Intel 32 nm Clarkdale Chip Brought Forward to Q4 2009

[tastegw]

i can remember tv ads for intel from way back, but i have never seen a AMD ad.

imo this is why intel is currently ahead, alot of people have never heard of AMD before.

 

[Darren]

i can remember tv ads for intel from way back, but i have never seen a AMD ad.

imo this is why intel is currently ahead, alot of people have never heard of AMD before.

I'm not sure what you mean by ahead, but ahead financially Intel's got a bigger marketing budget for adverts and probably lots of sponsors too. Ahead architecturally AMD have dominated previously for a long time despite AMD being almost known out of the enthusiast community.

 

[tastegw]

market share is what i meant.

 

[Assimilator]

Remember there was the Athlon XP vs the Pentium 4, AMD had the better architecture.

There was the Athlon MP vs the P4 Xeon, AMD had the better architecture.

There was the AMD Duron vs the Celeron, AMD had the better architecture.

There was the Athlon FX vs the Pentium 4 Extreme Edition, AMD had the better architecture.

There was the Celeron D vs Sempron, AMD had the better architecture.

There was the Athlon 64 vs the Pentium D, AMD had the better architecture

There was the Athlon X2 64 vs the Pentium D, AMD had the better architecture.

Wrong.

Between the P3 and the Core 2 line, AMD's processor architecture was more efficient, clock-for-clock, than Intel's. That does not necessarily mean that AMD's architecture was better.

In fact, I would argue that Intel's architecture won out in the end because they took elements from the P3 and P4D to create the monster that is Core 2; contrast that with AMD, who are still using the Athlon64 architecture and failing to produce chips that can compete with Intel's best.

 

[FordGT90Concept]

I'd almost say it was more of the reason why AMD's lead vanished, since the majority of computer users are buying for value, not speed. Even though AMD had chips that were a better value, you had to buy Intel, since hardly anyone was selling pre-built computers with AMD chips.

In USA, only Dell didn't offer AMD chips. Dell started offering them not long after Core 2 Duo came out because of popular demand (low price).

Intel wasn't found guilty in the USA.


Pentium 4 had a more facinating architecture than Athlon 64 featuring long pipes and a brand new technology, Hyper-Threading. Yes, it couldn't beat Athlon 64 but there is very little changes in AMD processors since the K6. AMD's offerings, therefore, aren't as interesting to disect.


Core 2 is a direct decendant from Core, Pentium M, and Pentium III. Little came from Pentium 4 except the process (65nm).

 

[Darren]

Wrong.

Between the P3 and the Core 2 line, AMD's processor architecture was more efficient, clock-for-clock, than Intel's. That does not necessarily mean that AMD's architecture was better.

If AMD's processors prior to the Core 2 line were faster clock for clock and were priced cheaper than the competition (Intel) that is the definition of a better architecture and better competitive pricing as far as I'm concerned. I do not believe anyone can dispute that.

In fact, I would argue that Intel's architecture won out in the end because they took elements from the P3 and P4D to create the monster that is Core 2; contrast that with AMD, who are still using the Athlon64 architecture and failing to produce chips that can compete with Intel's best.

Perhaps the Core 2 being derived from the P4D and P3 was Intel's saviour but it has little to do with anything I said in post #25 or from what you quoted from me. You can not claim that I'm wrong when I didn't disagree with the P3 architecture being exploited or disagree with the Core 2 being better than the Athlon 64 architecture. I was just merely pointed out 7 events from the top of my head where AMD dominated the architectures in the past. But I give Intel the Core 2 and i7 range, 2/7 is not bad.

 

[FordGT90Concept]

Pentium 4, Xeon, Pentium 4 EE, Pentium 4 D vs Athlon XP, Athlon MP, Athlon 64 FX, Athlon 64 X2: Intel was better. Pentium 4 (and sons) was much better because of Hyper-Threading.

Duron, Semperon vs Celeron: both are crap.


Just because one outperforms the other doesn't necessarily mean it is a better architecture. I consider NetBurst the best architecture out there because it represented a paradigm shift. It attempted to rewrite how processors are designed and the first attempt didn't work out so great. They fixed the bugs in Nehalem and now NetBurst has returned with a vengence. Eight logical cores for the price of four physical cores; Hyper-Threading has come a long way.

