News Bits: Intel Refresh Coming This Spring, MSI Wind Allows Overclocking

Intel Montevina refresh coming this spring

Intel will update its current Centrino 2 "Montevina" platform this spring with new chipsets and two new processors. April will see the launch of the GM47 chipset for high-end notebooks and the GL43 for entry-level. July/August will see the launch of the next-generation Capella platform, which will use the GM55 chipset.

The two new processors launching with the Centrino 2 refresh are the Core 2 Duo T9900 and the P8800. Intel will introduce several new processors for the Montevina platform before the end of this year, including the Core 2 Quad Q9600, and Core 2 Duo T9800, T9550, P9600, P8700, SP9600, SL9600, and SU9600. In addition, we should also see entry-level Celeron T1700 and T1600 dual-core processors launched before the end of the year.

Full Story (DigiTimes.com)

MSI Wind BIOS update allows overclocking

MSI just released a BIOS 1.09 update for its Wind netbook. This BIOS allows users to easily overclock their Wind's processors by pressing [Fn] + [F10] for an 8%, 15%, or 24% increase in clockspeed. A 24% overclock can reportedly yield up to a 30% performance increase over stock, which helps in processor-intensive applications such as Photoshop.

Full Story (ElectricVagabond.com)
Via (Gizmodo.com)
NotebookReview.com Review for the MSI Wind

Quanta to produce next-gen Intel Classmate PCs

According to a DigiTImes report, Quanta has reportedly landed orders from Intel for its next-generation Classmate PC. The new Classmate PCs will start shipping in the first quarter of 2009; shipment volume for 2009 is expected to be about two million units, up from 100,000 in 2007 and 500,000 in 2008.

Production of the current Intel Classmate will not stop with the introduction of the new version, and it will still be manufactured by Elitegroup Computer Systems.

Full Story (DigiTimes.com)

 

Hopefully when i7 is launched we will know the clock speed, capabilities, expected price, and release date of the Intel Core 2 Quad Q9600.

 

Man, the Wind is always looking more and more like a first choice for a netbook.

 

lets go intel lets go

 

Thanks for the latest New Bits, Chaz

Intel refresh for Spring...nice


Cin

 

Overclocking on the MSI Wind? Can you say sweet?

 

Yeah, thanks for the keeping us up to date Chaz. I love reading News Bits!

(OK, my REAL reason for posting was because I wanted to to promote my team too!)

 

Much obliged, always a pleasure.

 

Think about the heat that will come from that netbook after OC

 

Montevina's not even the dominant standard chipset for most new laptops yet... In less than a year, it's gonna be replaced?

 

tech. is moving very fast this days

 

I think digitimes made a small mistake. New quad to be released is Q9000 (2ghz quad core at $348).

T9900 at 3.06ghz and 35W TDP is very impressive.

 

Doesn't matter, it is still Core 2 Duo, and the Core architecture has been around since Jan 5, 2006. It is time to move on.

 

Thanks for the bits.

MSI Wind overclocking, now that is awesome. At this point, I would say it's the absolute best choice for netbooks.

 

i was trying to decide between an msi wind and an acer aspire one

being able to oc the wind just seals the deal

wind it is

 

My, my. While some people go thru the complex o'clocking process here comes a laptop that oc's for your preferences. This thing, at an overclocked state, does this also shortens the cpu's life similar to laptops which were manually overclocked? thanks.

 

1: Overclocking does not necessarily shorten CPU's life. If it's a pure overclocking (only increase CPU's frequency), then there is almost no lifespan decrease, However, if it's a overclocking boosted with a overvolting, then there might a lifespan decrease. If you overvolt too much, you can kill a CPU, But the only consequence of an extreme overclocking is system instability.

2: Also, even if overvolting can cause a lifetime decrease, the lifespan of a CPU is around 10 years so the decrease is not that dramatic.

3: Heat increase only happens with overvolting.

4: Most low-range CPU's can be overclocked, since most of them are just underclocked high-range CPUs

That's why manufacture dares to overclock their CPUs. They know that if they don't overvolt or don't overvolt too much, there is almost no repercussion.

