Next-gen Intel notebook chips to exceed 3.0GHz

[AppleInsider]

Intel this spring will launch its next-generation Centrino notebook platform alongside a half dozen new 45 nanometer mobile chips that will eventually make their way into Apple's MacBook and MacBook Pro offerings.

The platform, long referenced by its code-name "Montevina," will be officially dubbed Centrino 2 when it makes its debut at the Computex Taipei 2008 conference, which runs June 3 - 7, according to DigiTimes.

Although Intel has used the Centrino brand name for four generations of its notebook platforms -- Carmel, Sonoma, Napa and Santa Rosa -- the unchanging brand name has reportedly resulted in lower market recognition, as consumers are sometimes unable to identify the differences.

The Santa Clara, Calif.-based chip maker hopes the new marketing strategy will "give consumers the feeling of a tangible upgrade and will hopefully spur replacement demand," according to the report.

It will simultaneously deliver six new 45-nanometer notebook processors, running at speeds between 2.26GHz - 3.06GHz, all of which will sport a 1066MHz front-side bus, compared to today's models which scale up to 2.6GHz and operate on an 800MHz bus.

In addition, DigiTimes reports that Intel will then launch seven new 45-nm small form factor processors, like those used in Apple's new MacBook Air, sometime during the third quarter of the year.

The introductions should provide Apple with a means to refresh its mainstream MacBook and MacBook Pro family of notebook systems sometime during the summer educational buying season, and its MacBook Air sub-notebook line in time for the holidays.

Meanwhile, MacBook Pro models including Intel's just-released 45-nanometer Penryn processors -- the last of the Santa Rosa generation -- remain on tap for a release anytime in the next few weeks.

[ View this article at AppleInsider.com ]

[TenoBell]

The 3.06Ghz will more than likely be in the iMac refresh this year and in the MBP sometime in 2009.

[Rokken]

Are they going to update MacBook alongside MacBook Pro?

[Cubert]

What is up with 3GHz being a barrier in chip production? I know clock speed is not as important as it was 10 years ago, but it just seems that Intel is having problems similar to what IBM had several years ago.

[Pine]

Should be expect only a processor upgrade with the Penryn's with a major overhaul when the Montevina's come out or will the overhaul come with the Penryn update? The powerbook will chug until we find out I guess.

That Montevina announcement is pretty close to when WWDC will be...

[backtomac]

Originally Posted by Cubert

What is up with 3GHz being a barrier in chip production? I know clock speed is not as important as it was 10 years ago, but it just seems that Intel is having problems similar to what IBM had several years ago.

It's not a barrier. Intel have no one pushing them. Intel's desktop chips easily overclock to 3.6 ghz. With no additional cooling.

When the core 2s were first released they hit 3 ghz. Intel needed a competitive product then to win back market share from AMD. Since then AMD have had nothing to challenge Intel and advances have slowed. Although in all fairness Intel have increased the number of cores to 4 per chip on the desktop and work station.

[Abster2core]

Originally Posted by backtomac

Since then AMD have had nothing to challenge Intel and advances have slowed.

Probably the ludicrous reason I've ever heard.

[Phlake]

Simply because no one has said it yet:

3 GHz in 2008

[backtomac]

Originally Posted by Abster2core

Probably the ludicrous reason I've ever heard.

You're right. I forgot to mention that the G5 hasn't hit 3 ghz yet either.

So Intel don't have IBM pushing them either.

[Abster2core]

Originally Posted by Cubert

What is up with 3GHz being a barrier in chip production? I know clock speed is not as important as it was 10 years ago, but it just seems that Intel is having problems similar to what IBM had several years ago.

Perhaps a little background homework and research is appropriate. At least, it is better than some of the responses you'll get here.

http://www.associatedcontent.com/art...trial_and.html

A better blogging site might be Intel' http://blogs.intel.com/research/

[brianus]

Hmm. I wonder what this does to Apple's release schedule. With the MBP they've averaged about 9 months between upgrades thus far. If they release a Penryn MBP soon and a Montevina in, say, July, there'd be gaps of just 7 and then 5 months, respectively. 5 months! One wonders if they'll just skip the current updates altogether and just wait till June...

