Looking back, IPC, Intel Pentium 4 vs AMD Athlon XP

humbug · 23 Jul 2018 at 01:48

Most of know about these two chips and their story, but just how much faster than Intel was AMD's god chip.

Have a guess....... nope, it was a lot more than that.

Part 2, Intel's Core 2, Intel takes the lead back.

MasterOC · 23 Jul 2018 at 09:01

Had a barton core xp chip and a 5000+ x2 , still have them in boards mounted on the wall now infact . Pentium 4 was renowned trash tbf but still seemed to smash AMD when it came to sales in pre builds , schools used p4 , sales enviroments and even in professional spaces they all seemed to have contracts with intel. Pretty sure they got fines for these shenanigans . Shame really AMD had the better products .

d_brennen · 23 Jul 2018 at 09:05

Had a Duron 700 clocked to 1.1GHz then an Athlon XP 3200+, somewhat later an Athlon 64. Epic chips of their time, back when it was worth upgrading

Armageus · 23 Jul 2018 at 09:12

Struggling with the comparisons - not exactly like for like in terms of what was available at the time

Athlon X2 6000+ (3Ghz) was Feb 2007 release date
Athlon X2 5000+ was Sept 2006

Pentium D 930 was Jan 2006
Pentium D 830 was May 2005

humbug · 23 Jul 2018 at 09:16

Armageus said:
Struggling with the comparisons - not exactly like for like in terms of what was available at the time

Athlon X2 6000+ (3Ghz) was Feb 2007 release date
Athlon X2 5000+ was Sept 2006

Pentium D 930 was Jan 2006
Pentium D 830 was May 2005

He did explain and show the 930 and 830 are the same chip, as are the 5000+ and the 6000+

humbug · 23 Jul 2018 at 09:26

humbug said:
He did explain and show the 930 and 830 are the same chip, as are the 5000+ and the 6000+

Actually i will agree he has over complicated it by testing how much the difference in the amount of cache levels makes to performance, i mean really this is unrelated and confuses it, he is one of those tech reviewers who completely over thinks things and then spends 20 minutes explaining his unnecessary and needlessly complicated reasoning.

He should have just picked the D 930 and the 5000+ and be done with it in 4 minutes instead of 20.

eddiew · 23 Jul 2018 at 10:30

Man, I remember my first experience with AMD's "venice" architecture around that time... forgot to plug the CPU fan in while installing Windows, but it still didn't get hot to the touch. Try that with a P4!

Google tells me it was 130nm with 1.4vcore... what the heck happened with Ryzen's 12nm and 1.5v spikes?

(I like Ryzen, but those voltages bother me.)

Armageus · 23 Jul 2018 at 10:47

humbug said:
He should have just picked the D 930 and the 5000+ and be done with it in 4 minutes instead of 20.

Yep or the comparable 2005 Athlon 64 X2 4400+ (Toledo) vs the Pentium D830.

At this point AMD still had the performance advantage, however by July 2006 the first Core2Duo chips (E6300/E6400/E6600/E6700) were released (which are the chips the X2 5000+ should be compared against), and by Jan 2007 the Core2Quad Q6600 was released which would have competed with the X2 6000+

humbug · 23 Jul 2018 at 14:28

Armageus said:
Yep or the comparable 2005 Athlon 64 X2 4400+ (Toledo) vs the Pentium D830.

At this point AMD still had the performance advantage, however by July 2006 the first Core2Duo chips (E6300/E6400/E6600/E6700) were released (which are the chips the X2 5000+ should be compared against), and by Jan 2007 the Core2Quad Q6600 was released which would have competed with the X2 6000+

I think that's what's coming next, he's making a series of these.

Zeed · 23 Jul 2018 at 17:37

Only proof tjat AMD is crap at selling any product.
Now they got another and possibly LAST chance to outplay Intel that messed up with 10nm process prediction.

3000 series will be Epic win if they dont mess up. All it needs is mainstream chip that will do 4.4-4.5 on all 8 cores and they have intel by balls in gaming and productivity.

If they go 12 cpre mainstream sacryficing core clock tjat is ONLY thing that zen+ is lacking. Will be like shoting themselfes in foot.

Thwy got mpst on theirs aide many would jump from intel if zen2 had more Mhz.

Zeed · 23 Jul 2018 at 17:39

humbug said:
I think that's what's coming next, he's making a series of these.

Wounder why people waste theirs time on crearing useless content pike that?? Its like rewrighting history books just to rewright them..

humbug · 23 Jul 2018 at 17:45

I don't think we ever really understood the real IPC difference between these CPU's ^^^^

Zeed said:
Only proof tjat AMD is crap at selling any product.
Now they got another and possibly LAST chance to outplay Intel that messed up with 10nm process prediction.

3000 series will be Epic win if they dont mess up. All it needs is mainstream chip that will do 4.4-4.5 on all 8 cores and they have intel by balls in gaming and productivity.

If they go 12 cpre mainstream sacryficing core clock tjat is ONLY thing that zen+ is lacking. Will be like shoting themselfes in foot.

