Sorry about missrembering which Px had the 20 stage pipeline.
As for Hyper Pipeline, it requires that the stage that you intend to jump to is empty, i.e. a bubble in the pipeline, which probably makes it's usefulness limited. Ahmdals law is a good thing to apply (Intel seems to miss that sometimes). I would like to say that these kind of small improvements simply increase the complexity of the construction.
As for the P4's BPU, good or bad, it will still fail sometimes. It is not possible to predict all jump properly, thus you will have big penalties when not doing so if you use a big number of pipeline stages.
I'm not claiming that the idea is new nor that it is mine. I want the mainstream processors to use it. There are lots of variations on this, but I don't think that the VLIW (ok, EPIC in Intel lingo) thingie will yeild better performance than truly separated CPU cores on one chip and the latter is probably easier to scale if done right.
How long will this hunt for more GHz continue? I'd say that if the major industry companies (Intel, AMD...) would make a since long needed move to a better architecture we could achieve more performance with less means.
What do I have against high frequencies? For starters, high speed, fully syncronized digital constructions rely on switching millions of transistors at the same time (each clock cycle), this burns lots of power which is a limiting factor today.
Also, high frequency does not imply high performance, the CPU still needs to do something each stage, for example older Pentiums (P3, if I remember right) had a 20 (yes twenty) stage pipeline. This yeilds huge penalties for miss predictions for branches etc.
This GHz hunting also leads to other problems, such as huge electromagnetic disturbances in the chip, and in busses, etc. The solution to this is to add more wires and pull them in different directions to compensate. This only wastes more power and emits even more heat.
What I suggest, now when we have lots of transistors to play with, are asyncronous designs! Yes they are harder to design and verify, but that is largely because the lack of supporting tools.
This would reduce the power needs, let the designers make longer critical paths in their constructions (just clock that part slower), and reduce the need for registers used to balance pipe-lines etc.
Another move could be to introduce simpler, but parallell CPUs, perhaps on the same piece of silicon. The software systems of today are multi-threaded already, so why not make the hardware capable of _true_ multi tasking...
Of course the will not copy BIOS, motherboard layout, etc. if they where to do a new Pentium based platform. I'm stating that.
The last section is about the micro processor architecture (I try to point that out). What I'm stating is that the PPC is a RISC machine, which is better, from an architecural point of view, than an x86, that is CISC machine.
The only reason I can see to use a CISC machine today is backwards compatability. This is invalid in the current case, as Apple will have to recompile the whole thing anyway if they are going to make a platform switch.
If Apple was to do Intel (read Px) based hardware, would they then go for a standard PC? Probably not as this means that their users can go to Win or *nix too easily. As they then would have to develop their own special little system, they would still have performance problems (fewer bucks spent on HW development) and expensive hardware (monopoly, or close to).
Since this rumour has been around for a long time without anything actually happening, I'd say that Apple will keep on building proper RISC based machines. We can all agree that it would be a step backwards to go from PPC to x86 from an architectual point of view, can't we.
The only super hero I care for is Tux. I cannot comment on the classic super heroes, as the site could not bear the load. Since there are several tux sites, they will probably survive the load better. Check
here,
here,
here,
here,
here,
here or
here.
I'd say that this looks more like an Apache worm than a Linux worm. It does not seem too bad though, "Get your Apache systems patched and update your antivirus software and you should be fine." (from the Slapper.C article).
This shows that Linux+Apache is so widely accepted that it is a legitimate virus target. Enjoy it!
This is too early to give in to Microsoft. As neither version has any significant market advantage yet it is not good to make the systems one-way compatible. This only makes it easier for customers to move to.Net, not the otherway around.
The priority must be to compete with.Net, not to become the little brother of it. There are a number of points that need to be equally good/better than.Net:
1. Ease of use (both user-wise and coder-wise).
2. Security and user control of information
3. User base (on both sides again).
The first point is the reason of the project from the start and must be maintained.
The second point is the advantage, no-one can reach me, and on-one can reach the customer-records of a competing company without authorization. Not only geek users should be afraid of giving too much info away, also the companies utilizing these platforms must be aware and protect their customer bases.
The third point is probably the pass/fail issue of the entire project. It must get adopted, from the average user and by the service providing companies.
This is a good point to insert Linux onto the desktop of average Joe. Despite Lindows licensing tricks they open a door for open source software.
