AMD64 = IA-32e

[Black Jack]

Regarding whether Intel is going to be using the AMD64 language or not, and
answer is not, it's using the IA-32e language. Other than the name it's
exactly the same as AMD64, right down to the number of registers, and lack
of support for Virtual-8086 mode, etc.

Here is AMD's AMD64 documentation:
http://www.amd.com/us-en/Processors/DevelopWithAMD/0,,30_2252_875_7044,00.html

Here is Intel's IA-32e documentation:
Vol 1: http://www.intel.com/technology/64bitextensions/300834.htm
Vol 2: http://www.intel.com/technology/64bitextensions/300835.htm

It doesn't even look like Intel did much to change the formatting of their
PDFs from AMD's.

Yousuf Khan

 

[steve harris]

Regarding whether Intel is going to be using the AMD64 language or not, and
answer is not, it's using the IA-32e language. Other than the name it's
exactly the same as AMD64, right down to the number of registers, and lack
of support for Virtual-8086 mode, etc.

Here is AMD's AMD64 documentation:
http://www.amd.com/us-en/Processors/DevelopWithAMD/0,,30_2252_875_7044,00.html

Here is Intel's IA-32e documentation:
Vol 1: http://www.intel.com/technology/64bitextensions/300834.htm
Vol 2: http://www.intel.com/technology/64bitextensions/300835.htm

It doesn't even look like Intel did much to change the formatting of their
PDFs from AMD's.

Yousuf Khan

http://www.investorshub.com/boards/read_msg.asp?message_id=2404565

 

[joe smith]

Here is Intel's IA-32e documentation:

http://www.investorshub.com/boards/read_msg.asp?message_id=2404565

This is all very good news, even for little troll a-holes like me. I must
applaud Intel's decision to put their pride aside and doing a very good
decision for themselves and their customers. Don't have to think of
medium-term (shy on using long-term in context of desktop PC's..) effects of
the investment: the latest cutting-edge X86 extended instructions will be
usable no matter which vendor's way you go on this regard. SSE3 is another
issue, but atleast can't go wrong as far as 64-bit X86 spin-offs are
concerned. Very nice.

From technical perspective, it looks amazingly nice to look at the number
and width of the registers side-by-side (yes, identical sets.. that's the
point). I don't know if Microsoft's earlier announcement regarding support
for more than one 64-bit X86 based architechture has anything to do with
this, but if it has, I'm glad Microsoft used their monopoly position FOR
ONCE (okay, this is a bit strong wording but just think is mildly amusing)
for common good.

Without possibility of variety, things can go stagnant easily.. this is why
systems where most available architechtures work (thinking of UNIX and the
mindset that goes with that) without being locked to a single (or few)
hardwired standards like what kind of CPU is inside the machine, is a good
thing. I'm ofcourse referring to compiling sourcecode against widely adopted
standards like POSIX and X11 for instance.

The bytecode mindset Microsoft has in mind, which is processor architechture
independent (Yup, that's .NET, atleast in theory is a way forward which
will give wider opportunity to optimize (okay, atleast match) shipped
application with the client's hardware better. This is a step forward, even
if at first it is a step backwards (as far as performance is concerned, .NET
is yet to demonstrate being of higher performance than traditional languages
like C or C++ for example, but it's getting closer all the time, and when
the packed datatype support/optimizations and better runtimes arrive, who
knows when, or when the difference on things that count is made neglible by
the Moore's Law there shouldn't be much room for complaints).

That put aside, as long as statically compiled software (or dll runtimes)
are the way most of the software is written for the desktop, it's atleast a
good thing overall for the industry and customers alike that there is less
different architechtures that have to be taken care of. It's easy to go
where the fence is the slowest. It's the nature of man. If the fence is the
lowest, where work of a few benefit the many, that is the route that leads
to best results -on average-, and that's what .NET has chance of delivering.
If certain company plays it's cards right and does it's job well enough.
Meanwhile, this thread I am replying to, for some odd reason make me think
that 64-bit adoptation rate will be increased, which is not the part that
makes the developer inside me smile, it's the fact that there will be MORE
REGISTERS, which shows as much as 15% average performance increase with only
recompilation, and this is with early versions of compilers for the new,
improved ISA. IA32e, AMD64, whatever it's being called at the time..

Ofcourse I could be ****ing totally wrong, in that case forget it, I'm just
trolling anyway. If I was right, please worship me.. I was ofcourse serious.
Right.

 

[gaffo]

Regarding whether Intel is going to be using the AMD64 language or not, and
answer is not, it's using the IA-32e language. Other than the name it's
exactly the same as AMD64, right down to the number of registers, and lack
of support for Virtual-8086 mode, etc.