 

overclock on wind is cool, its not new though.

Asus has there hybrid engine on the EEE's and it overclocks too, only 1.7ghz vs 1.6ghz stock BUT... that lead way for us to easily use programs like SETFSB, Clockgen, and others to overclock higher.

Most people can only hit about 2ghz before things go bad though, so I guess the wind is saying they can do the same.

Performance gain is very negligible though, its a single core atom. It may make the difference if your borderline from ok and slow but it wont make things magically be fast and smooth.

Wind is great and the main competitor to the EEE but I disagree that its the #1 choice as its hard as heck to find it with a 6cell battery (or at least it was) and it cost more than the EEE with the same/lesser specs.

 

^agree,eee 1000h is a first choice now IMO

 

The other thing is the Wind will void your warranty if you upgrade the ram big downfall.

First thing I did with my 1000H before it even booted for the first time was take 5 minutes to install a 250gb HDD and 2gb ram

EEE 1000H has only 1 flaw, the strange right shift key, and thats easily fixed with a software remap of the key to the up arrow.

Its good to see the competition is keeping things tight though. Finally the 6 cell Wind is out there.

 

That's what you think but situation on market is different .

Aspire one and Eee are far away from competition (I'm talking about shiped/sold numbers) .

Acer just replaced HP as top PC (overall) vendor in EMEA region because of Aspire one.

http://news.cnet.com/8301-1001_3-10066106-92.html
http://notechie.com/the-netbook-boostent-sales-of-pc/
http://www.itpro.co.uk/607155/netbooks-help-acer-top-hp-in-european-sales

 

Im not talking about sold numbers although I know the EEE is doing crazy good. Im talking about similar units.

The Aspire is not the same kind of unit because you cant touch type on it with its crazy small keyboard, it does not have bluetooth, its 9" not 10", ect.

The Wind & EEE 1000H are mirrors of each other basically and thus competing products directly.

edit: Here we go Jerry Shen an Asus CEO divulged this info not long ago:

 

did this have black colour

 

Well, there's also the lack of numlock and capslock lights... That's the one thing keeping me from getting an EEE1kH.

 

And here I am, on the verge of buying an X9100; I guess I'll just play the waiting game for the T9900, a chip I can be 100% sure won't burn a hole through my keyboard.

 

Montevina is a platform, not a CPU. Montevina's successor will use Core 2 Duo, etc., too.

 

I know, I messed up. Montevina is the platform, its sucscessor WILL NOT use Core 2 Duo. Core i7 (Nehalem) is a new architecture, and it will not use Core 2 Duo. It will take the Core architecture, refine it, and add new things to it.

 

NotebookReview.com: Thanks for the link

ViciousXUSMC: "Performance gain is very negligible though, its a single core atom. It may make the difference if your borderline from ok and slow but it wont make things magically be fast and smooth."
What do you base this on? Performance gain is substantial - over 40% increase over OEM installed Wind Bioses. It is a notable increase in performance that everyone who has upgraded has remaked on. No need to simply dismiss it as negligible.

"Wind is great and the main competitor to the EEE but I disagree that its the #1 choice as its hard as heck to find it with a 6cell battery (or at least it was) and it cost more than the EEE with the same/lesser specs."
You can't compare which machine is better based on availability that is flawed logic.

"The other thing is the Wind will void your warranty if you upgrade the ram big downfall."
This is incorrect.

"EEE 1000H has only 1 flaw, the strange right shift key, and thats easily fixed with a software remap of the key to the up arrow."
A machine with only 1 flaw? Why wouldn't the whole world get it! The 1000H is a great machine and is on par with the Wind, however there are some definite downfalls such as the aesthetics, the weight and size... all of which lose out to the Wind. Fortunately the touchpad is better and it has Wireless N.

Good to see this topic is generating some quality discussion though.