[Pine]

Originally Posted by brianus

Hmm. I wonder what this does to Apple's release schedule. With the MBP they've averaged about 9 months between upgrades thus far. If they release a Penryn MBP soon and a Montevina in, say, July, there'd be gaps of just 7 and then 5 months, respectively. 5 months! One wonders if they'll just skip the current updates altogether and just wait till June...

God I hope not. Occasionally they have done quick updates, but on the other hand it would make sense. Apple used to be ahead of the game with processors. Although I would buy a MacBook Pro with a Penyrn processor, every other company already has them out, so what is the innovation from Apple? Would they be content just playing catch up?

[shanmugam]

Originally Posted by Abster2core

Probably the ludicrous reason I've ever heard.

sometimes truth may be hard to digest ... no one is pushing intel ... already rumors are CPU delivery dealys are part of intel's part of slowing down the engine

[shanmugam]

is it not too early for WWDC rant? where is the updated MBP first?

[mjteix]

For those who want to know some details about the prices/models, here they are:


IMO, this is the REAL penryn update. Montevina should be in every Mac except the Mac Pro:
2.40/2.53/2.80 MacBook Pro (25/35W)
2.26/2.53/3.06 iMac and a quad-core option later in the fall (25/35/45W)
2.26/2.40 MacBooks (25W-3MB)
2.26/2.40 Mac mini (25W-3MB)
1.60/1.86 MacBook Air (while I think Apple could only use the 1.86GHz cpu given the price difference $32).

Of course, other products could be released using the "S" small package series in the small notebooks/desktops area. But these chips are $75 more expensive than their P counterparts,

And yes, the initial launch of Montevina will be really close to the WWDC.
June: P, T, X series
Q3: SP, SL, SU, U, 723 series

[Abster2core]

Originally Posted by shanmugam

sometimes truth may be hard to digest ... no one is pushing intel ... already rumors are CPU delivery dealys are part of intel's part of slowing down the engine

The second most ludicrous answer I've ever heard.

Perhaps a perusal through the Harvard Business School Working Knowledge will help direct you on the right path. Intel's, and a matter of fact, Apple's reference lectures are not there because of ridiculous business practices as you and some others so offer.

Do yourself a favor. Explore Intel's commitment to research, and the absolute and relative millions they like Apple, and AMD, spend to finance it.

[cameronj]

Originally Posted by Abster2core

The second most ludicrous answer I've ever heard.

Perhaps a perusal through the Harvard Business School Working Knowledge will help direct you on the right path. Intel's, and a matter of fact, Apple's reference lectures are not there because of ridiculous business practices as you and some others so offer.

Do yourself a favor. Explore Intel's commitment to research, and the absolute and relative millions they like Apple, and AMD, spend to finance it.

Are you nuts?

You think Intel is going to push forward when there is no competition?

[backtomac]

Originally Posted by Abster2core

The second most ludicrous answer I've ever heard.

Perhaps a perusal through the Harvard Business School Working Knowledge will help direct you on the right path. Intel's, and a matter of fact, Apple's reference lectures are not there because of ridiculous business practices as you and some others so offer.

Do yourself a favor. Explore Intel's commitment to research, and the absolute and relative millions they like Apple, and AMD, spend to finance it.

No one is questioning Intels commitment to research.

Do you really think that the highest clocking Penryn available would be 3.2 ghz if Barcelona was at 3 ghz?

[Hattig]

Originally Posted by backtomac

You're right. I forgot to mention that the G5 hasn't hit 3 ghz yet either.

So Intel don't have IBM pushing them either.

Well, the G5 did go dual-core instead.

Cell (designed by IBM) has hit >6GHz though (in its 45nm incarnation) ... shame that the PPU is in-order, even if it can run two threads at the same time, as that limits its actual general purpose processing power.

(I imagine that means that the XBox360 CPU could hit 6GHz as well as it uses a very similar PPU core to Cell, apart from the fact that won't ever happen as the CPU has no use outside of the console)

These new mobile chips sound great though. I wouldn't mind a 3GHz 24" iMac or MacBook

[Abster2core]

Originally Posted by cameronj

Are you nuts?

You think Intel is going to push forward when there is no competition?

