the powerflex 386SX. Many other top ranked references refer to a storm door with that model number. Another blog states that his old computer was a 286sx "meaning it did not even have a math coprocessor." That later part makes it clear that he is getting it confused with a 486SX. So if you want to insinuate this is not just another error, please provide a reference to a 286SX chip (from any chip maker) that was a 16 bit processor with 8 bit databus as the foldoc article states. I can't find one. BTW, the IP manager speculated the 286SX was an ICT product. The ICT 286-S-X was apparently a 28-pin integrated circuit socket, not a processor. The foldoc author is likely confusing a 286SX with the 386SX which was a 32-bit processor with a 16-bit databus. Given the consistent near misses in the statements in that source article I am beginning to wonder whether this was a test or a practical joke by someone who knowingly introduced these errors to see how far it would get. And 13 years later it apparently is still being cited. Its a testament to how unreliable the internet can be if people aren't sufficiently skeptical of what they read. And its interesting to see that google reports 11 sites that have essentially identical first paragraph or more of content from the foldoc article and no doubt there are others that are simply copies of the wikipedia article so its not simply an idle theory that wikipedia and foldoc has propagating this false information all over the internet for at least a decade -- the date of the last update on the foldoc article was 1997 but it may very well have been published for years before that. --Awitko (talk) 23:12, 7 February 2010 (UTC)

One thing that is coming back to me that I want to correct here and hopefully eventually have incorporated in the article: I now recall the extra pin on the 487SX was simply a mechanical key so that the end user would know how to orient the upgrade chip in the socket. Without that pin, I believe the user might have rotated it 90 or 180 degrees from the desired orientation and still have been able to insert in the socket. The extra signal I referred to came from one of the other pins -- (I think UP#). This signal line was driven to ground when the upgrade device was present thereby disabling the 486SX. Since it has been almost 20 years, it would be worth confirming this in the 487SX databook before including in the article but I think this is right. Another point - I said below that the option to disable the floating point unit existed in the original 486DX. I speculated that this was just for debug purposes and the idea to create a 486SX may have been an afterthought motivated by price discrimination. But I was not part of the product definition process for the 486DX so for all I know their were plans for the possibility of a 486SX at the time the 486DX was being designed and that was at least one reason why this capability existed on the 486DX. --Awitko (talk) 18:26, 9 February 2010 (UTC)

ZIF Sockets and FPU Usage

It appears that whoever wrote this entry only knows of 486 chips from a history book. The lack of FPU didn't aversely effect contemporary games, because the lack of the FPU was standard on machines of the era. 486 chips didn't have ZIF sockets, they had LIF sockets. The ZIF wasn't introduced until the Pentium was released. 67.169.145.35 (talk) 16:43, 19 January 2009 (UTC)

This is clearly erroneous. ZIF sockets were not ubiquitous in the same way that they are now, but they were certainly in widespread use. I have a couple of socket 2 and 3 motherboards upstairs with ZIF sockets. See http://www.nuggetlab.com/comptia_equipment.htm for images of 486 ZIF sockets. CrispMuncher (talk) 19:27, 19 January 2009 (UTC)

FPU Disabled/Broken

IMHO the 486 SX was not a 486 Dx with a defective fpu. It was a marketing creation a 486 DX with afpu disabled same production costs but less features.

As so far as I've read, the SX actually had higher production costs than a DX, and that it was sold to try and compete with AMD's midend space. Either way, the SX chips were horrendously slow.

I disagree. Intel certainly said at the time that the FPUs were tested and disabled if not functional. This fits in well with contemporary industry observations that they initially had great difficulty ramping up the yield of the 486DX. As for performance, if running with a 32 bit bus (it could be configured to run with a 16 bit bus like the 386SX), it was identical to a similarly clocked 486 if floating point calculations were not used. CrispMuncher (talk) 14:49, 18 July 2008 (UTC)

CrispMuncher I saw your undo of my edit and your comment. I'm not sure how these disagreements get resolved on Wikipedia. I was a product engineer for the i486DX and the first two versions of the i486SX and I managed the small design project for the first 486SX in a plastic package. I wrote the programs that tested the chips in production worldwide so I can tell you for a fact that is not true that units were built according to whether the FPU passed. And I can tell you that a laser did not disconnect the FPU. It was a bond pad that had a bondwire tied to power or ground that disabled the FPU. You mention an Intel cite that supports your point - please point to it. You are also wrong about the CPUID - it did exist. See for example this cite. http://www.gilanet.com/david/Intel486.htm. I was intimately familar with those numbers at the time and those look right to me. So where do we go from here? Awitko (talk) 23:51, 3 November 2008 (UTC)awitko

I'll dig out a cite in the next day or two, but your edit itself admitted the existence of public statements that this was the case. I'm sure I can find a contemporary press release, datasheet or whatever but that means going up in my loft which I'm not doing tonight. Regarding the CPUID instruction, your reference actually confirms my assertion - CPUID was introduced in 1992 whereas the 486SX was introduced a year earlier. My memory was that later steppings introduced the instruction but according to http://www.intel.com/Assets/PDF/appnote/241618.pdf (page 21, and footnotes on page 23) the CPUID instruction never made it to the 486SX line. It's possible this is a simplification and it did appear on some chips, but given that the troublesome statements are already limited to early chips then any discussion of CPUID is obviously an irrelevance.