As Windows licenses are expensive it would be nice to see for example Dell to supply computers pre-installed with Lindows. This would still let the average user to use Excel, Word, etc. But allows easy installation and adoption of applications such as KOffice, Gimp etc.
Today it is hard to get a Windows computer to run *nix (open source) apps in a native looking way, but this could change all that.
The skill of writing small games is highly relevent again as new mobile phones with the ability to run custom software is starting to appear.
Also, it would be nice if some game programmers ever asked them selves 'can I do this with less code and memory consumption?'. The answer would most probably be yes in many cases.
An example of this is the use of large look-up tables instead of doing a medium-sized calculation. As the caches can't hold an entire game today, the penalty for using a large look-up table is probably pretty big, since it is spread over a larger section of the memory than a medium (properly aligned) calculation routine.
When discussing code size, I must say that the best (in the amusing, fun, addictive way) I've had was Super Cars II on the Amiga. Great game, only one (or two) 720kB floppys. It had what many of today's games lack of: gameplay.
* Lower Total Cost of Ownership for the desktop, including lower user training costs.
We've got problems at my work with people thinking that they are fully fledged programmers since they can record two macros and cut'n'paste the results into a super-macro. Of course users need to be educated, otherwise they will not be able to use the applications properly. (One example is people insisting on using spaces when they try to indent text, then go to the IT department and complain about the lines not being properly aligned...)
* Centralized and efficient security policy administration.
Oh, what? Surely one can pull the TP-cable out of *nix boxes too. Even the 'central' one in the basement... Security can not be a reason to use M$ software.
* Greater flexibility and management functionality from products that offer a broader range of management solutions that integrate with non-Microsoft environments.
This is M$ key to new markets. Take a standard, implement it, expand it in your solution in order to make your app 'integrate' with others, but not the other way around. A good application should be able both to import and export data properly. (M$ Word RTFs crash my FrameMaker... portable format - not).
* Greater productivity and reliability attributed to less downtime.
Again, you do not get less downtime by buying an expensive system with big flaws. They probably pay loads of $$$ to get a guaranteed time to support arrives and press the 'reset' button. *nix usually do not fail as ofter as Win*, thus no need to advertise that support will arrive in 2.3ms. The lack of service can be because it is not needed, not because it is an ingnored flaw.
* Extended support for a large base of software applications.
Since most advertised software is commersial, and they probably do not look for software them selves (just ask for it in a formal way and have companies make offers). Just use KDE as the German government intends to do. This does not only give a better quality of the software, but also save loads of license $$$.
But since Bill payed Bush's campain, Bush has to give the money back to Bill. As he doesn't fancy paying up at few $$$, he just takes the $$$ from the tax payers. Bomb the hell out of a few arabs and the software sums looks small in the contents.
It seems like this is a part of HP's plan to 'do an IBM', i.e. become a provider of complete solutions (HW+SW+Consulting). They've got the HW and consulting, but still need a big SW platform to sell and promote.
As for scaryness, yes it is a threat to the freedom online. We have to hope that Liberty Alliance will succeed and that average Joe will become aware of the lack of integrity this type of solutions can result in.
Still, there are mobil phone networks there too (Koria, USA, Canada) that can handle speech. Thus making the 3G introduction slower than expected.
As for the typing, I too prefer to speak. If I want to send a short text message I am not ready to pay the _outrageous_ prices for SMSs charged here in Sweden (Euro 0.1). I prefer sending e-mail using a palmtop and GPRS (2.5G, i.e. slow packet based radio transfers). I can type (scribble?) faster/easier/better on my Palm, and send it cheaper over GPRS (charged per MB).
You mention geographical coverage, 3G will hopefully be even better. Image one number, one phone and one bill for the entire world.
About not needing 3G, do you need the internet? Do you need a coke? Marketing and selling a product is not about needing, it is about wanting. To make 3G desireable, it must add some value to the user, for example gaming, on-line ordering, trafic information, p0rn, you name it. We need contents to the technology.
Concerning shares, I'm actually thinking about buying shares in Ericson. This is a huge company with an enormous know-how and competative technology. Their problem is the outrageous pricing of the 3G licenses on the European continent. In Sweden (among other) there was a 'beauty contest' instead of an auction. This shows that the Swedish state has (despite their incompetent government) realised that it will pay off in the end if the 3G market can take of, instead of trying to make short-term profits on selling licenses and bankrupting the service providers.