If you really think that there is no competition, I have some land in the Everglades for you.

Do you really think that Intel isn't aware of guys such as these; http://www.sciencedaily.com/releases...0112083626.htm or aren't researching such likewise themselves or in collaboration with others?

Do you really think that universities, even those who enjoy being the beneficiaries of Intel's (Worldwide) Research Grant Programs are solely exclusive to Intel?

You don't stay number one sticking your head in the sand.

[wk127001]

The reason no one is past roughly 3GHz is because it's physically not possible without better, much more expensive conductors. Usually a college level electric circuits class covers this.

[solipsism]

Originally Posted by Pine

Should be expect only a processor upgrade with the Penryn's with a major overhaul when the Montevina's come out or will the overhaul come with the Penryn update? The powerbook will chug until we find out I guess.

I agree with melgross, any case overhaul will most likely happen with the Montevina. There are so many significant changes compared to Santa Rosa to Penryn upgrade.

Originally Posted by brianus

Hmm. I wonder what this does to Apple's release schedule. With the MBP they've averaged about 9 months between upgrades thus far. If they release a Penryn MBP soon and a Montevina in, say, July, there'd be gaps of just 7 and then 5 months, respectively. 5 months! One wonders if they'll just skip the current updates altogether and just wait till June...

I don't foresee a Montevina until August—October unless Intel has production flying along by then.

Originally Posted by Abster2core

If you really think that there is no competition, I have some land in the Everglades for you.

There is competition, but Intel has also slowed down releases because they not as directly challenged as years past. There have been many rumours of this floating around the nets and I agree. Intel is still maintaining their clear lead, but they also aren't pushing ahead with new releases when they can milk the current tech a little longer. In other words, make more dinero.

[JeffDM]

Originally Posted by cameronj

Are you nuts?

You think Intel is going to push forward when there is no competition?

Even if there was no competition, Intel has to compete with their existing product in the field. Computers can easily last 10+ years but part of the reason people usually don't use them that long is because new ones are so much faster. If the new ones weren't faster, then there's less reason to upgrade. The progress might slow down, but generally it won't stop for the above reason.

[backtomac]

Originally Posted by wk127001

The reason no one is past roughly 3GHz is because it's physically not possible without better, much more expensive conductors. Usually a college level electric circuits class covers this.

Current Intel chips, the old first gen core 2 chips easily overclock to 3.6 ghz. On stock cooling.

Intel could release release 3.6 ghz chips(desktop) now if they wanted to.

They don't have to.

[backtomac]

Originally Posted by Abster2core

If you really think that there is no competition, I have some land in the Everglades for you.

Do you really think that Intel isn't aware of guys such as these; http://www.sciencedaily.com/releases...0112083626.htm or aren't researching such likewise themselves or in collaboration with others?
.

And what product do they have?

Research is nice but I don't see a product that competes with existing products that Intel have.

Intel aren't afraid of vapor ware. They don't have to be. It's one thing to talk of theoretical performance but quite another bring it to market at quantities that meet demand. A hard lesson that AMD is learning now.

You still don't have a good answer for why Intel haven't tapped the potential of the existing chips.

I'll give you a hint. If and when AMD are able to ramp up clock speeds on Barcelona/Phenom you'll see some rapid clock speed increases in Intel's chips. The potential is there. They're just sitting on it now.

[Bigc]

Originally Posted by backtomac

And what product do they have?

Research is nice but I don't see a product that competes with existing products that Intel have.

Intel aren't afraid of vapor ware. They don't have to be. It's one thing to talk of theoretical performance but quite another bring it to market at quantities that meet demand. A hard lesson that AMD is learning now.

You still don't have a good answer for why Intel haven't tapped the potential of the existing chips.

I'll give you a hint. If and when AMD are able to ramp up clock speeds on Barcelona/Phenom you'll see some rapid clock speed increases in Intel's chips. The potential is there. They're just sitting on it now.

In order for Intel to make newer better chips they have to make a profit to pay for the research and only they know what is required...

[JeffDM]

Originally Posted by backtomac

I'll give you a hint. If and when AMD are able to ramp up clock speeds on Barcelona/Phenom you'll see some rapid clock speed increases in Intel's chips. The potential is there. They're just sitting on it now.

