Sign in with
Sign up | Sign in
Your question

Hammer bandwidth concerns

Tags:
  • CPUs
  • Bandwidth
  • Controller
  • Memory
Last response: in CPUs
Share
April 20, 2002 10:16:29 PM

The Clawhammer's onboard memory controller's going to be limited to DDR333. Pardon me, but don't this give it a serious deficiency when compared to the Pentium 4 B (2700MB/s bandwidth as compared to 4200MB/s)?

Since AMD would need something like DDR533 (which will NEVER happen) to match the P4 B, will it be badly affected by having such little bandwidth?

More about : hammer bandwidth concerns

April 20, 2002 10:28:26 PM

It doesn't need high bandwidth like the P4. It only has 2 more stages in its pipelines, nothin got fret about.
Besides, from what I heard, Hyper Transport is a different method of bandwidth transfering, so it will be something new and tons better.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
April 20, 2002 10:34:33 PM

The p4s extremely long pipeline requires much more bandwith to keep full and processing than the axp or hammers(which will only be 2 stages longer) thus the hammer wont need gobs of bandwidth to keep up with the p4.


Also the sledge hammer will have a dual ddr controler, which provides more bandwith than rdram with less latency.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
Related resources
Can't find your answer ? Ask !
April 20, 2002 11:15:24 PM

Ah, now I see. I suppose with the Hammer, memory transfers would 'hit the ceiling' faster than it would with the P4, but that hardly ever happens outside of video editing and professional OpenGL work (which would be catered for by Sledgehammer).
April 21, 2002 12:27:11 AM

I'm really looking forward to this Dual Channel DDR. Maybe even DDR II sometime soon. Dual Channel will increase bandwidth substantially and I hope it'll be a success.
April 21, 2002 5:24:17 AM

Yeah, what everyone else said.

But what im waiting on is Serial ATA. SATA coupled with hyper-transport is gonna kick a$$.

Benchmarks are like sex, everybody loves doing it, everybody thinks they are good at it.
April 21, 2002 5:35:47 AM

blah....sorry, but i am not too happy with what i have read about serial ata...individual cables per drive...more controllers needed on the motherboard...seems like it might drive the cost up a bit, and make the motherboard need to be bigger, and more cables jumbles in the case...sure they are smaller cables, and it will be fast as hell, but heat man....i need airflow!
but hey, if they can figure out how to make it cheap and able to have one cable for more than one device...then sure.
oh well...oh, and i would hate to start a flame war by this post...lol...but i cant stop it...

-DAvid

-Live, Learn, then build your own computer!-
April 21, 2002 7:08:04 AM

techie!!! so nice to see u again

My tech advice here is not free. Email your credit card detials to mynic@hotmail.com :smile: :wink:
April 21, 2002 9:52:55 AM

I'm starting to think serial ATA is a folk lore.. lol, been hearing random things about it for like 1 and a half and its yet to be released.. heh

Computer: $2000 Internet Access: $40 Registering for forums: Free A good signature: Priceless
April 21, 2002 1:50:32 PM

Same thing as DDR II.
Wish they would hurry their asses doing new things!
We home users want 10000RPM IDE drives, we wanted Dual Channel DDR P4s since the beginning!

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
April 21, 2002 5:01:18 PM

10,000 RPM IDE drives? i'd rather not have the whiny noise that something like that would produce... they have to find someway to make it quiet... AND have fairly high density...
but for serial ata, i *think* that there is (relatively) unlimited cable length, so i think you could route the cables around the edges of your case and thus have MUCH improved airflow...

(bb || !bb) - Shakespeare
April 21, 2002 5:14:47 PM

i also wonder, about the socket setup in the hammer architecture, will they have to change it everytime they "upgrade" the ondie memory controller for more bandwith?

(bb || !bb) - Shakespeare
April 21, 2002 6:17:35 PM

No. Using that logic, you'd need different DIMM sizes for DDR200, DDR266 and DDR333.
April 21, 2002 7:51:11 PM

i guess i never thought of it that way, good point...

(bb || !bb) - Shakespeare
April 21, 2002 7:53:41 PM

Trust me, if AMD is still AMD, then we won't get more socket switching for a while after Hammer. That is the whole goal of saving money on new mobos. But Hammer is a very new method of computing in terms of mobo design and installation, so neither of us can determine if staying with same socket means same mobo can function on new Hammer versions.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
April 22, 2002 12:06:15 AM

As we all know, new advancements in memory technology take place every few months some of these advancements actually make it to production. Look at DDR and how far it has come in a short time. The socket and pins stay the same. Bandwidth should not be an issue.