3G will not fail! Everyone just need to remember that it took more than 10 years for GSM to explode here in Europe, and it will probably take even longer for 3G since GSM allready handles talking. What the suppliers do not see is that we lack good services and a good way of charging for it. What is needed is a global standard for micropayments. I think that it would be great to get all the micropayments on my phonebill, even better if I could surf over to my service provider and check my spendings over the web using my phone/computer/camera/mp3-player/calendar/gamecent re...
And by the by, why isn'n there a plug-in enabling the new photo and video phone to show their images on a TV (when connected to the powergrid, the batteries will burn otherwise), and a plug-in to be able to play more advanced games... it would be (ta-da) the return of the cartridge games...
The Swedish city if Gothenburg (quite close to Oslo) has a similar system. I do not think that it uses GPS, but rather tags that tell sensors at each stop where each veichle is. At each stop there are signs either giving a time (if the location of the veichle is known) or a time, prefixed by 'ca', to show that it is using the time table. The system has been developed by a Swedish company called Hogia Persontrafiksystem AB. The site is in Swedish, but holds information about their products that can be interesting (since you discuss Oslo, I presume that you can read Swedish, it is closely related to Norweigan).
If you have a virus threat problem, I assume that you are connected to an external (out of your control) network. If this network happens to be the internet (most likely), just try Symantec's Security Check. It scans for viruses over the net (with a bit of ActiveX magic... It seems that M$ security misses can be useful sometimes;-P)
My key point is not only that my employer would gain from it (he does sponsor both copy+computer+internet access). The software house would also gain from it, so they should not only have this licensing form, they should promote it. If I knew that if I used tool X for work I would also get to use tool X at home, legaly, for free, without having to bug my employes about it. Also, I do not think that a requirement that a percentage of the home use has to be work orientated is good, I use the app and learn how to use it with my own projects. The interest is not in me working from home, but me gaining experience from home (and not having to take a $20k course every once in a while).
I feel that all software houses selling propretary software should allow several installs of their products. If I use a product at work (for example a word processor) and I want to use the same product at home, I do not see the problem. I would never pay $5k for an office suite, but rather use a free alternative or a pirated copy. This would be good for my employer too, since I would gain experience with the software that I'm supposed to use at work. I'd even stretch it to being good for the software house, as I have experience from their product I'd want it at my next work too.
On the other side, I do not see why a company making commersial use of a free alternative should be able to do so without giving back to the community.
As a good example of good licensing I would like to point out trolltech's licensing model (faq here)for Qt. I can play around with it and make GPL'ed software without any fees, but if I want to get some $$$ from selling the software I need to give them some $$$ first.
I subscribed for the major newspaper in my region (western Sweden) and decided to drop my subscription after three months due to two problems: 1) I have to carry my paper recycling to the recycle station once a week instead of once every other month, 2) I had to get up 30 minutes earlier than usual each morning because I browsed the articles and always found something interesting.
Now I read the same articles on quiets moments from work, where I also do my bigger downloads and burn them out on a CDRW and bring them home. At home I have a 56k modem which is good enough for mail and/. on the weekends and the odd app that needs registration.
As for giving up TV, yes, it is a good idea. I think that I will miss the local news though, as they are difficult to find on the net.
Don't flame me for saying this, but if Linux is ever going to become a mainstream desktop OS this is very needed. An average office user does not care about Gnome/KDE differences (even though I as a programmer prefer KDE). We geeks must realize that enterprice isn't interested in themeabilities, cool features, etc. but _stability_ and _productivity_.
What RH is doing is not removing all the features, just streamlining. As anyone interested in the extras can reactivate them quite easily, this kind of standardization must be a _good_ thing for the community.
I would not recommend anyone to optimize modern x86 asm by hand. If you know your way around disassembled code you know enough to find any (rare) compiler mistakes. Any other operation is usually done better by a compiler. (Please don't yell at me with small, hand optimized special cases, compilers do a good job today if your application isn't very special.)
If you would try to hand optimize asm code for a modern cpu you must concider many issues, among them the reordering of instruction in the processor, the different layouts of the pipelines in different processor models (even intels differ to other intels), cache effects (I suppose that you must link everything statically and control where in memory your code will end up)...
You must also unroll loops, change the access patterns to 2D data structures to improve cache performance, avoid inner loop data dependencies, etc. It is simply too much to handle by hand.
As you probably know a higher level language such as C/C++ and don't write a highly optimized operating system (or work without an OS) you do not need, and should not want to, optimize your asm code by hand!
Slashdot Mirror
Sorry about missrembering which Px had the 20 stage pipeline.