Spencer
Electric Vagabond dot com

 

Well the warranty issue has been well known and documented for awhile:

> http://www.bit-tech.net/news/2008/07/24/msi-wind-upgrading-memory-voids-warranty/1

> http://blog.laptopmag.com/msi-upgrading-wind-ram-voids-warranty

> http://www.warp2search.net/contentt...ind__upgrading_memory_voids_the_warranty.html

> http://www.overclock3d.net/news.php?/misc_content/msi_wind_ram_upgrade_issue/1

As you can see in the last source though looks like MSI took note of all the upset people and now will allow you to upgrade the ram, but only if you contact them first, use approved ram, and you have to then install a new warranty sticker after. So its still a major pain.

As for CPU performance, I am not speaking word of mouth. I have overclocked mine to 2.1ghz and you do not really see a increase in speed when doing something like loading photoshop, it does not fix the choppy playback of a 1080p video file. It will however increase your superpi score, or if you were at the border of 720p being smooth or not, it may fix that issue.

Thats exactly what I have experienced and exactly what I stated, I have a whole forum of people with the same opinion that I can link too for you in our EEE overclocking threads.

Yes you can when one product you can get with a 6 cell battery and 7 hour battery life, and the other you could only find with a 3 cell battery and 3 hour battery life. That puts them in totally different leagues.

Just now finally do I see the MSI wind with 6 cell on neweggs site, before now people have been trying everything to get a 6 cell, even going as to so far to buy the special "hearts" edition because it was the first semi wildly available version with the 6 cell and then they bought stickers to cover the ugly hearts up with.

Did you not just contradict yourself? In a notebook there are really only 2 main points for aesthetics, the keyboard and the touchpad. If the EEE has the better touchpad thats 50% better already, and the keyboard issue with the EEE has an easy fix.

 

Hey not trying to start a cyber-argument here or prove you wrong at all just clarifying a few things;

We are talking about the Wind here not the Eee. Sure they have similar specs but if you head on over to the Wind Forums (can't link as I don't have enough posts on this site) you'll see everyone ranting and raving about the performance increase.

Haha thats me included on the hearts one - my Wind is the Wind for Love edition unfortunately. However there is an abundance of vinyls and skins online to fix this for a few bucks. Availability and price are really tangential to a direct comparison of two pieces of hardware. They are caused by economic and market factors (such as SCM and SvsD) and not hardware factors. I can understand what you are saying though.

I think you might be a bit confused what I meant;
definition of Aesthetics in the Merriam Webster dictionary

perhaps you thought i was talking about ergonomics? In this case I would agree with you the 1000H has better ergonomics because of the large functional touchpad.

I'm not trying to say this netbook is better than that netbook and so on - I run an analysis site so enthusiasts and consumers can make educated decisions. I personally think that the MSI Wind and Eee PC 1000H are 1st equal in the netbook stakes although each one is suited to different people better.

 

Try identical specs, chipset cpu, all that is the same. Wireless or bluetooth may be different but thats not going to effect the cpu.

Also I find too many people just suffer the "sugar pill" effect when something like this comes out of the blue. I mean I do hardware reviews and build computers as a main hobby I know a performance difference when I see one and wont let numbers lead my thought process.

Percentages can be misleading, 40% overclock... wow thats big.

But its a 1.6ghz single core atom... break it down into reality its like saying instead of opening photoshop in 2 seconds now I can open it in 1.2 seconds. Thats 40% but only .8 seconds a difference, and of course the cpu wont speed up a program linearly like that. The HDD and things are still going to slow it down and keep it closer to the full 2 seconds.

If you have a 4ghz quad core and get a 40% overclock, now your talking a big jump! (I have a q6600 @ 3.7ghz and max it out often )

So moral of the story do not use 40% as your number use the real life gain, 400mhz. Anybody that is a computer enthusiast can tell you that even on a core2duo that is about 3x faster than an atom that 400mhz is small, and since the atom is seriously about 1/3 the power clock for clock that is only relatively equal to 130mhz difference on a computer system you may be familiar with, and that's totally insignificant.

 

I just realized that the T9900 is going to be a GM47. This means it's incompatible with my notebook, and I'll have to "settle" for the 2.93Ghz T9800.