That is the key point to bear in mind - those comments relate to early examples. I don't assert that there were no later chips which were intended from the beginning to be SXes and lacked an FPU altogether. Given your experience I would be a fool to argue about your knowledge of particular steppings, and certainly I'm sure you have much to add on this article of benefit. However, is it possible that you are mistaken over dates? If the particular designs you were working on were later on in the 486-on-the-desktop period things could easily have changed by the time you started those particular projects. Finally, one of the weaknesses of Wikipedia is that sources are needed for contentious material. Expert opinion such as your own is welcome but still needs to be backed up by publically available information. CrispMuncher (talk) 20:13, 4 November 2008 (UTC)

The version you describe is plastered all over the internet, but it is simply untrue. Absolutely, no doubt. The problem is that this was an internal implementation issue that Intel had no interest in discussing publicly so I am virtually certain you won't find anything in official Intel literature. I was at Intel from 1988 to 1999. I was on the debug team for the original i486DX. The original i486DX had the disable floating point (DFP) pad from the beginning. I think it was initially conceived as a debug option and only later did Intel decided to use it to segment the market. I was involved on the original i486SX from the very beginning. I was on a team of about 12 that were recognized for the implementation of the original i486SX. I was involved in both versions that had an FPU and we always disabled using a bond wire option. And versions after I was involved had the FPU removed. As far as the CPUID, the reference shows CPUID for the early versions of the i486DX which go back to 1989 so that definitely predated the introduction of the i486SX. How the CPUID was accessed is not really relevant here, but although there may not have been an instruction at the time (I can't remember), there was a way to get the CPUID on the databus. For example, when I designed the i486SX for the plastic package, I started with a process shrink version of the i486DX and one of the several changes I made was to change the hardwired CPUID.

Crispmuncher. If you look on page 11 of your appnote you will see it state: "Later, with the advent of the Intel386 processor, Intel implemented processor signature identification that provided the processor family, model, and stepping numbers to software, but only upon reset." So the CPUID was accessable for the 486SX without using the CPUID instruction. Also, if you look at http://www.intel.com/pressroom/kits/quickreffam.htm#i486 at the last page you will see a reference to 486SX versions with (1) 1.2 million transistors on a 1.0 micron process which is the same number of transistors at the 80486DX listed there (thus this version is the one with a FPU that was disabled) and (2) 0.9 million transistors on a 0.8 micron process (this is the one with the FPU removed). I worked on an interim product in the PQFP package that had 1.2 million transistors on the 0.8 micron process - which is not listed there. Awitko (talk) 04:48, 10 November 2008 (UTC)

Crispmuncher. I also want to address your earlier point about i486DX yields. I gather than you believe the motivation behind the i486SX was to salvage defective units during i486DX manufacturing. The i486SX was announced about 20 months after production 486DX units started shipping. At that point, we were running high volumes and had good yields. The motivation behind the i486SX was price discrimination. We could compete with the mid-range AMD product using our lower performance i486SX without having to compromise pricing on our high end i486DX product. You find price discrimination as a way to increase profits in many businesses. Think first class and coach tickets on airlines. If all seats were first class, many of the coach customers would be lost. If all seats were coach, they would lose the premium that the first class customers would have been willing to pay. It works even if the differences in prices have little or nothing to do with the cost of providing the different levels of service. For example, Vista Home, Professional, and Ultimate are three products that have different prices but all three cost Microsoft the same amount to produce. The i486SX strategy also allowed us to sell highly profitable i487 upgrade chips to end users that wanted to add FPU capability to their i486SX PC. Awitko (talk) 17:41, 10 November 2008 (UTC)

And I don't know the origin of the story that i486SX units had defective FPU units that were permanently disconnected by laser. It was certainly convenient for Intel that people believed it. For the early i486SX in a PGA package, one could flip the lid off the package and. if they had a steady hand and knew which bond pad disabled the FPU, they could disconnect the bondwire to reenable the FPU. They probably could have identified the bond pad by comparing the bond wires of an i486DX and an i486SX on the same process (same die size). The chip would still have a CPUID that identifed it as a i486SX - which probably would have caused problems with some software recognizing that an FPU was available, but it would have an enabled FPU like the i486DX. Some of the FPUs would not work because they had not been tested, but many of them would work. The lack of testing did not stop people from overclocking CPUa to run at higher operating frequencies than tested by Intel. Many would have been happy to make that modification to save hundreds of dollars on the i486DX or i487 upgrade. Of course, Intel would not have liked that. BTW, this modification could only have been done in the early units - it could not have been done in the i486SX units that were in a plastic package (not accessible) or the ones with the FPU removed (not available!). Awitko (talk) 19:36, 10 November 2008 (UTC)