 

[Darren]

FordGT90Concept,

The Semprons were good, used to beat out the low end Pentium 4s which cost like 75% more, with a small overclock it was competing with the high end Pentium 4s. From an engineering stand point you are correct. Its like the argument "PS3s cell processor vs 360s IMB Tri-core". PS3 fan boys can boast about its architecture superiority and pretend to be engineers all they want, just like Intel fan boys that defend Intel’s slow ass Pentium 4s and Ds (back before the core 2 duos) when as customers all that matters is performance not theoretical performance, actual performance relative to price.

Edit:


My post in #25 was ambiguous but the message was easy to convey AMD had faster processors at numerous points in time, one can argue theoretically that AMDs had faster processors despite having a slower architecture. But the post was suppose to interpreted as if we are basing architecture superiority to actual performance opposed to theoretical.

Edit 2:

A 1.4 GHz Pentium 4 is faster than a 2.4 GHz Semperon. That's a generational gap even. Semperons and Celerons both have butchered caches and because of that, it takes them a lot longer (more clocks) to do everything.


Customers only care about getting something that works. Performance is rarely a concern--cost is. $400-600 PCs are still the best selling segment.

It depends what generation of sempron and what generation of Pentium 4 we are talking about. Remember the sempron spanned accross socket A and 754!

 

[FordGT90Concept]

The Semprons were good, used to beat out the low end Pentium 4s which cost like 75% more, with a small overclock it was competing with the high end Pentium 4s. From an engineering stand point you are correct. Its like the argument "PS3s cell processor vs 360s IMB Tri-core". PS3 fan boys can boast about its architecture superiority and pretend to be engineers all they want, just like Intel fan boys that defend Intel’s slow ass Pentium 4s and Ds (back before the core 2 duos) when as customers all that matters is performance not theoretical performance, actual performance relative to price.

A 1.4 GHz Pentium 4 is faster than a 2.4 GHz Semperon. That's a generational gap even. Semperons and Celerons both have butchered caches and because of that, it takes them a lot longer (more clocks) to do everything.


Customers only care about getting something that works. Performance is rarely a concern--cost is. $400-600 PCs are still the best selling segment.

My post in #25 was ambiguous but the message was easy to convey AMD had faster processors at numerous points in time, one can argue theoretically that AMDs had faster processors despite having a slower architecture. But the post was suppose to interpreted as if we are basing architecture superiority to actual performance opposed to theoretical.

Faster does not necessarily mean a better architecture. I define architecture supremecy by innovation. In your example of PS3 vs Xbox 360, I'd say it is a tie. PS3 has the Cell Broadband Engine which uses SPEs while Xbox360's tri-core processor can handle two threads per core. They are both fairly innovative in their own regards. Performance is based largely on clockspeed. In examining an architecture, I ignore the clockspeed nullifying performance.

It depends what generation of sempron and what generation of Pentium 4 we are talking about. Remember the sempron spanned accross socket A and 754!

Socket 423 Willamette vs Socket 754 Palermo

I used both. Willamette flew on XP and Palermo could barely boot. The story is the same for Celerons. It takes at least twice as long to do everything on a Celeron/Sempron everything else being equal.

 

[Darren]

I agree with you, innovation is just as important as performance, however in the previous posts everyone seems to view sheer performance as the measuring stick and hence why I disagreed that if performance is the only indicator than AMD have been doing well in prior years. Personally I like prices that are relative to performance, innovation is good too but if the processor has 45 GBs of cache and 100 cores but costs £10,000 I'm still not buying it. We are going to have to agree to disagree for the majority of this architectural debate as I'm sure that we are not making much progress.

As for the Sempron I was reading a classic Xbit Labs review of the Sempron 2600 @ 2.5 GHz destroying a Pentium 4 E @ 3.4 GHz. Interesting read.
Here

To be fair the slighty OC'd sempron destroyed even the high end Athlons. Semprons were absolute beasts and workhorses. Ultra cheap too.

 

[Laurijan]

I agree with you, innovation is just as important as performance, however in the previous posts everyone seems to view sheer performance as the measuring stick and hence why I disagreed that if performance is the only indicator than AMD have been doing well in prior years. Personally I like prices that are relative to performance, innovation is good too but if the processor has 45 GBs of cache and 100 cores but costs £10,000 I'm still not buying it. We are going to have to agree to disagree for the majority of this architectural debate as I'm sure that we are not making much progress.