 

Intel doesn't get it.

They need to come up with a tightly integrated system on a chip (SOC) that integrates:

Intel Atom dual core with a 1mb onboard cache
Chipset including memory controller
x4500 graphics

All made on a state of the art process (32nm) and consuming no more than 5watt TDP.

Then they need a comms chipset that dos 802.11n dual band (3 antenna), Bluetooth 2.x, and Wimax all on one card (and as few chips as possible) with a similarly low TDP.

Both of these chipsets need to be managed with some pretty sophisticated firmware / software to drop the actual power usage on average down to about 2 watt or lower.

The package need to support DDR3 (up to 4gb) for low power usage.

Then package it with an Intel SSD, call it Centrino 2 Lite.

Sell it t the box makers as an entire package.

The day when Intel can sell an Intel Atom with a clunky 945 or 950 chipset is fast coming to an end.

 

The x4500 for an IGP produces a lot of heat, it would be hard to fit into something like a netbook.

 

Um, if you read my post, it called for all 3 devices to be made on a single chip, presumably not at 65nm which is what the present x4500 is made on.

65nm is obsolete technology. the present Intel Atom n270 is made on a 45nm line --- so what you said would imply that Intel would take a step backward to build an integrated MPU, Cache, Chipset, and Graphics device on a 65nm line.

If the identical system on a chip is done on a 32nm process (read some serious die shrinks), and add to it a few power saving features, shouldn't be much of a problem to get it down to a couple of watts max out of a 5 watt MPU, Chipset, Graphics budget.

There is no factual basis for you to suggest that a die shrunk 32nm X4500 would run as hot as the current 65nm part.

Making the X4500 on a 65nm obsolete line is the old fashioned Intel Business model where power consumption is a distant 3rd or 4th consideration.

Integrating all 3 pieces on a state of the art chip on a 32nm line is what the overpaid monopolist ding-a-lngs at Intel need to figure out.

With AMD flat on its back and about to get carnal with foreign investors, there is no pressure on Intel to figure it out... until AMD releases such a chip.

 

According to the above news item, Montevina's successor, Calpella, will be out before Nehalem (IIRC, Nehalem for notebooks won't be out before fall 2009, at the earliest), so it will have to use Core 2 CPUs at launch.

 

I have no idea where you had gotten this entire rebuttal from my one little reply, I simply was telling you that the x4500 runs hot, its not like the lower end IGP's where they use low power and low heat so if you were thinking of putting it in a netbook it wont work well.

That is a fact and that is all I said.

 

I think the launch of Calpella in July/August are the Quad Core's under the Nehalem Core i7 brand, not Core 2 Duo.

 

I can see you have no idea!

That is not a fact (the X4500 necessarily runs hot and it "wont work well" in a netbook) and that is precisely my point.

Basic performance specs of these parts are available from Intel, including their TDP and what process it is made on.

If you want to talk about a hot part, the present Intel 945 GCGraphics core used in most Atom n270 netbooks roughly consumes about 22 watts (max) TDP, vs. the Intel Atom n270 processor's 1-2 watt TDP.

See: http://ark.intel.com/chipset.aspx?familyID=28994


The Intel G45 Chipset (with the X4500 core) has a maximum TDP of 24 watt for the 65nm version. See Table 2 here:

http://www.intel.com/Assets/PDF/designguide/319972.pdf

The difference between the current chips OK for netbooks according to you (945) is 22 watt max TDP, vs. the G45's 24 watt max TDP.

In actual use, given the more sophisticated speed controls and throttling on the G45 chipset / graphics, it can use far less power than the older, 945 chipset / graphics.

However, I was not referring to these legacy parts, but what Intel can build in mid 2009 for a future netbook.

Made on a 32nm process, the G45 chipset and graphics will use about 25 to 30 percent of the power of the 65nm part. It would work great in a netbook.



It is factually incorrect to say "lower end IGPs use low power and low heat".

They are, a) less complex parts, b) can run at lower clock frequencies and / or lower voltages, which in turn, generates less waste heat.

