Intel said 6-7 years ago to reach 10Ghz by 2008

[paadness]

I saw this years ago on some new channel where intel saying like 10 Ghz im sure.

AMD was right, dual core making progress. 10 Ghz goal lost.

Damn u Intel.

 

[Hyperlite]

brilliant foresight for you.

 

[clarkey01]

Originally posted by: paadness
I saw this years ago on some new channel where intel saying like 10 Ghz im sure.

AMD was right, dual core making progress. 10 Ghz goal lost.

Damn u Intel.

They said two years ago that Nehalem would hit 10.25 Ghz..In 2005.

 

[bjc112]

I don't even think we will see Intel @ 5ghz before 2007.

With AMD's current heat out put and power consumption, I would expect to see dual cores in the 3.5ghz range by then..


A man can hope!

😛

 

[thegimp03]

I remember them thinking they could hit 10 ghz by then too. Little did they know the factor heat would play. 🙂

 

[theMan]

planning ahead always helps. if intel wants to survive, they are going to have to start fresh and redesign their chips.

 

[bjc112]

Originally posted by: theman
planning ahead always helps. if intel wants to survive, they are going to have to start fresh and redesign their chips.

intel could sit on the same chips for another year or so and still have NO problem surviing..

 

[akugami]

Intel did redesign their chips, it's called the Pentium M, this is a vastly superior CPU to the Pentium 4. There is no way I will consider a P4 system, if they make dual core PM's then i would have to consider it after factoring in if they improved their dual core architecture from the current bandwith starved implementation.

 

[cbehnken]

Originally posted by: akugami
Intel did redesign their chips, it's called the Pentium M, this is a vastly superior CPU to the Pentium 4. There is no way I will consider a P4 system, if they make dual core PM's then i would have to consider it after factoring in if they improved their dual core architecture from the current bandwith starved implementation.

I hate to tell you buddy, by the Pentium M isn't a P4 redesign. It's based off the P3 and was developed alongside the P4.

 

[dmens]

if they make dual core PM's then i would have to consider it after factoring in if they improved their dual core architecture from the current bandwith starved implementation.

Soon there will be fully buffered DIMM DDR2/DDR3 DIMM's and the CSI interface estimated at PC6400 equivalent combined with larger caches... that weakens the case for an integrated memory controller for low-mid desktop and mobile parts.

Servers and hi-end desktops still need them though, even with massive last level caches.

As for the processing cores themselves, P-M will do just fine, after some widening.

 

[bunnyfubbles]

GHz is a joke, Intel just promoted it because its a marketable term that your everyday average joe can understand or at least acknowledge that "more is better", however that isn't always true as Intel brilliantly dissplayed with their P4 line.

I dont' want a 10GHz CPU if I can have a quad core 3GHz CPU that is faster in all but the most trivial apps and situations. Parallelism, not Hz, is the future until we move away from binary computing.

 

[RichUK]

i read somewhere that on current technology we can not go over 50Ghz, and i also read that there is another technology that will allow over 100Ghz .. just thought i would let you guys know 😉

Anyways its not about speed its about efficiency! (ipc) .. and intel have proved that speed is not a good formula when it comes to thermals .. PresHOOOOTTTTSS hahaha

 

[Elcs]

Originally posted by: bjc112
I don't even think we will see Intel @ 5ghz before 2007.

With AMD's current heat out put and power consumption, I would expect to see dual cores in the 3.5ghz range by then..


A man can hope!

😛

Can we say Dual P4 2.5's = 5.0Ghz? 😛

Id not be suprised if Intel try and use that logic to dig theirselves out of such a hole.

 

[DrMrLordX]

Originally posted by: Elcs
Can we say Dual P4 2.5's = 5.0Ghz? 😛

Id not be suprised if Intel try and use that logic to dig theirselves out of such a hole.

They've already got dual-core CPUs in the 2.8-3.2 ghz range, and they're not selling them as 5.6 ghz processors(or anything of that sort). So, no, they won't be doing that.

And, yes, I know availability of the Pentium D is poor/non-existant. Just making a point about the marketting of the chips, that's all.