As for the Sempron I was reading a classic Xbit Labs review of the Sempron 2600 @ 2.5 GHz destroying a Pentium 4 E @ 3.4 GHz. Interesting read.
Here

To be fair the slighty OC'd sempron destroyed even the high end Athlons. Semprons were absolute beasts and workhorses. Ultra cheap too.

I know.. I had a 745 Sempron 3000+.. oced to well over 3GHz once.. it could motor any game i threw at it (Doom3, Battlefield Vietnam, BFII)..
Heck it still motors most of the games.. i sold it to a friend with my mobo and he uses that setup till he has some cash to update it and he plays games like Fallout 3, The Witcher and NHL 2009 with some drawbacks in eyecandy

 

[cray86]

Thats small 32nm! I wonder where the limit of silicon is

Edit: So do anyone know where the limit it?

http://en.wikipedia.org/wiki/11_nanometer

 

[Laurijan]

http://en.wikipedia.org/wiki/11_nanometer

Cool find thx!

 

[FordGT90Concept]

To be fair the slighty OC'd sempron destroyed even the high end Athlons. Semprons were absolute beasts and workhorses. Ultra cheap too.

I wouldn't call a 36% overclock "slight." Processors that have less cache tend to be able to run higher clocks at lower voltages. That doesn't mean they are getting more work done though. The 4000+ was clocked at 2.4 GHz and the FX-55 was at 2.6 GHz. I don't doubt that either of those processors would hand it to a 2.5 GHz Sempron.

 

[Guru Janitor]

AMD have been behind for around 3 years. But how long were Intel behind before that? longer than 3 years I would presume



Remember there was the Athlon XP vs the Pentium 4, AMD had the better architecture.

There was the Athlon MP vs the P4 Xeon, AMD had the better architecture.

There was the AMD Duron vs the Celeron, AMD had the better architecture.

There was the Athlon FX vs the Pentium 4 Extreme Edition, AMD had the better architecture

There was the Celeron D vs empron, AMD had the better architecture.

There was the Athlon 64 vs the Pentium D, AMD had the better architecture

There was the Athlon X2 64 vs the Pentium D, AMD had the better architecture.


My point AMD has been dominating the performance for a long time, Intel’s 3 year reign has been very short in comparison to AMDs. Even with Intel’s slower architectures Intel still overpriced their CPUs but some how managed to maintain a larger market share than AMD.
.

Thats wayyyyy offff. Intel held the crown since the early 70's. Then they messed up with P4, and AMD had the crown for a few years. Now after Core 2, and now with Core i* Intel is once again holding the crown. AMD has had to deal with the back seat a lot longer then Intel had to in the past few years....

That being said, I don't find it surprsing Intel is slated to release 32nm tech before AMD, its just simple economics, but will Intel have it better than AMD will eventually do? I honestly don't know, 32nm is small...and its pretty much new territory for mass production, although so was 45nm...and 65...oh well, well see.

 

[cdawall]

I agree with you, innovation is just as important as performance, however in the previous posts everyone seems to view sheer performance as the measuring stick and hence why I disagreed that if performance is the only indicator than AMD have been doing well in prior years. Personally I like prices that are relative to performance, innovation is good too but if the processor has 45 GBs of cache and 100 cores but costs £10,000 I'm still not buying it. We are going to have to agree to disagree for the majority of this architectural debate as I'm sure that we are not making much progress.

As for the Sempron I was reading a classic Xbit Labs review of the Sempron 2600 @ 2.5 GHz destroying a Pentium 4 E @ 3.4 GHz. Interesting read.
Here

To be fair the slighty OC'd sempron destroyed even the high end Athlons. Semprons were absolute beasts and workhorses. Ultra cheap too.

not quite they are about 15% weaker clock for clock against a standard athlon same goes on the sempron X2 vs athlon X2 a windsor chip clocked at 2.8ghz could out perform my sempronX2@3.2ghz 256KB L2 vs 1MB L2 cache makes alot of difference.

Thats wayyyyy offff. Intel held the crown since the early 70's. Then they messed up with P4, and AMD had the crown for a few years. Now after Core 2, and now with Core i* Intel is once again holding the crown. AMD has had to deal with the back seat a lot longer then Intel had to in the past few years....

That being said, I don't find it surprsing Intel is slated to release 32nm tech before AMD, its just simple economics, but will Intel have it better than AMD will eventually do? I honestly don't know, 32nm is small...and its pretty much new territory for mass production, although so was 45nm...and 65...oh well, well see.

not 100% correct P3 vs K6-2/K7 was pretty much a toss up a 1600+ and P3 1.4ghz ran about the same cost the same etc.