There are greater problems to overcome, first off is getting microsoft to develop a WHQL x86-64 OS version. There is limitations to the existing 86x Windows that prevent it from ever running on x86-64 (I have posted links to this information). not to mention the time it takes to get the above WHQL version to market. My best guess is 3 years for microsoft to do so. win286~win95~win98, NT4~NT2000~WinXP

There is another problems in the architecture, maybe will be addressed in later revs. Ill give AMD the benifit of the doubt as I do not know more than their engineers in what they want to get to market first. My guess is AMD will get it to market and fix problems later as the OS's are being developed.

Also, Hypertransport is working out great for all those developing it.

You are limited to what your mind can perceive.<P ID="edit"><FONT SIZE=-1><EM>Edited by fugger on 04/21/02 05:07 PM.</EM></FONT></P>
April 22, 2002 12:39:54 AM

Quote:
There are greater problems to overcome, first off is getting microsoft to develop a WHQL x86-64 OS version. There is limitations to the existing 86x Windows that prevent it from ever running on x86-64 (I have posted links to this information). not to mention the time it takes to get the above WHQL version to market. My best guess is 3 years for microsoft to do so. win286~win95~win98, NT4~NT2000~WinXP



Then how do you explain windows 64 for the itanium, you are sorely mistaken.


There are strong signs there ALREADY IS a windows for the hammer, so why dont you take the fud and fff elsewhere :-).



:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 22, 2002 12:43:19 AM

Quote:
My best guess is 3 years for microsoft to do so. win286~win95~win98, NT4~NT2000~WinXP

Windows xp 64 for itanium fits in your schedule how exactly?


If microsoft makes a windows for the tiny number of itaniums sold, there is no logical reason(asides from what your intel fanboy imaginary friend tells you) for m$ not to release a windows xp x86-64.


And your timeframe is laughable especially considering said windows is not the next generation of windows, and its not replacing xp, it, like the ia64 windows, is a niche product for a niche market.


The hammers 32 bit performance should be greater than anything out when its released, and the 64 bit is only sweeter, you constantly forget this when you post your fff's, the hammer does NOT need a windowsx86-64 to succeed, it runs current window's faster than anything.*


(*according to amd which remains to be seen)

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 22, 2002 12:50:57 AM

"Then how do you explain windows 64 for the itanium, you are sorely mistaken."

Microsoft started that version of windows at least 4 years ago (someone find an accurate date?), wtf is wrong with you. do some research. Itanium has been in the planning for a long time and pushed back a few times.

You are limited to what your mind can perceive.
April 22, 2002 12:59:30 AM

Quote:
Microsoft started that version of windows at least 4 years ago (someone find an accurate date?), wtf is wrong with you. do some research. Itanium has been in the planning for a long time and pushed back a few times.


The hammer has been in the works for several years as well, and you seem to think windows wont work in 64 bit, prey tell, where do you get your inside information?


You have 0 clue what your talking about, and you have shown in the past you dont know what the hammer even is, let alone how it works.(remember your dual core one fills with 0s in 32 bit mode bs comments from a month or so ago).


Please dont spread your fff's without linkage to back up your statements, you have 0 credibility and it would save everyone tons of time refuting your crap if you would just be direct and prove what you say.

Once a liar always a liar.



:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 22, 2002 1:16:14 AM

"If microsoft makes a windows for the tiny number of itaniums sold, there is no logical reason(asides from what your intel fanboy imaginary friend tells you) for m$ not to release a windows xp x86-64."

The Itanium was the step for windows to advance to 64bit specificly. Alpha in a sense died as a primary platform. Seems you forgot about the Alpha. As far as tny numbers. thats from the inquirer (lemming bible). the numbers the boasted was somthing like 30 million dollars in the first month of production as they so nicely refered to as a "quarter", when most people understand a "quarter" as 3 months. Itanium is a huge sucess, too bad you are clueless to the uses. your a blind lemming that will blow any point out of whack to suit your needs.

"And your timeframe is laughable especially considering said windows is not the next generation of windows, and its not replacing xp, it, like the ia64 windows, is a niche product for a niche market"

No, its realistic for Microsoft to re-engineer a whole new OS to overcome the current limitations with true x86 version of the OS.