 

[TuxDave]

I find it really funny how everyone is an expert now. Two years ago, everyone on the forum was rooting for higher and higher clock speeds for both AMD and Intel processors. Take whatever AMD had at that time, I'd read people saying stuff like "Yeah, now once this AMD processor scales up in speed, we'd REALLY be rocking Intel's boat." No one suggested "Hey, maybe we shouldn't be increasing clock speed. We should really be looking into parallelism and dual cores"

Now that the theme of parallelism comes to fruitation, everyone's like "duuuh... higher clock speeds wasn't the answer! Stoopad Intel"

 

[Tweakin]

Originally posted by: TuxDave
I find it really funny how everyone is an expert now. Two years ago, everyone on the forum was rooting for higher and higher clock speeds for both AMD and Intel processors. Take whatever AMD had at that time, I'd read people saying stuff like "Yeah, now once this AMD processor scales up in speed, we'd REALLY be rocking Intel's boat." No one suggested "Hey, maybe we shouldn't be increasing clock speed. We should really be looking into parallelism and dual cores"

Now that the theme of parallelism comes to fruitation, everyone's like "duuuh... higher clock speeds wasn't the answer! Stoopad Intel"

:beer:

 

[Chode Messiah]

I think AMD would probably reach 5ghz before Intel, unless there is a serious thermal redisign for intel cpus. We ought to sell prescotts to the military so they can burn terrorists out of caves with the immense heat!:roll:

 

[Insomniak]

Originally posted by: TuxDave
I find it really funny how everyone is an expert now. Two years ago, everyone on the forum was rooting for higher and higher clock speeds for both AMD and Intel processors. Take whatever AMD had at that time, I'd read people saying stuff like "Yeah, now once this AMD processor scales up in speed, we'd REALLY be rocking Intel's boat." No one suggested "Hey, maybe we shouldn't be increasing clock speed. We should really be looking into parallelism and dual cores"

Now that the theme of parallelism comes to fruitation, everyone's like "duuuh... higher clock speeds wasn't the answer! Stoopad Intel"

My position has always been simple: If the performance is higher, I don't care about the method (all other things being equal). If you could take p2 450Mhz processors and OC them to 12 TeraHertz, I'd have one in my rig.

That said, I've been saying since 2002 that ramping clockspeed was not the answer. I wasn't touting parallelism at the time either - I was simply of the opinion that Intel's "speed kills" method was getting rather silly, especially with the predicted (and now verified) problems of running high clockspeeds and long pipelines on small manufacturing processes.

My call was this: Processors need to become more efficient, not necessarily faster or wider. How they gain the efficiency was up to them, but we're rapidly approaching the barrier where we will no longer be able to shrink dies and tack more stages onto the pipeline. That's coming to an end, and sooner than anything thinks to boot.

Processor companies need to be looking to pull every ounce of performance out of every transistor, and AMD has been taking that road moreso than Intel for the past 5 years. Thus, I have though highly of AMD during this time, and I think it goes without saying that their processors are more elegantly designed and more capable than Intel's comparable offerings. Intel, however, seems to have learned from their mistakes and are following AMD down the wider trail. However, while this may buy some time, efficiency still needs to be the number 1 priority for everyone at this point. The other options are running out.

 

[bunnyfubbles]

Originally posted by: TuxDave
I find it really funny how everyone is an expert now. Two years ago, everyone on the forum was rooting for higher and higher clock speeds for both AMD and Intel processors. Take whatever AMD had at that time, I'd read people saying stuff like "Yeah, now once this AMD processor scales up in speed, we'd REALLY be rocking Intel's boat." No one suggested "Hey, maybe we shouldn't be increasing clock speed. We should really be looking into parallelism and dual cores"

Now that the theme of parallelism comes to fruitation, everyone's like "duuuh... higher clock speeds wasn't the answer! Stoopad Intel"

Considering super computers use parallelism it has been obvious for quite some time that singe CPUs (or single core) would not cut it forever.

Then there will be an eventual limit to parallelism because communication between the different parts won't be efficient for increasing speed or processing power.

 

[Chode Messiah]

good point. U can only have so much bandwith

 

[DrMrLordX]

Originally posted by: TuxDave

Now that the theme of parallelism comes to fruitation, everyone's like "duuuh... higher clock speeds wasn't the answer! Stoopad Intel"

Eh? I can't speak for anyone else, but I wasn't alone when the POS 1.8 ghz Willamettes came out and had serious problems beating 1.4 ghz Thunderbirds that had been on the market for awhile. It was obvious at that time that a shift to the Netburst architecture for more mhz didn't necessarily result in increased performance. We all learned that having a better core design was sometimes preferable to having a core that could scale to higher speeds.

I also remember a lot of Intel faithful moaning and groaning about how Intel pushed aside the Tualatin in favor of Netburst CPU development, etc etc(yes, Tualatins were eventually released, but only after lengthy delays).

So, some of us have been skeptical of higher clockspeeds for some time now. It didn't take the Prescott to teach me that.

 

[dmens]

Originally posted by: DrMrLordX
It didn't take the Prescott to teach me that.

congrats, you're a rare bird.