 

[btarunr]

Let's get back to topic. It's not that every news post about AMD or Intel should descend into an Intel vs. AMD discussion.

 

[Wile E]

Can't wait to see how well 32nm clocks, and if they are able to keep electron migration in check on such a small process.

 

[FordGT90Concept]

That being said, I don't find it surprsing Intel is slated to release 32nm tech before AMD, its just simple economics, but will Intel have it better than AMD will eventually do? I honestly don't know, 32nm is small...and its pretty much new territory for mass production, although so was 45nm...and 65...oh well, well see.

AMD co-develops new processes with IBM. AMDs 90 and 45 are pretty good but their 65 was crap. Intel has never really stumbled with a new process but there is a first for everything.

 

[Guru Janitor]

AMD co-develops new processes with IBM. AMDs 90 and 45 are pretty good but their 65 was crap. Intel has never really stumbled with a new process but there is a first for everything.

I know that, but this is really small...what is it, two processes above nanoelectronics? I'm not saying they are gonna mess something up, or its not gonna be the best it can be when they release, or in the coming months after, I'm just saying its interesting how its gonna play out going smaller and smaller.

 

[btarunr]

AMD co-develops new processes with IBM. AMDs 90 and 45 are pretty good but their 65 was crap. Intel has never really stumbled with a new process but there is a first for everything.

Intel stumbled with 90 nm.

 

[mdm-adph]

Intel wasn't found guilty in the USA.

Given the traditional pro-business-no-matter-who-gets-screwed-over political climate in the US, the fact that Intel wasn't found guilty in the USA means absolutely nothing to me.

Just because the US won't take a stand on something doesn't make it okay. The US has never joined international treaties banning the use of landmines, for example -- does that make them okay, too?

The fact remains is that Intel has been a very, very naughty company in the past, and everything they do can be seen in suspect because of that, which was what I was trying to get at.

 

[a_ump]

So what is the limit to silicon's manufacturing process? 11nm? what will TSMC/Global Foundries use then? wonder if once we get to 11nm, there'll be like a delay where companies(Nvidia, Intel, AMD) are stuck using 11nm for a while till they come up with something else, or is there already something in the works to replace silicon? i'm not too savy on this subject so if there's a good read or just general info on it post it

 

[FordGT90Concept]

I know that, but this is really small...what is it, two processes above nanoelectronics? I'm not saying they are gonna mess something up, or its not gonna be the best it can be when they release, or in the coming months after, I'm just saying its interesting how its gonna play out going smaller and smaller.

Nanoelectronics are around 10nm.

Intel stumbled with 90 nm.

I was thinking the same thing actually. It wasn't as bad of a stumble as AMD's 65nm but a stumble nevertheless.

Given the traditional pro-business-no-matter-who-gets-screwed-over political climate in the US, the fact that Intel wasn't found guilty in the USA means absolutely nothing to me.

It should. Screwing business means screwing employees which means screwing the economy. Anti-trust laws only come in to play in the USA when there is a blatant violation of fair business practices (e.g. buying up all your competitors so you have no competition).

Just because the US won't take a stand on something doesn't make it okay. The US has never joined international treaties banning the use of landmines, for example -- does that make them okay, too?

Considering USA is about the only country that still conducts wars, most countries have little use for landmines anymore. We do. Landmines are excellent area denial weapons and much cheaper than $100,000 bombs.

The fact remains is that Intel has been a very, very naughty company in the past, and everything they do can be seen in suspect because of that, which was what I was trying to get at.

You can't stop a mob determined to burn a "witch" (see Salem Witch Trials). Your perception of "fact" varies from mine. Just because a court in EU and a court in Japan considers it fact doesn't make it a "universal truth."

 

[Flyordie]

If what I read was correct AMD may skip 32nm and go straight to 22nm if GLOBALFOUNDRIES has their way. This is more rumor than it is fact but I think it might be true... 45nm for AMD clocks real well and has decent efficiency so... why change it*?
*so soon
---OFF TOPIC BELOW---
Also, Intel has yet to be found guilty in the US because the cases are not over yet... Intel's final trial is set for August 09... the judge is very pissed off at Intel for screwing with him over and over again... so I doubt this will end well for Intel.
---
As for the FTC investigation... FTC has said that they will take all the findings from the AMD vs Intel case to assist in putting down their possible fine.
---
As for the state of NY... same applies...