Here is AMD's AMD64 documentation:
http://www.amd.com/us-en/Processors/DevelopWithAMD/0,,30_2252_875_7044,00.html

Here is Intel's IA-32e documentation:
Vol 1: http://www.intel.com/technology/64bitextensions/300834.htm
Vol 2: http://www.intel.com/technology/64bitextensions/300835.htm

It doesn't even look like Intel did much to change the formatting of their
PDFs from AMD's.

Yousuf Khan

Virtual-8086 mode?........what is that?

so the AMD64 chip will not be able to run 16-bit XT programs natively?

--





RUSSERT: Are you prepared to lose?

BUSH: No, I'm not going to lose.

RUSSERT: If you did, what would you do?

BUSH: Well, I don't plan on losing. I've got a vision for what I want to
do for the country.
See, I know exactly where I want to lead.................And we got
changing times
here in America, too., 2/8/04


"And that's very important for, I think, the people to understand where
I'm coming from,
to know that this is a dangerous world. I wish it wasn't. I'm a war
president.
I make decisions here in the Oval Office in foreign policy matters with
war on my mind.
- pResident of the United State of America, 2/8/04


"Let's talk about the nuclear proposition for a minute. We know that
based on intelligence, that he has been very, very good at hiding
these kinds of efforts. He's had years to get good at it and we know
he has been absolutely devoted to trying to acquire nuclear weapons.
And we believe he has, in fact, reconstituted nuclear weapons."
- Vice President Dick Cheney, on "Meet the Press", 3/16/03


"I don't know anybody that I can think of who has contended that the
Iraqis had nuclear weapons."
- Defense Secretary Donald Rumsfeld, 6/24/03


"I think in this case international law
stood in the way of doing the right thing (invading Iraq)."
- Richard Perle


"He (Saddam Hussein) has not developed any significant capability with
respect to weapons of mass destruction. He is unable to project
conventional power against his neighbours."
- Colin Powell February 24 2001


"We have been successful for the last ten years in keeping
him from developing those weapons and we will continue to be successful."

"He threatens not the United States."

"But I also thought that we had pretty
much removed his stings and frankly for ten years we really have."

'But what is interesting is that with the regime that has been in place
for the past ten years, I think a pretty good job has been done of
keeping him from breaking out and suddenly showing up one day and saying
"look what I got." He hasn't been able to do that.'
- Colin Powell February 26 2001

 

[Rob Stow]

Virtual-8086 mode?........what is that?

so the AMD64 chip will not be able to run 16-bit XT programs natively?

I'd hate to generalize from one very limited test, but about a
week ago I was unable to boot an Opty dualie or an Athlon64 from
a MS-DOS 3.3 floppy. No problems with MS-DOS 6.22.

I would have tried an even earlier DOS, but 3.3 was the oldest
disk image we could find - not that we tried too hard

 

[Alex Johnson]

Virtual-8086 mode?........what is that?

so the AMD64 chip will not be able to run 16-bit XT programs natively?

Do a google search for "virtual 8086" (not quoted). You will find
plenty of explanations in the first page.

V86 allows a 32-bit protected processor to emulate multiple 8086
processors by restricting access to some instructions, changing fault
behavior, and locking down translation and addressing to a special
subset of what the chip normally allows.

You might be able to run real mode (DOS for example) but you will not be
able to run 64-bit OS or programs that put the processor into extended
mode and then use V86 mode access to BIOS. Through "legacy mode" you
can still do 32-bit OS + V86, if I understand correctly.

Alex

 

[jack]

<snip>

OT to your post, but did you find yerself a new alias, Yousuf? ;-)

J.

 

[G]

Virtual-8086 mode?........what is that?

so the AMD64 chip will not be able to run 16-bit XT programs natively?

From what others have written, you should have no problems in Real
Mode (ie actually booting to DOS), or V86 Mode under 386 Mode (ie
running 32 bit Windows).

But apparently in 64 bit extended (long?) mode it drops V86 Mode.

So you can run 32bit and 16bit apps together, or 32bit and 64bit apps
together, but not all three at once.

I think.

 

[spinlock]

Wow, those guys at AMD sure are smart.

They created x86-64 before Intel.

Who would have thought to extend the EAX register
from 32 to 64 bits! And then they extended all the
other registers to 64 bits. Infuckingcredible!!!!

Wow, that was real genius! AMD engineers sure
deserve a lot of credit for beating Intel to the 64bit punch.

I am really surprised that no one at Intel thought of this
before the AMD guys did!

 

[Rob Stow]

Wow, those guys at AMD sure are smart.

They created x86-64 before Intel.