"The hammers 32 bit performance should be greater than anything out when its released, and the 64 bit is only sweeter, you constantly forget this when you post your fff's, the hammer does NOT need a windowsx86-64 to succeed, it runs current window's faster than anything.*"

Purely a guess on your part about the performance as you do not know what will be available when the Hammer is released. and you have no clue how the Hammer will perform period. Pure speculation and wishfull thinking.

Consumers will not pay extra for a x86-64 CPU that is made for server and workstation applications to run just 32bit windows at home. There are no 64bit applications yet for consumers. put 2 plus 2 together and you have no market on release except for Linux servers and development.

From what I have read there is a Linux version that works on the Hammer in theory, and that the software simulator was able to produce a compiler that works.

No one was able to touch the demo at IDF, it was running a canned demo. I have very little faith in canned demo's.

The only Hammer on display was one in the socket on a non operating machine with HSF off.

Windows 32bit on one running MS word/excel loop and the other at a linux command prompt. Like I said no one was allowed to touch them. The cases were closed. It may have been smoke and mirrors from AMD during IDF like some of the people who attended the showing claimed. We know AMD did this during IDF to disrupt any momentum Intel had with developers and to capture new deveopers at the gathering. As most of the developers to show up would never consider AMD in the first place.

I was not allowed to attend the closed doors events, but there a lot of reviews published on this subject on the net.

In the end Microsoft has stated <A HREF="http://msdn.microsoft.com/library/default.asp?url=/libr..." target="_new">x86-64 and the limitations of doing both.</A>

You are limited to what your mind can perceive.
April 22, 2002 1:22:22 AM

Quote:

In the end Microsoft has stated x86-64 and the limitations of doing both.


Hmm, that's the limitation of IA64. The x86-64 should be able to able to jump from 32-bit mode to 64-bit mode and vise versa seemlessly.


:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
April 22, 2002 1:26:10 AM

"The hammer has been in the works for several years as well, and you seem to think windows wont work in 64 bit, prey tell, where do you get your inside information?"

Look at the above link at atteched from Microsoft. ROFL@U

The Hammer will zero extend 32 bit code to 64bits, using software code morphing. These zeros will take up bandwidth on the bus and is a very lame approach to solving the problem. A much better approach was to use SMP with the dual cores while operating @ 32bits right? The Hammer cannot do this.

You forget I work with the <A HREF="http://www.pmc-sierra.com/products/details/rm9000x2/" target="_new">PMC-Sierra</A>. I have good reason to stay current with the Hammer and its development.

Even when I said the Hammer was dual 32bit cores you tried to call me a liar and I had to shut you up. Now that you admit the dual cores is true your still trying to call me a liar, Get a life loser.

You are limited to what your mind can perceive.
April 22, 2002 1:28:21 AM

Read more, this page is reference from x86-64. This is why IA64 windows cannot operate on the Hammer.

stuck? type x86-64 in the search field.

You are limited to what your mind can perceive.

I change AI to IA, my mistake so you dont need to re-quote what I had in order to prove me wrong.<P ID="edit"><FONT SIZE=-1><EM>Edited by fugger on 04/21/02 06:31 PM.</EM></FONT></P>
April 22, 2002 1:32:06 AM

Quote:

Read more, this page is reference from x86-64. This is why AI64 windows cannot operate on the Hammer.

stuck? type x86-64 in the search field.

Lol, that's pathetic. There's x86 and 64 on the page, but not x86-64. You're trying to hard to make a point that doesn't exist.

:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
April 22, 2002 1:33:27 AM

Going from IA64 to x86-64 is not going to work, and going from current Windows XP to x86-64 is not going to work.

The limitation is as explained on that page. reguardless what OS it is for. the limitations are the same currently.

You are limited to what your mind can perceive.
April 22, 2002 1:35:21 AM

Ok, have it your way.

you obviously know better than Micorsoft or me.

Its not what you want to hear so it must be wrong. so go poke your head back under the earth.

You are limited to what your mind can perceive.
April 22, 2002 1:54:23 AM

No, I know a bit about 16-bit to 32-bit thunking under the x86. I can call a 16-bit dll from a 32-bit app, then there's no reason for a 64-bit Hammer which is 100% compatible with x86-32? x86-64 is completely backward compatible, hence the "x86" in the name.