I thought Intel chips have generally been hugely overclockable, even when AMD chips were beating Intel in terms of speed. The way that Intel critics put it, Intel could have / should have produced a 4GHz P4 chip because their chips could overclock beyond that in free air, but I think it ignores certain considerations that Intel uses to determine whether to market at that speed. I don't know if the overclockers were really testing the chip over the entire environmental range like -10C to 85C or whatever spec that Intel's consumer chips are supposed to meet.

[tuneman07]

Well as usual I wonder what real world differences are going to be here. Anyone have any guesses? Is a 3.0ghz 45nm chip going to save a ton of battery and be noticeably faster? I like to know before I get too excited and I am no computer processor guy lol

[Rolo]

Why is the subject of this article at odds with the body text? Nowhere in the article is there any mention of any chip that will exceed 3 GHz. Why the hype?

The real advantages of Montevina:
266 MHz faster bus speed
More efficient chipset that's 40% smaller
Optional integrated Wi-Fi and WiMAX
Integrated hi-def video decode

The key to Montevina is the smaller size of the chipset will allow for sleeker notebooks. It's all about improved portability and wireless connectivity, not raw power. The speed improvements will be very modest.

[Lorre]

Originally Posted by wk127001

The reason no one is past roughly 3GHz is because it's physically not possible without better, much more expensive conductors. Usually a college level electric circuits class covers this.

Indeed, hand't it something to do with pipelining as well? At 3 GHz, light travels like 4 inches per clock-pulse, I know transistors are rediculously small nowadays, bubt still, something nice to think about, and quite obvious why we'll never have 10 GHz computers

Also, I'm under the impression that the GHz-myth is slowly being replaced with the cache-myth.

[melgross]

Originally Posted by Cubert

What is up with 3GHz being a barrier in chip production? I know clock speed is not as important as it was 10 years ago, but it just seems that Intel is having problems similar to what IBM had several years ago.

It's not a problem. It's just that when the various companies re=tooled their lines for multiple cores, the power draw and heatsinking became so much more difficult, that they traded off some speed for the extra cores. as new process technology comes out, those speeds have continues to rise.

But remember that they are talking about mobile chips, which have always been slower than desktop chips because of those very same power and heat problems. Desktop chips have been over 3.0 GHz for some time now.

[melgross]

Originally Posted by cameronj

Are you nuts?

You think Intel is going to push forward when there is no competition?

It's interesting that some people don't understand that Intel competes against itself almost as much as it competes against others.

If chip performance doesn't rise in a timely manner, then Intel's sales fall, because why buy new machines if the performance is little changed?

While Intel may feel rushed when AMd has more competition, and so rushes to market a bit faster, we are only talking about a month or two, here and there. Not enough to make much of a difference, except to those few who slaver at the thought of something newer three months after they bought their last machine, or who can pretend that they will but a new machine.

If Intel can breath for a month longer on a design, and so squeeze a bit more profit out of it, and give the new chips one more iteration before going to production, that's a good thing.

[melgross]

Originally Posted by backtomac

Current Intel chips, the old first gen core 2 chips easily overclock to 3.6 ghz. On stock cooling.

Intel could release release 3.6 ghz chips(desktop) now if they wanted to.

They don't have to.

But when doing so, they consume much more power. This is a power game as much as a performance game these days.

[melgross]

Originally Posted by tuneman07

Well as usual I wonder what real world differences are going to be here. Anyone have any guesses? Is a 3.0ghz 45nm chip going to save a ton of battery and be noticeably faster? I like to know before I get too excited and I am no computer processor guy lol

There are tradeoffs. Either much less power, much faster chips, or some combo. Usually it the combo choice.

[melgross]

Originally Posted by Lorre

Indeed, hand't it something to do with pipelining as well? At 3 GHz, light travels like 4 inches per clock-pulse, I know transistors are rediculously small nowadays, bubt still, something nice to think about, and quite obvious why we'll never have 10 GHz computers

Also, I'm under the impression that the GHz-myth is slowly being replaced with the cache-myth.

It's an interesting thought.