Who would have thought to extend the EAX register
from 32 to 64 bits! And then they extended all the
other registers to 64 bits. Infuckingcredible!!!!

Ignoring your sarcasm, the extended registers are only
half the register story for AMD64 - they also doubled the
number of registers.

I think most of the performance gains we are seeing with the
limited amount (so far) of 64-bit optimized software are due
to having more registers than to any other single factor except
the integrated memory controller.

Wow, that was real genius! AMD engineers sure
deserve a lot of credit for beating Intel to the 64bit punch.

You bet they do. Intel really screwed the pooch by not
jumping on the AMD64 bandwagon a few years ago.

And Intel's AMD64 clone looks like it will be handicapped by
1.) Memory latencies because they are still going to rely on
external memory controllers.
2.) Inability to do cheap, fast, and easy SMP because they will
still rely on the chipset to provide the glue logic instead
of building it into the processors like AMD did.

AMD will now, with good justification, be able to say that
they are the "real thing", while Intel's johnny-come-lately
chip is merely a cheap clone that omits a couple of very
important features.

Remember how AMD was mocked when the first started making
clones of Intel chips ? I'll bet you there is a lot of
glee at AMD now that Intel is the one doing the cloning.

I am really surprised that no one at Intel thought of this
before the AMD guys did!

Who says they didn't ? But /if/ they did think of it first,
then the Intel execs really screwed the pooch by not
pursuing the idea.

 

[RusH]

(e-mail address removed) (Black Jack) wrote :

http://www.eetimes.com/semi/news/OEG20040217S0033

"The big databases have already been converted over to Itanium, and
they will be on Itanium forever," Fister added.

Define "forever" heheheheheh.

Pozdrawiam.

 

[Robert Myers]

I think most of the performance gains we are seeing with the
limited amount (so far) of 64-bit optimized software are due
to having more registers than to any other single factor except
the integrated memory controller.

How would you tell the difference?

The across-the-board winner, practically all problems, practically all
coding styles, has almost got to be reduced latency. In the same
category as increasing the size of the cache, only much better,
because you don't have to work so hard to suck stuff into the cache.

More register names. My guess is that the biggest benefit is to
compiler writers and hand coders.

RM

 

[Robert Myers]

(e-mail address removed) (Black Jack) wrote :

http://www.eetimes.com/semi/news/OEG20040217S0033

"The big databases have already been converted over to Itanium, and
they will be on Itanium forever," Fister added.

Define "forever" heheheheheh.

I suppose that Intel has done so well out of just dumb, blind luck,
and that given the same dumb, blind luck, you could do just as well.

Intel has a strategy here. These are not stupid people, and the
strategy may be working better than you think.

Your belief, I suppose, is that Intel management is whistling past the
graveyard. Their strategy was to get big customers locked into an
absolutely unique architecture. To the extent that Intel has
succeeded, those customers are not going back to x86.

Intel's strategy is built more on human psychology than it is on any
principal of engineering. Once people have made a major move, they
have an emotional investment in the decision they've made and will
reconsider only faced with overwhelming evidence.

While Opteron may have tremendous appeal to companies trying to do
things on the cheap, those weren't the customers Intel was after in
the first place, and the fact is that Itanium outperforms the
competition on the applications that matter to those customers.

RM

 

[gaffo]

From what others have written, you should have no problems in Real
Mode (ie actually booting to DOS), or V86 Mode under 386 Mode (ie
running 32 bit Windows).

But apparently in 64 bit extended (long?) mode it drops V86 Mode.

So you can run 32bit and 16bit apps together, or 32bit and 64bit apps
together, but not all three at once.

I think.

I see - so Real Mode is still supported.

thats what i as wondering about - thanks!........................so i
can run Moslo and play Digdug on my new shiny Opteron monster ;-).


if I had one.............

--





RUSSERT: Are you prepared to lose?

BUSH: No, I'm not going to lose.

RUSSERT: If you did, what would you do?

BUSH: Well, I don't plan on losing. I've got a vision for what I want to
do for the country.
See, I know exactly where I want to lead.................And we got
changing times
here in America, too., 2/8/04


"And that's very important for, I think, the people to understand where
I'm coming from,
to know that this is a dangerous world. I wish it wasn't. I'm a war
president.
I make decisions here in the Oval Office in foreign policy matters with
war on my mind.
- pResident of the United State of America, 2/8/04


"Let's talk about the nuclear proposition for a minute. We know that
based on intelligence, that he has been very, very good at hiding
these kinds of efforts. He's had years to get good at it and we know
he has been absolutely devoted to trying to acquire nuclear weapons.
And we believe he has, in fact, reconstituted nuclear weapons."
- Vice President Dick Cheney, on "Meet the Press", 3/16/03


"I don't know anybody that I can think of who has contended that the
Iraqis had nuclear weapons."
- Defense Secretary Donald Rumsfeld, 6/24/03


"I think in this case international law
stood in the way of doing the right thing (invading Iraq)."
- Richard Perle


"He (Saddam Hussein) has not developed any significant capability with
respect to weapons of mass destruction. He is unable to project
conventional power against his neighbours."
- Colin Powell February 24 2001


"We have been successful for the last ten years in keeping
him from developing those weapons and we will continue to be successful."