As for Hyper Pipeline, it requires that the stage that you intend to jump to is empty, i.e. a bubble in the pipeline, which probably makes it's usefulness limited. Ahmdals law is a good thing to apply (Intel seems to miss that sometimes). I would like to say that these kind of small improvements simply increase the complexity of the construction.
As for the P4's BPU, good or bad, it will still fail sometimes. It is not possible to predict all jump properly, thus you will have big penalties when not doing so if you use a big number of pipeline stages.
I'm not claiming that the idea is new nor that it is mine. I want the mainstream processors to use it. There are lots of variations on this, but I don't think that the VLIW (ok, EPIC in Intel lingo) thingie will yeild better performance than truly separated CPU cores on one chip and the latter is probably easier to scale if done right.
How long will this hunt for more GHz continue? I'd say that if the major industry companies (Intel, AMD...) would make a since long needed move to a better architecture we could achieve more performance with less means.
What do I have against high frequencies? For starters, high speed, fully syncronized digital constructions rely on switching millions of transistors at the same time (each clock cycle), this burns lots of power which is a limiting factor today.
Also, high frequency does not imply high performance, the CPU still needs to do something each stage, for example older Pentiums (P3, if I remember right) had a 20 (yes twenty) stage pipeline. This yeilds huge penalties for miss predictions for branches etc.
This GHz hunting also leads to other problems, such as huge electromagnetic disturbances in the chip, and in busses, etc. The solution to this is to add more wires and pull them in different directions to compensate. This only wastes more power and emits even more heat.
What I suggest, now when we have lots of transistors to play with, are asyncronous designs! Yes they are harder to design and verify, but that is largely because the lack of supporting tools.
This would reduce the power needs, let the designers make longer critical paths in their constructions (just clock that part slower), and reduce the need for registers used to balance pipe-lines etc.
Another move could be to introduce simpler, but parallell CPUs, perhaps on the same piece of silicon. The software systems of today are multi-threaded already, so why not make the hardware capable of _true_ multi tasking...
Of course the will not copy BIOS, motherboard layout, etc. if they where to do a new Pentium based platform. I'm stating that.
The last section is about the micro processor architecture (I try to point that out). What I'm stating is that the PPC is a RISC machine, which is better, from an architecural point of view, than an x86, that is CISC machine.
The only reason I can see to use a CISC machine today is backwards compatability. This is invalid in the current case, as Apple will have to recompile the whole thing anyway if they are going to make a platform switch.
If Apple was to do Intel (read Px) based hardware, would they then go for a standard PC? Probably not as this means that their users can go to Win or *nix too easily. As they then would have to develop their own special little system, they would still have performance problems (fewer bucks spent on HW development) and expensive hardware (monopoly, or close to).
Since this rumour has been around for a long time without anything actually happening, I'd say that Apple will keep on building proper RISC based machines. We can all agree that it would be a step backwards to go from PPC to x86 from an architectual point of view, can't we.
The only super hero I care for is Tux. I cannot comment on the classic super heroes, as the site could not bear the load. Since there are several tux sites, they will probably survive the load better. Check here, here, here, here, here, here or here.
I'd say that this looks more like an Apache worm than a Linux worm. It does not seem too bad though, "Get your Apache systems patched and update your antivirus software and you should be fine." (from the Slapper.C article).
This shows that Linux+Apache is so widely accepted that it is a legitimate virus target. Enjoy it!
This is too early to give in to Microsoft. As neither version has any significant market advantage yet it is not good to make the systems one-way compatible. This only makes it easier for customers to move to .Net, not the otherway around.
.Net, not to become the little brother of it. There are a number of points that need to be equally good/better than .Net:
The priority must be to compete with
1. Ease of use (both user-wise and coder-wise).
2. Security and user control of information
3. User base (on both sides again).
The first point is the reason of the project from the start and must be maintained.
The second point is the advantage, no-one can reach me, and on-one can reach the customer-records of a competing company without authorization. Not only geek users should be afraid of giving too much info away, also the companies utilizing these platforms must be aware and protect their customer bases.
The third point is probably the pass/fail issue of the entire project. It must get adopted, from the average user and by the service providing companies.
This is a good point to insert Linux onto the desktop of average Joe. Despite Lindows licensing tricks they open a door for open source software.
As Windows licenses are expensive it would be nice to see for example Dell to supply computers pre-installed with Lindows. This would still let the average user to use Excel, Word, etc. But allows easy installation and adoption of applications such as KOffice, Gimp etc.
