Originally posted by: backlinkbuilder I ran into the same issue as agner. Thanks for the help guys I really do appreciate it. I was about to start pulling hair out.
This bot make me laugh, to the favorites.
Originally posted by: extremeseos0007 Really the two companies are competing with each other......which is good for this market
The unregulated competition has a number of undesired effects as I have explained earlier in this thread.
The XOP coding scheme is clear evidence of some of these undesired effects. The XOP prefix byte is 8F, which is a POP instruction in the original x86 instruction set. Overlap with existing instructions can only be avoided if the m-bits have a value bigger than or equal to 8. Intel's VEX prefix does not have this restriction. This makes the compatibility with Intel's VEX prefix imperfect. Instructions with VEX prefix and XOP prefix can't have the same m-bit values, though it would be more logical if they had. This problem could have been avoided if AMD had chosen to use the 62 byte for the XOP prefix, as I have proposed in the beginning of this thread. Even better, AMD could have used part of the huge opcode space defined by the VEX prefix. If some of the new instructions introduced by AMD become so popular that Intel will copy them (as it happened with x64), then Intel have to modify their instruction decoder to handle the different m-bit pattern.
I have no inside information so I don't know what negotiations have taken place between Intel and AMD. But it is obvious, that if Intel had been willing to negotiate and treated AMD fairly, then they would have allowed AMD to use either the 62 byte prefix or part of the VEX opcode space and guaranteed that Intel would not use the same code combinations except when copying instructions invented by AMD.
The fact that AMD has chosen the suboptimal solution of using the 8F byte is evidence that fair negotiations have been impossible. I find it most likely that Intel is to blame for this outcome, since it gives Intel an unfair advantage over AMD. Apparently, AMD have scratched their heads to find an obscure corner of the opcode map that it was unlikely that Intel would use. This has happened before in the history of AMD instruction set extensions.
The market situation now is so that if Intel and AMD happen to assign different instructions to the same code bytes (God forbid it!) then Intel is likely to win. AMD have often had to chose outlandish byte combinations in order to avoid the risk that Intel might have something else in the pipeline that happens to use the same code for something else.
This is a clear market failure causing suboptimal technical solutions and compatibility problems. The costs of this is considerable because no existing opcode can be changed. We have to live with this patchwork of weird instruction codes in all future as long as we are making backwards-compatible microprocessors. Both the hardware industry and the software industry will be affected by these compatibility problems for many years in the future.
It is obvious that no voluntary agreement between AMD and Intel is possible. The only solution is outside intervention.
What would a petition help? Do you think a big multinational company will change its profitable policy just because 100000 random people sign a petition in some Facebook group? And I think it would be difficult even to find 100 people who understand the technical issues. Maybe it would help if we can get some IT journalists and politicians interested in the issue.
Just imagine what if this (not existed yet) petition would be signed by famous and powerful names: John Carmack (id Software), Tim Sweeney (Epic Games) and so on!
Originally posted by: agner What would a petition help? Do you think a big multinational company will change its profitable policy just because 100000 random people sign a petition in some Facebook group?
IT journalists and politicians won't take any participation in this petition until it will be signed by famous programmers.
Originally posted by: agner And I think it would be difficult even to find 100 people who understand the technical issues. Maybe it would help if we can get some IT journalists and politicians interested in the issue.
No, I want IT journalists to write about the issue. Nobody will be interested unless there is public awareness about the problem.
But it is difficult to get media attention about a technical issue that people don't understand. Most people will just think, as extremeseos0007 wrote:
Really the two companies are competing with each other......which is good for this market
This will be the initial reaction of most people. Economists have told us again and again that competition is good. But competition is not good for compatibility. Maybe standard organizations like IEEE or ISO could be interested.
Originally posted by: sj1009I think that Intel won't make any steps to create such a public standardization committee.
Of course not. The present situation gives Intel a big competitive advantage over AMD. That's why somebody has to put pressure on Intel to cooperate. I don't know if AMD can sue Intel for unfair competition. I think they have tried that already, but I don't know. At least that would be difficult and expensive and take a very long time.
The pressure might come from government organizations, from the EU, from professional organizations like IEEE, from the software industry, from IT journalists, from political organizations, ...