"He threatens not the United States."

"But I also thought that we had pretty
much removed his stings and frankly for ten years we really have."

'But what is interesting is that with the regime that has been in place
for the past ten years, I think a pretty good job has been done of
keeping him from breaking out and suddenly showing up one day and saying
"look what I got." He hasn't been able to do that.'
- Colin Powell February 26 2001

 

[Felger Carbon]

I think most of the performance gains we are seeing with the
limited amount (so far) of 64-bit optimized software are due
to having more registers than to any other single factor except
the integrated memory controller.

And Intel's AMD64 clone looks like it will be handicapped by
1.) Memory latencies because they are still going to rely on
external memory controllers.

Correct. Don't forget the constipated Intel FSB, esp. in SMP
configurations. This is a problem that the Opteron doesn't have.

2.) Inability to do cheap, fast, and easy SMP because they will
still rely on the chipset to provide the glue logic instead
of building it into the processors like AMD did.

You got it right, Rob. Just thought I'd toss in my two cents. ;-)

 

[Samuel Barber]

Wow, those guys at AMD sure are smart.

They created x86-64 before Intel.
Yep.

Wow, that was real genius! AMD engineers sure
deserve a lot of credit for beating Intel to the 64bit punch.

AMD does deserve credit. And so does Intel for quickly following AMD's
lead.

I am really surprised that no one at Intel thought of this
before the AMD guys did!

Don't be. Intel thought of it, but didn't implement it - probably
because management didn't want to "confuse" the market with an Itanium
competitor.

Sam

 

[lyon_wonder]

I am really surprised that no one at Intel thought of this

before the AMD guys did!

blind loyalty to Itanic and it's IA64 ISA

 

[Tony Hill]

One of the MANY legacy modes available in x86. In particular it was
used to run 16-bit real-mode code while the processor as a whole was
running under a 32-bit protected mode operating system.

When operating in 64-bit long mode, no. If it's operating in 32-bit
protected mode it can still do Virtual-8086 mode. Basically it just
means that if you want to run old 16-bit code while running WinXP
64-bit Edition you need to emulate things.

I'd hate to generalize from one very limited test, but about a
week ago I was unable to boot an Opty dualie or an Athlon64 from
a MS-DOS 3.3 floppy. No problems with MS-DOS 6.22.

That's an unrelated problem, there's no way in hell MS-DOS 3.3 was
booting up in 64-bit Long Mode! :>

 

[Tony Hill]

How would you tell the difference?

The across-the-board winner, practically all problems, practically all
coding styles, has almost got to be reduced latency. In the same
category as increasing the size of the cache, only much better,
because you don't have to work so hard to suck stuff into the cache.

More register names. My guess is that the biggest benefit is to
compiler writers and hand coders.

I think he means 64-bit code on Opteron/Athlon64 vs. 32-bit code on
Opteron/Athlon64, ie identical hardware, only differences is the
software. He's also almost certainly right, since 64-bit should, if
all else were equal, be slower than 32-bit. However because AMD
doubled the number of registers we sometimes (~70-80% of the time from
what I've seen so far) see a performance benefit.

 

[steve harris]

I suppose that Intel has done so well out of just dumb, blind luck,
and that given the same dumb, blind luck, you could do just as well.

Intel has a strategy here. These are not stupid people, and the
strategy may be working better than you think.

Your belief, I suppose, is that Intel management is whistling past the
graveyard. Their strategy was to get big customers locked into an
absolutely unique architecture. To the extent that Intel has
succeeded, those customers are not going back to x86.

Intel's strategy is built more on human psychology than it is on any
principal of engineering. Once people have made a major move, they
have an emotional investment in the decision they've made and will
reconsider only faced with overwhelming evidence.

While Opteron may have tremendous appeal to companies trying to do
things on the cheap, those weren't the customers Intel was after in
the first place, and the fact is that Itanium outperforms the
competition on the applications that matter to those customers.

RM

How about Intel is about making money and Merced is nothing but a money
making proprietary processor that AMD would not be allowed to produce?

Rambus was an Intel opportunity to corner the market. Intel's Rambus
shareholdings came back to bite them in the butt.

Intel could care less what the customer wants IMO.