Today it is hard to get a Windows computer to run *nix (open source) apps in a native looking way, but this could change all that.
OSNews ran this a few weeks ago. It is gold, unless you sold/threw away/lost your floppys too.
The skill of writing small games is highly relevent again as new mobile phones with the ability to run custom software is starting to appear.
Also, it would be nice if some game programmers ever asked them selves 'can I do this with less code and memory consumption?'. The answer would most probably be yes in many cases.
An example of this is the use of large look-up tables instead of doing a medium-sized calculation. As the caches can't hold an entire game today, the penalty for using a large look-up table is probably pretty big, since it is spread over a larger section of the memory than a medium (properly aligned) calculation routine.
When discussing code size, I must say that the best (in the amusing, fun, addictive way) I've had was Super Cars II on the Amiga. Great game, only one (or two) 720kB floppys. It had what many of today's games lack of: gameplay.
Ok, lets cover them one by one:
* Lower Total Cost of Ownership for the desktop, including lower user training costs.
We've got problems at my work with people thinking that they are fully fledged programmers since they can record two macros and cut'n'paste the results into a super-macro. Of course users need to be educated, otherwise they will not be able to use the applications properly. (One example is people insisting on using spaces when they try to indent text, then go to the IT department and complain about the lines not being properly aligned...)
* Centralized and efficient security policy administration.
Oh, what? Surely one can pull the TP-cable out of *nix boxes too. Even the 'central' one in the basement... Security can not be a reason to use M$ software.
* Greater flexibility and management functionality from products that offer a broader range of management solutions that integrate with non-Microsoft environments.
This is M$ key to new markets. Take a standard, implement it, expand it in your solution in order to make your app 'integrate' with others, but not the other way around. A good application should be able both to import and export data properly. (M$ Word RTFs crash my FrameMaker... portable format - not).
* Greater productivity and reliability attributed to less downtime.
Again, you do not get less downtime by buying an expensive system with big flaws. They probably pay loads of $$$ to get a guaranteed time to support arrives and press the 'reset' button. *nix usually do not fail as ofter as Win*, thus no need to advertise that support will arrive in 2.3ms. The lack of service can be because it is not needed, not because it is an ingnored flaw.
* Extended support for a large base of software applications.
Since most advertised software is commersial, and they probably do not look for software them selves (just ask for it in a formal way and have companies make offers). Just use KDE as the German government intends to do. This does not only give a better quality of the software, but also save loads of license $$$.
But since Bill payed Bush's campain, Bush has to give the money back to Bill. As he doesn't fancy paying up at few $$$, he just takes the $$$ from the tax payers. Bomb the hell out of a few arabs and the software sums looks small in the contents.
It seems like this is a part of HP's plan to 'do an IBM', i.e. become a provider of complete solutions (HW+SW+Consulting). They've got the HW and consulting, but still need a big SW platform to sell and promote.
As for scaryness, yes it is a threat to the freedom online. We have to hope that Liberty Alliance will succeed and that average Joe will become aware of the lack of integrity this type of solutions can result in.
Yes, you are right. I like cows, but not their dung... :-)
Still, there are mobil phone networks there too (Koria, USA, Canada) that can handle speech. Thus making the 3G introduction slower than expected.
As for the typing, I too prefer to speak. If I want to send a short text message I am not ready to pay the _outrageous_ prices for SMSs charged here in Sweden (Euro 0.1). I prefer sending e-mail using a palmtop and GPRS (2.5G, i.e. slow packet based radio transfers). I can type (scribble?) faster/easier/better on my Palm, and send it cheaper over GPRS (charged per MB).
You mention geographical coverage, 3G will hopefully be even better. Image one number, one phone and one bill for the entire world.
About not needing 3G, do you need the internet? Do you need a coke? Marketing and selling a product is not about needing, it is about wanting. To make 3G desireable, it must add some value to the user, for example gaming, on-line ordering, trafic information, p0rn, you name it. We need contents to the technology.
Concerning shares, I'm actually thinking about buying shares in Ericson. This is a huge company with an enormous know-how and competative technology. Their problem is the outrageous pricing of the 3G licenses on the European continent. In Sweden (among other) there was a 'beauty contest' instead of an auction. This shows that the Swedish state has (despite their incompetent government) realised that it will pay off in the end if the 3G market can take of, instead of trying to make short-term profits on selling licenses and bankrupting the service providers.