Thwy got mpst on theirs aide many would jump from intel if zen2 had more Mhz.

You are right AMD have never been good at marketing, tho i would argue they are getting better these days, still in regards to this there is so much more to it, to stop AMD making any headway with their architecture lead and benefiting from it Intel not only gave away thier CPU's to the likes of Dell but on top of that also paid them $800m a year not to use AMD's CPU's at all, thats the most extreme example but Intel were doing this with all OEM's that matter.

Now when you have your vastly larger rival with endlessly deep pockets giving away their CPU's and paying 'where it matters' not to use yours you're not selling enough of your own CPU's or making any money.

I think this did far more damage to AMD than most people realise, i do think its a direct reason as to where we are at now.

humbug · 23 Jul 2018 at 17:59

@Zeed please watch this to understand the history

Gerard · 23 Jul 2018 at 18:10

Intel's P4 chips had massively long pipelines for the p4, wasn't it like 20+ stage compared to much lower for the amd equivalents? I remember apple doing a comparison vs the g4 cpu, they seemed to think they coined the term "megahertz myth"

Zeed · 23 Jul 2018 at 19:56

humbug said:
@Zeed please watch this to understand the history

I know and I like how Intel and NV thinks Thats why they made LOADS OF MONEY. If You got option You destroy competition before they are threat.

In capitalist world You dont make money from being nice. From being nice you get screwed over by competition like AMD been by Intel and NV.
If amd wont learn play the dirty game they always will be behind on money

kitfit1 · 23 Jul 2018 at 20:20

Zeed said:
I know and I like how Intel and NV thinks Thats why they made LOADS OF MONEY. If You got option You destroy competition before they are threat.

In capitalist world You dont make money from being nice. From being nice you get screwed over by competition like AMD been by Intel and NV.
If amd wont learn play the dirty game they always will be behind on money

+1 absolutely.
There is another thread going on at the moment where Intel and AMD IPC is being called out, revolving around Cinebench single Core runs. Thing is though, my 2700x runs nicely at 184 at only 4.25Ghz and does so time after time. Peeps that are running a 4790K at 4.6Ghz with a score of 190 seem to think that it's better than a 2700x. What complete rubbish. I know that if my 2700x could run at 4.6Ghz, it would blow a 4790K into the weeds. Thing is though, i firmly believe that most other peeps think the same thing. All this crap talk of single thread one upmanship is just that............................it's crap and not worth even thinking about.

Zeed · 23 Jul 2018 at 20:27

kitfit1 said:
+1 absolutely.
There is another thread going on at the moment where Intel and AMD IPC is being called out, revolving around Cinebench single Core runs. Thing is though, my 2700x runs nicely at 184 at only 4.25Ghz and does so time after time. Peeps that are running a 4790K at 4.6Ghz with a score of 190 seem to think that it's better than a 2700x. What complete rubbish. I know that if my 2700x could run at 4.6Ghz, it would blow a 4790K into the weeds. Thing is though, i firmly believe that most other peeps think the same thing. All this crap talk of single thread one upmanship is just that............................it's crap and not worth even thinking about.

Like jumping from 5820k to 1700x i noticed system and games run Smoother even cause AVERAGE fps ware about same Minimal frame drops ware higher.
2700x is an FANTASTIC CPU !!!
WoW runs better than even and thats in 3D !!!!

Till AMD gets CB ST score passing Intell they will always be seen as lesser gaming product and tahts a fact.

mmj_uk · 24 Jul 2018 at 14:27

The problem AMD had back in the early days of P4 was that 1.8A Northwood overclocked to 2.7ghz+ and the 2.4A to 3.3ghz+ and AMD just couldn't compete, they were the cheap alternative. They also relied a lot on crap VIA chipsets which I imagine is what put industry off (on top of anti-competitive behaviour by Intel). Athlon64 and x2 is about the time their CPU's were really attractive to top end users again.

Gerard · 24 Jul 2018 at 14:43

mmj_uk said:
The problem AMD had back in the early days of P4 was that 1.8A Northwood overclocked to 2.7ghz+ and the 2.4A to 3.3ghz+ and AMD just couldn't compete, they were the cheap alternative. They also relied a lot on crap VIA chipsets. Athlon64 and x2 is about the time their CPU's were really attractive to top end users again.

The pentium 4's were pretty bad chips on the whole, a 20 stage pipeline increasing to 31 stage, they overclocked well but they were like furnaces to cool, especially prescott. It was a bad series of chips for intel, the RD-ram requirement on the first iteration that went the way of the dodo. That few years of netburst was like amd's few years of the bulldozer core, they were locked into it for several years as they had nothing else close to ready, mainly in the design stages of anything new.

I had a prescott 3.4 running at 4.00 with a corsair hydrocool external cooler to keep it tamed, which it barely did.

smsmasters · 24 Jul 2018 at 15:00

I had an AMD K6-2 333 MHz in 1998 and it was generally better than Intel at the time.