The reason why Intel uses such large caches is because the frontside buss isn't fast enough for all the memory hits the cpu needs to make. Some programs are affected severly by this, while others aren't.

Once Intel moves to Nehalen later this year, with its onboard memory controller (like AMD), then the much wider bandwidth, with the ability of each core to access the memory directly, Intel will be shrinking the caches somewhat, and the large size will be less useful.

10GHz, well it's possible.

[backtomac]

Originally Posted by melgross

It's interesting that some people don't understand that Intel competes against itself almost as much as it competes against others.

If chip performance doesn't rise in a timely manner, then Intel's sales fall, because why buy new machines if the performance is little changed?

While Intel may feel rushed when AMd has more competition, and so rushes to market a bit faster, we are only talking about a month or two, here and there. Not enough to make much of a difference, except to those few who slaver at the thought of something newer three months after they bought their last machine, or who can pretend that they will but a new machine.

If Intel can breath for a month longer on a design, and so squeeze a bit more profit out of it, and give the new chips one more iteration before going to production, that's a good thing.

I agree with this but still think Intel would ramp up their clock speeds more aggressively if Barcelona/Phenom were more competitive performance wise.

The top performing chips are the ones with the highest price tags and greatest margins so there's more at stake than just bragging rights.

[melgross]

Originally Posted by backtomac

I agree with this but still think Intel would ramp up their clock speeds more aggressively if Barcelona/Phenom were more competitive performance wise.

The top performing chips are the ones with the highest price tags and greatest margins so there's more at stake than just bragging rights.

I don't think there would be too much difference.

Even if there was, there would be at least as many people complaining that manufacturers are obsoleting their machines too fast, as there are now people who are complaining that things are moving too slowly.

How many times have we heard the complaint that Apple is obsoleting their machine too quickly? Especially just after someone buys a new one a few weeks before some Apple show.

[Tailpipe]

I've been trying to analyse the implications of Centrino 2 in terms of what this may mean for the MacBook Pro and MacBook ranges.

First of all, there can be no doubt that the Montevina chipset with Penryn 45 nm processors represents a real step-up in terms of performance. We'll get faster Macs with longer batttery life and smaller form factors.

While we know for sure that a new MacBook Pro is in the pipeline, the $1,000,000 question is will Apple simply refresh the existing MBP model with Penryn now and wait until September before launching the all new version?

I think the answer depends on whether the form factor can be reduced without needing Montevina components. If it can be, expect brand new design of MacBook Pros imminently. If not, then it will be in 6 month's time for sure. From the little i know, I'm pretty sure that the MBP update WILL feature a new case design now.

In terms of the MacBook, I don't think this will be updated until Montevina is ready. Therefore, my money is on new MacBooks in May or June.

Of the chips listed above, to me the most interesting are the SP9400 and Sl9400 running at 2.4 Ghz and 1.86 Ghz respectively and using 25 watts or 17 watts of power respectively. With package size of 22 mm, these processors should be ideal for use in the MacBook Air, which currently offers only 1.6 GHz and 1.8 Ghz chips. Montevina low voltage chips would provide quite a step-up in performance without a power-consumption penalty.

What do you think?

[JeffDM]

Originally Posted by Tailpipe

Of the chips listed above, to me the most interesting are the SP9400 and Sl9400 running at 2.4 Ghz and 1.86 Ghz respectively and using 25 watts or 17 watts of power respectively. With package size of 22 mm, these processors should be ideal for use in the MacBook Air, which currently offers only 1.6 GHz and 1.8 Ghz chips. Montevina low voltage chips would provide quite a step-up in performance without a power-consumption penalty.

MBAir uses 20W parts, and it seems like they use their own qualification bin.

http://www.anandtech.com/mac/showdoc.aspx?i=3203

I don't think it would be smart to go to a higher watt chip. Maybe a refresh would have 1.8GHz & 2GHz chips.

[solipsism]

Originally Posted by JeffDM

I don't think it would be smart to go to a higher watt chip. Maybe a refresh would have 1.8GHz & 2GHz chips.

I'm thinking they will go from the current non-Santa Rosa Merom chip to the 22x22mm Montevinas in the Fall. There may also be a price drop for the MBA as the Montevina chips will more standardized.