Here's an example written in Borland Delphi 5:

function GetSystemResources(SysResource: Word): Word;
var
ThunkTrash: array[0..$20] of Word;
hInst16: THandle;
GFSR: Pointer;

begin
Result := 0;
ThunkTrash[0] := hInst16;

hInst16 := LoadLibrary16('user.exe');
if hInst16 < 32 then raise Exception.Create('Cannot load USER.EXE');

FreeLibrary16(hInst16);

GFSR := GetProcAddress16(hInst16, 'GetFreeSystemResources');
if GFSR = nil then
raise Exception.Create('Cannot get address of GetFreeSystemResources');

asm
push SysResource
mov edx, GFSR
call QT_Thunk
mov Result, ax
end;

{ end; }
end;


Note this is not compatible with WinNT and up because they have unlimited system resources.

:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
April 22, 2002 3:29:37 PM

Quote:
Purely a guess on your part about the performance as you do not know what will be available when the Hammer is released. and you have no clue how the Hammer will perform period. Pure speculation and wishfull thinking.



Which is why I put the * there, are you blind?

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 22, 2002 3:30:55 PM

Quote:
The only Hammer on display was one in the socket on a non operating machine with HSF off.


The hammer was in the box and running, you should look up your information before you make stuff up. There were 2 hammers running, one was running windows(32) and the other was running linux in dual 64/32 mode.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 22, 2002 3:33:00 PM

LoL fugger, from your link.


Quote:
A 32-bit DLL expects 4K pages and x86-style exception handling. On the Intel Itanium processor, the native page size is 8K and the WOW64 emulator simulates 4K pages with assistance from the kernel memory manager and the Intel Itanium processor. WOW64 implements x86 exception handling as "pure"—no unwinds from x86 code to the Intel Itanium processor and back to x86—because exceptions on an x86 machine do not unwind from user mode to kernel mode and back to user mode.



Those are negatives for dual 32/64 ITANIUM systems, not amdx86-64


run along and get some new info troll.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 22, 2002 3:41:21 PM

I've had enough of Fugger! He thinks he knows more about hardware and software than anyone else on here!

:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
April 22, 2002 3:43:16 PM

Quote:
Even when I said the Hammer was dual 32bit cores you tried to call me a liar and I had to shut you up. Now that you admit the dual cores is true your still trying to call me a liar, Get a life loser.


The hammer is NOT dual 32 bit cores, learn about what your talking about <A HREF="http://www.anandtech.com/cpu/showdoc.html?i=1546&p=5" target="_new">Here</A>.


Quote:
In an interesting but definitely not poorly chosen move, AMD has decided to stick with the K7's 3 ALUs, AGUs and FPUs. Although this may seem far from technical, the justification honestly comes down to "if it ain't broke, don't fix it." We're sure that AMD has done much more extensive profiling on the usage of the Athlon's execution units than we have, but it's safe to assume that the Athlon has no problem of running out of hands to work with.

That's great but we're still back to square one, how does AMD plan to increase IPC on the Hammer?

The answer to that comes in three of the major enhancements over the K7 architecture:

1) integrated memory controller & North Bridge
2) vastly improved branch prediction unit, and
3) what AMD likes to call "large workload TLBs"

There are no cute names for these benefits of the Hammer architecture and we'd have it no other way, so now it's time to dig into what really makes the Hammer special.


So, please run along fugger, you are not educated enough on this subject to make any claims, let alone say amd will fail or is making "mistakes".







You claimed the linux hammer demo was just a command line, looks like your proven a <A HREF="http://www.eweek.com/article/0,3658,s=702&a=23885,00.as..." target="_new">liar</A> once more.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 22, 2002 3:45:08 PM

YOu owned that troll amd_man.


To everyone else, fugger is spreading fud and fff(fugger fabiricated facts) do not believe anything he says without references, he is a known and proven troll and liar.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
April 22, 2002 4:11:39 PM

Quote:

So, please run along fugger, you are not educated enough on this subject to make any claims, let alone say amd will fail or is making "mistakes".

Agreed, the kind of statements he make are so far off it's pathetic. Not knowing is one thing, but pretending that you know it and bashing a company with his fake knowledge is another.

:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
April 22, 2002 5:07:13 PM

"The hammer was in the box and running, you should look up your information before you make stuff up. There were 2 hammers running, one was running windows(32) and the other was running linux in dual 64/32 mode."

