ecosoft

Distinguished
Jun 23, 2004
137
0
18,680
Dudes! All of 'ya! Is everyone posting to this Forum a childish ass-ole??

Who REALLY gives red rat's ass about "Fanboyism"? CPU fanboyism is nothing, NO THING, different than the old redneck automotive Ford vs Chevy BS, except the terms argued now revolve around electronics devoted to personal computers. BFD!

Try examining the controversy from the next level up ... call it the business viewpoint, if ya' like, but whatever it's valid because history has demonstrated this viewpoint ... all businesses that have competition have only one of two choices for continued survival. Both choices involve the size of their gonads ... BALLS!

Intel, being the first semiconductor mfg'r to consolidate a large chunk of the industry ... anyone remember Fairchild Semi, Digital Equipment Corp., Data General, Wang, etc. who, previous to Intel's dominance, all made their own CPU's. As a result of outstandingly aggressive expansion minded (monopoly?) management Intel grew to 20+ mfg plants around the world and their internal organisation grew likewise (many separate power centers) so that EVERY decision was an agonizing process (SLOW as Diahera flowing in the Winter!). Intel had no effective competition even with the advent of AMD and their K5-6 x86 CPU's. Intel was literally fat, dumb and happy and remained that way only until recently ... less than a year ago!!

Intel looked down it's nose at the initial AMD K7's, and from their point-of-view (PoV) it was logical, so they stayed their x86 architectural course. It wasn't until AMD popped out the K8 series that Intel slowly began to "notice" and even then, they stayed their architectural designs by attempting to move the market with higher and higher clock ratings, but at the expense of superhuge pipelines and humongo thermal envelopes. Intel could "afford" to be complacent because they have 1.) hugely deep pockets, 2.) huge business IT mindshare ... just like in the mainframe days when it was always said "ya' can't ever go wrong choosing IBM" (anyone remember Borroughs, Honeywell, RCA, Sperry-Rand, Control Data, etc.) and 3.) aggressive marketing i.e. "buy our sh-t and no one else's and we will subsidize your advertising with $$$ and a "can't get it out of your mind Intel Inside" tone jingle. It worked!

The hook for their strategy working was the economic side, which dictated "keep doing what we'er doing" 'cause the other guys, AMD, didn't have the financial balls to pose a real challenge. HOWEVER, the other side of having balls is "where do 'ya choose to be ballsy?"

AMD took the only feasible option open to them ... CPU architecture. Now then, architecture includes more than how ya' layout the specific functional circuitry (ALU's, FPU's, etc., etc.) but also how ya' build 'em. AMD got in bed with IBM for both mfg capacity AND expertise... SOI, etc. (Do you know IBM holds more tech patents than anyone on the planet?)

So, we have the "biggest" CPU mfg'r now fending off a shrimp that has a superior architecture and you Fanboys are arguing who's going to win 'da game. In the end we consumers win!

In the short term I predict Intel will not substantially alter their architecture because of numerous internal "vested interests". Rather, Intel is using it's bucks-up position to drive their production towards the 65, 45 and 22nm nodes because all else being equal ya' get less heat, more curcuitry per square inch, more chips per wafer, etc., etc. AMD only recently began employing 300mm wafers, and they still have yet to reach volume 65nm manufacture. Even so, AMD Opterons are kickin' butt in the server market because of their superior architecture, thus lower total cost of ownership.

As for Intel's boast of 4-way, 8-way, gazillion-way CPU's in the future BFD! Until software ... and that means MORE than games (how many businesses run game software in the office?) can seamlessly accomodate multi-thread execution, the number of cores per chip is a genuine "so F...'in what?" for the vast majority of end-users! :D
 

Caboose-1

Distinguished
Mar 5, 2006
1,864
0
19,780
Dudes! All of 'ya! Is everyone posting to this Forum a childish ass-ole??

Who REALLY gives red rat's ass about "Fanboyism"? CPU fanboyism is nothing, NO THING, different than the old redneck automotive Ford vs Chevy BS, except the terms argued now revolve around electronics devoted to personal computers. BFD!

Try examining the controversy from the next level up ... call it the business viewpoint, if ya' like, but whatever it's valid because history has demonstrated this viewpoint ... all businesses that have competition have only one of two choices for continued survival. Both choices involve the size of their gonads ... BALLS!

Intel, being the first semiconductor mfg'r to consolidate a large chunk of the industry ... anyone remember Fairchild Semi, Digital Equipment Corp., Data General, Wang, etc. who, previous to Intel's dominance, all made their own CPU's. As a result of outstandingly aggressive expansion minded (monopoly?) management Intel grew to 20+ mfg plants around the world and their internal organisation grew likewise (many separate power centers) so that EVERY decision was an agonizing process (SLOW as Diahera flowing in the Winter!). Intel had no effective competition even with the advent of AMD and their K5-6 x86 CPU's. Intel was literally fat, dumb and happy and remained that way only until recently ... less than a year ago!!