3G will not fail! Everyone just need to remember that it took more than 10 years for GSM to explode here in Europe, and it will probably take even longer for 3G since GSM allready handles talking. What the suppliers do not see is that we lack good services and a good way of charging for it. What is needed is a global standard for micropayments. I think that it would be great to get all the micropayments on my phonebill, even better if I could surf over to my service provider and check my spendings over the web using my phone/computer/camera/mp3-player/calendar/gamecent re...
And by the by, why isn'n there a plug-in enabling the new photo and video phone to show their images on a TV (when connected to the powergrid, the batteries will burn otherwise), and a plug-in to be able to play more advanced games... it would be (ta-da) the return of the cartridge games...
The Swedish city if Gothenburg (quite close to Oslo) has a similar system. I do not think that it uses GPS, but rather tags that tell sensors at each stop where each veichle is. At each stop there are signs either giving a time (if the location of the veichle is known) or a time, prefixed by 'ca', to show that it is using the time table. The system has been developed by a Swedish company called Hogia Persontrafiksystem AB. The site is in Swedish, but holds information about their products that can be interesting (since you discuss Oslo, I presume that you can read Swedish, it is closely related to Norweigan).
If you have a virus threat problem, I assume that you are connected to an external (out of your control) network. If this network happens to be the internet (most likely), just try Symantec's Security Check. It scans for viruses over the net (with a bit of ActiveX magic... It seems that M$ security misses can be useful sometimes ;-P)
My key point is not only that my employer would gain from it (he does sponsor both copy+computer+internet access). The software house would also gain from it, so they should not only have this licensing form, they should promote it. If I knew that if I used tool X for work I would also get to use tool X at home, legaly, for free, without having to bug my employes about it. Also, I do not think that a requirement that a percentage of the home use has to be work orientated is good, I use the app and learn how to use it with my own projects. The interest is not in me working from home, but me gaining experience from home (and not having to take a $20k course every once in a while).
I feel that all software houses selling propretary software should allow several installs of their products. If I use a product at work (for example a word processor) and I want to use the same product at home, I do not see the problem. I would never pay $5k for an office suite, but rather use a free alternative or a pirated copy. This would be good for my employer too, since I would gain experience with the software that I'm supposed to use at work. I'd even stretch it to being good for the software house, as I have experience from their product I'd want it at my next work too.
On the other side, I do not see why a company making commersial use of a free alternative should be able to do so without giving back to the community.
As a good example of good licensing I would like to point out trolltech's licensing model (faq here)for Qt. I can play around with it and make GPL'ed software without any fees, but if I want to get some $$$ from selling the software I need to give them some $$$ first.
I subscribed for the major newspaper in my region (western Sweden) and decided to drop my subscription after three months due to two problems: 1) I have to carry my paper recycling to the recycle station once a week instead of once every other month, 2) I had to get up 30 minutes earlier than usual each morning because I browsed the articles and always found something interesting. /. on the weekends and the odd app that needs registration.
Now I read the same articles on quiets moments from work, where I also do my bigger downloads and burn them out on a CDRW and bring them home. At home I have a 56k modem which is good enough for mail and
As for giving up TV, yes, it is a good idea. I think that I will miss the local news though, as they are difficult to find on the net.
Don't flame me for saying this, but if Linux is ever going to become a mainstream desktop OS this is very needed. An average office user does not care about Gnome/KDE differences (even though I as a programmer prefer KDE). We geeks must realize that enterprice isn't interested in themeabilities, cool features, etc. but _stability_ and _productivity_.
What RH is doing is not removing all the features, just streamlining. As anyone interested in the extras can reactivate them quite easily, this kind of standardization must be a _good_ thing for the community.
I would not recommend anyone to optimize modern x86 asm by hand. If you know your way around disassembled code you know enough to find any (rare) compiler mistakes. Any other operation is usually done better by a compiler. (Please don't yell at me with small, hand optimized special cases, compilers do a good job today if your application isn't very special.)
If you would try to hand optimize asm code for a modern cpu you must concider many issues, among them the reordering of instruction in the processor, the different layouts of the pipelines in different processor models (even intels differ to other intels), cache effects (I suppose that you must link everything statically and control where in memory your code will end up)...
You must also unroll loops, change the access patterns to 2D data structures to improve cache performance, avoid inner loop data dependencies, etc. It is simply too much to handle by hand.
As you probably know a higher level language such as C/C++ and don't write a highly optimized operating system (or work without an OS) you do not need, and should not want to, optimize your asm code by hand!