I was IDF tard, maybe you should rethink your brilliant statement. I relayed what I saw. GG

"To everyone else, fugger is spreading fud and fff(fugger fabiricated facts) do not believe anything he says without references, he is a known and proven troll and liar."

Total BS. face it you dont like what I say so you attack me perosnally. post some links to disprove or something. dont be such a pussy about it.



You are limited to what your mind can perceive.
April 22, 2002 5:13:04 PM

AMD_MAN so I take it you are no longer neutral.

First off you are as useless a Eden on any subject, not quite as bad as Matisaro.

Oh god I said something bad about your beloved AMD CPU so I must be evil, liar, troll, disllusioned, and know nothing.

Hell I posted positive stuff and you didnt run that into the ground. Go figure you two faced troll.

ROFL. now go bury your head in the dirt.

You are limited to what your mind can perceive.
April 22, 2002 5:14:48 PM

wasn't that link he posted at anandtech good enough to show you there isn't dual cores? that's what it looked like to me. and you're talking about the hammer not running. are we talking about the same event that both tom and anandtech did the articles on, showing the thing running win/linux?

[insert philosophical statement here]
April 22, 2002 5:37:17 PM

HAHAHAHAHAHAHAHA, dude your losing at every turn. WTF you get the brillant idea to post this?

"The hammer is NOT dual 32 bit cores, learn about what your talking about Here."

<A HREF="http://www.x86-64.org/documentation_folder/assembly" target="_new">And the crowd goes nuts!</A>

<A HREF="http://www.amd.com/us-en/assets/content_type/white_pape..." target="_new">x86-64 whitepapers explaining zero extending to second core. WTF are you talking about? do you even know?</A> You will pick anything to attack it.

<A HREF="http://www.planetunreal.com/ictus/3dgn/hardware/article..." target="_new">Q6--- With Sledgehammer, AMD plans to use dual 32-bit cores to speed up 32bit and run 64-bit applications. Intel is using a single 64-bit core, causing the need for emulation of 32-bit software. How do you feel about the 32/64 processor issue?
A: We think our approach to 64-bit computing is a win-win situation. Our approach is similar to what happened between the introduction of the 286 and the 386, when the x86 instruction set went from 16 to 32 bit. We plan on extending the x86 instruction set to 64 bits. The plan is to offer a processor that does not compromise 32 bit application performance to run the small number of applications that actually require 64 bits.</A>

I can dig up a thousand links proving it will have dual 32bit cores.

A Loser is you!



You are limited to what your mind can perceive.
April 22, 2002 5:38:49 PM

It is dual 32bit cores, that is common knowledge.

Not sure what Matisaro has been sniffing but its funny that he made that mistake. He gets lamer as the days go by.

You are limited to what your mind can perceive.
April 22, 2002 6:15:53 PM

k, thanks. but from that amd pdf. is that the current cpus do this now when accessing like 3dnow? is this correct or did i read it wrong? how would this effect hammer if processors now are doing this (on a different level i take it) and isn't hindering their performance?

[insert philosophical statement here]
April 22, 2002 6:29:37 PM

I dont think any processors are doing this now.

Zero extending will hinder to some degree, unknown untill it hits the market. SMP on a chip would have sold me on the Hammer design. But this is not the case.

You are limited to what your mind can perceive.
April 22, 2002 6:51:27 PM

The Hammer does not consist out of a <b>Dual execuation core</b> it merly uses extensions to the GPRs which are 64bit wide however - instructions on those GPRs are done with a <b>single 64 bit execution core (with 64bit ALUs)</b> much as current 32bit x86 processors execute legacy 16bit x86 code <b>as well as 32bit code</b> on the same execution core - as you said - when executing 16bit code on 32 bit GPRs the extra 16bit is being treated as zeros.
thus Hammer extends x86 code to 64 bit in the same way. treating the extra 32 bit as zeros - this shouldn't result in a preformance loss sence a single 64bit ALUs execute on 64bit GPRs in a single clock-cycle.

this is how it goes as far as I know - and does not contredict with any of the info you shown (at least offical one).


This post is best viewed with common sense enabled<P ID="edit"><FONT SIZE=-1><EM>Edited by iib on 04/22/02 09:54 PM.</EM></FONT></P>
April 22, 2002 7:13:37 PM

It is dual 32bit cores on a chip.