Intel looked down it's nose at the initial AMD K7's, and from their point-of-view (PoV) it was logical, so they stayed their x86 architectural course. It wasn't until AMD popped out the K8 series that Intel slowly began to "notice" and even then, they stayed their architectural designs by attempting to move the market with higher and higher clock ratings, but at the expense of superhuge pipelines and humongo thermal envelopes. Intel could "afford" to be complacent because they have 1.) hugely deep pockets, 2.) huge business IT mindshare ... just like in the mainframe days when it was always said "ya' can't ever go wrong choosing IBM" (anyone remember Borroughs, Honeywell, RCA, Sperry-Rand, Control Data, etc.) and 3.) aggressive marketing i.e. "buy our sh-t and no one else's and we will subsidize your advertising with $$$ and a "can't get it out of your mind Intel Inside" tone jingle. It worked!

The hook for their strategy working was the economic side, which dictated "keep doing what we'er doing" 'cause the other guys, AMD, didn't have the financial balls to pose a real challenge. HOWEVER, the other side of having balls is "where do 'ya choose to be ballsy?"

AMD took the only feasible option open to them ... CPU architecture. Now then, architecture includes more than how ya' layout the specific functional circuitry (ALU's, FPU's, etc., etc.) but also how ya' build 'em. AMD got in bed with IBM for both mfg capacity AND expertise... SOI, etc. (Do you know IBM holds more tech patents than anyone on the planet?)

So, we have the "biggest" CPU mfg'r now fending off a shrimp that has a superior architecture and you Fanboys are arguing who's going to win 'da game. In the end we consumers win!

In the short term I predict Intel will not substantially alter their architecture because of numerous internal "vested interests". Rather, Intel is using it's bucks-up position to drive their production towards the 65, 45 and 22nm nodes because all else being equal ya' get less heat, more curcuitry per square inch, more chips per wafer, etc., etc. AMD only recently began employing 300mm wafers, and they still have yet to reach volume 65nm manufacture. Even so, AMD Opterons are kickin' butt in the server market because of their superior architecture, thus lower total cost of ownership.

As for Intel's boast of 4-way, 8-way, gazillion-way CPU's in the future BFD! Until software ... and that means MORE than games (how many businesses run game software in the office?) can seamlessly accomodate multi-thread execution, the number of cores per chip is a genuine "so F...'in what?" for the vast majority of end-users! :D
I wonder how many times something like this has been posted. Lets see.....1..2..3
 

mpjesse

Splendid
when your signature comes to mind, the following even comes to mind:

(for howard stern fans only)

Jeff "The Drunk" falling down the stairs.

All ya'll stern fans know what i'm talkin' about. ;-)
 

Caboose-1

Distinguished
Mar 5, 2006
1,864
0
19,780
when your signature comes to mind, the following even comes to mind:

(for howard stern fans only)

Jeff "The Drunk" falling down the stairs.

All ya'll stern fans know what i'm talkin' about. ;-)
Hell yeah Howard Stern is the sh!t.
 

Zyklus

Distinguished
Feb 27, 2006
6
0
18,510
There are two kind of people I don't like in this world, people who are intolerant of other people's cultures, and the Dutch!

Thank god the rest of the world can't stand those dickhead Americans either..
 

Zyklus

Distinguished
Feb 27, 2006
6
0
18,510
haha like anyone in america cares what some scrawny third-world asshat thinks. shouldnt you be out trying to pimp your sister for money?

Well, I'm actually European, but with your great sense of American superiority I guess you would consider that third world as well.. :roll:
 

TabrisDarkPeace

Distinguished
Jan 11, 2006
1,378
0
19,280
The K5 and K6-2 at similar price points (TCO) and even at similar clock speeds in 3DNow! optimized applications (eg: Quake 2) had their asses handed to them once the Celeron was given 128 KB of integrated L2 cache.

At equal clock speeds the Celeron was a good 25 - 40% better performing than the K6-2, K6-III and K6-2+, each with large integrated caches, the later two with L2 cache integrated.

I think you are 'remembering' history the way you wanted it to happen instead of the way it actually happened.

Go ahead, look up some Quake II (or any other game of the time) benchmarks and compare the Pentium II / Celeron (pre SSE days) with the K6-2 which had the combined 3DNow! and MMX feature set.

The Celeron (Mendocino core, the MMX only one before the Celeron 533A, at only 250nm) was far more cost effective than the K6-2 at the time.

The K5 and K6, and K6-2 where quite attractive for a variety of reasons though. eg: The K6-2 was a very nice upgrade to Socket Super 7 systems, and saved purchasing a new mainboard, etc

Intel SSE and AMD 3DNow! actually share many common instructions, they are at least 90% identical from a code perspective. AMD and Intel share quite a lot when it comes to microprocessor architecture.

The Pentium 4 PreScott was 'less desirable' but other than that they compete with, and complement each other in a way that benefits us all.

I am in aggreement with you on many of your points.

Started a new thread to get people educated on the basics again [sigh]:
http://forumz.tomshardware.com/hardware/modules.php?name=Forums&file=viewtopic&t=177791
 

TabrisDarkPeace

Distinguished
Jan 11, 2006
1,378
0
19,280
True, but what would Intel 'share' with AMD to use AMD64 under another name (EM64T) ?.

Doesn't anyone else find it odd AMD had 3DNow! (floating) and basic MMX, while Intel had MMX (integer), then SSE (just 3DNow! under another name).

In so many ways they are still the one old and true 'INTegrated ELectronics'. ala: INT-EL. Which is why the 'e' was subcript in their old logo. As far as I am concerned they are still the same company, just based and run in different countries.

If software looked at the supported instructions & feature flags, instead of looking for the 'Genuine Intel' string even less people would notice the difference.