<A HREF="http://chip-architect.com/news/2001_10_02_Hammer_microa..." target="_new">Once again another link to more proof of the dual cores</A> Nice pretty pictures. Notice core 0 and core 1... Hmmm. notice AMD white papers speak of feeding a core zero's... Hmmm.

You are limited to what your mind can perceive.
April 22, 2002 7:17:33 PM

Well, from what was indicated on Anandtech, the size of the Hammer's package is about equal to that of the mPGA478 P4. Hammer's die is smaller than Northwood's, but there's no way that two of those would fit in that package. Hyperthreading wouldn't be possible, due to insufficient memory bandwidth (on Clawhammer, at least).
April 22, 2002 7:32:23 PM

this article was bassed on patenets issued by AMD and does not reflect the final Hammer desing - the writer Hans de Vries.

when AMD it self presented the Hammer desing it turned out very diffrent from what the writer Hans de Vries derived from AMDs patants - he mentions so on a <A HREF="http://chip-architect.com/news/2001_10_18_AMD_presents_..." target="_new">piece</A> he did some time after.
here is a qoute:
[qoute]
The Hammer is not based on the set of patents that we discussed a few weeks ago. These patents belong to an earlier K8 design which has been either abandoned or is possibly continued in one way or another for a future design. The currently presented Hammer resembles much more a highly tuned version of the Athlon Palomino.
[/qoute]

the parts AMD tuned to make the Hammer are well explained in the anandtech article. those voil down do the on-die memory controler, enhanced branch predictor, improved TLBs.

the execution core has remained the same as athlon except ofcource from its capbilty to handle 64 bit instructions.

the block diagrem of the Hammer which clearly shows one executiON unit is avalible at the AMD Hammer white paper PDF which you can download <A HREF="http://www.amd.com/us-en/assets/content_type/white_pape..." target="_new">here</A>.
you can clearly see a single core - fetching - decoding - executing - and despeching instruction bassed on the x86-64 ISA.


This post is best viewed with common sense enabled
April 22, 2002 7:33:26 PM

Well they are doing it, hard to believe. The heat spreader makes the P4 core seem larger than it really is. The Hammers are slightly larger than the P4 478.

<A HREF="http://images.anandtech.com/reviews/shows/2002/SpringID... 2/hammer/clawhammer.jpg" target="_new">Clawhammer 754Pins</A>

<A HREF="http://images.anandtech.com/reviews/shows/2002/SpringID... 2/hammer/sledgehammer2.jpg" target="_new">Sledgehammer over 900Pins</A>

You are limited to what your mind can perceive.
April 22, 2002 7:43:06 PM

"the execution core has remained the same as athlon except ofcource from its capbilty to handle 64 bit instructions."

To my understanding the cores are based off the K7 core.

Check out page 11 of whitepapers.

Also check page 7, 32bit operand size in 64bit mode.

"A quantity on which a mathematical or logical operation is performed."

You are limited to what your mind can perceive.
April 22, 2002 7:47:22 PM

First off, you're the only one who's agreeing with yourself. Go put a topic asking WHO here likes you or agrees with you, and we'll see.
Second of all, you have reached the limits, if AMD_Man has changed his sides, he of all "THGC angels", then it means you've crossed the line man. You better not call him a troll, he has NOT done anything trolling, merely defending a position with neutral comments. You OTOH treat AMD like dirt. Funny how you trolls think everyone favors you.

--
Thunderbirds in wintertime, Northwoods in summertime! :lol: 
April 22, 2002 7:48:37 PM

Fugger, you are not going to improve your credibility through your false information. What's the point of bashing the Hammer and AMD? Hmm, I can confirm your so-called software incompatibility claim is pure falsehood due to my own knowledge of programmer. However, my knowledge of hardware is more limited but I can also be fairly sure what you're saying is completely unintelligible. The number of pins has nothing to do with the core size. How can AMD fit two cores on a single chip when the total die size is so small? AMD hasn't removed anything. Dual-cores just don't make sense...the overhead of emulating 64-bit processing through 2 processors would make 64-bit no faster than 32-bits, obviously AMD engineers aren't stupid. You can find all the links you want Fugger, but it all seems like speculation to me! I find all this idle speculation and bashing distasteful. If you don't have a degree in computer engineering, then I suggest you stop trying to be something you're not Fugger.

:wink: <b><i>"A penny saved is a penny earned!"</i></b> :wink:
      • 1 / 3
      • 2
      • 3
      • Newest
!