However, often the low end IGP parts are made on older processes, like the 90 nm process that makes them more power hungry (read heat dissipated) than parts made on more modern or state of the art processes.

Therefore, it is factually incorrect to state that lower end IGPs use low power. In fact, the older 945 chipset, made on a 90nm process, does less work and consumes nearly the same energy as the far more powerful G45 chipset. Furthermore, semiconductor chips do not "use heat" but dissipate heat as a waste product.



In your original comment, you are referring to a 65nm X4500 core which is today's version that is made on a 1 generation back line which is less energy efficient than something made on a current state of the art line.

I referred to a future x4500 product that can be made by Intel on their 32nm line scheduled to begin production in 2H 2009.

The x4500 core can be implemented at 32nm, which, all things equal, will lower its power consumption by about 70% vs the current 65nm part unless some of that gain is given up in greater clock speed or other features. Needless to say, it will also be far faster than the 945 chipset.

At 32nm, the entire chipset / graphics combo it will have a TDP of about 5 watt, well below its present power consumption and TDP of 22 watt. It will be, comparatively, a very cool part even if it is clocked faster than the current part. As such, it is perfect for a netbook.

Pretty elementary lesson from semiconductor manufacturing of what happens when you shrink the lithography by 50%.


These basic issues of semiconductor device and IC design and engineering can be explained to you by any competent engineer, or by consulting Intel's online educational materials, or Wikipedia.

 

Your going from papers there, your looking at the chipset as a whole. Im going from user commends that have the IGP.

TDP is only a part of what to look at, you should look at idle/load draw figures and TCase too.

Plus fact of the matter is null anyways when Intel already said (word of mouth here) that they will not place the x4500 with the atom. This makes sense when you think about the Asus N10 they had to use the low end Nvidia GPU for that unit when the x4500 would have worked better. The only reason I can think of is Intel would not allow them to use the x4500.

 

I am working off the chip manufacturer's design and technical specifications which are presumably the most accurate data, as opposed to impressions and user comments from a public forum. These facts are supported by Intel's technical documentation, which are linked and also widely cited by other sources.

Can you tell me how forum users that you rely on for "facts" accurately measure the power consumption of the chipset without access to things like prototyping boards and instrumentation? It can be done.... but not exactly for the faint of heart.

Hearsay claims from unnamed and unreferenced forum posts that is not back with any formal documentation and data is rather dubious.

We can debate which is the proper figure to use, such as idle power, etc. but the TDP is the figure that gives you the upper limit, and probably more salient. An older generation chip, all things equal, doing the same work, will run closer to its TDP limit than a new generation chip. (Hotter in your words.)

The Atom n270 is made on a 45nm line, while the 945 chipset is a 7 year old 90nm part that originally was a great fit with MPUs that dissipated as much as 70 watts TDP.

That is why the hottest chip in an Intel Atom n270 netbook is the chipset, not the MPU.

To pair these two devices, one an energy miser, the other a profligate power hog, is the greatest misfit ever since horses were used to pull motor cars. The only reason this is done is to lower costs for the netbook makers making $300 netbooks with a slim margin. For a $500+ netbook it makes no sense for Intel or for the netbook maker or the end user.

Similarly, the Via C7M runs very hot because it is a large die part made on a 65nm line, and its chipset and graphics are, likewise, old parts. It's sole claim to fame is price (cheap) and availability when the Atom is on allocation.


Insofar as this being a moot point. Far from it. Intel is at this very moment formulating plans for 2H 2009 when the 32nm process becomes available.

If there is a market for it that they make good money on, e.g. a premium priced netbook that delivers very long battery life with a highy optimized integrated MPU, Chipset and GPU, they can have the product out in time for Christmas 2009. Netbooks priced above $500 makes this possible.

The lead time to die shrink the parts and integrate them on a single die are about 3 months, then 4 to six weeks for a mask set, and a manufacturing cycle of about 3 months, plus 2 months slack / rework time. About 10 months overall. Barely have the chips ready for Christmas 2009. More than likely, the die shrinks of discrete components (G45 chipset) to 45nm from 65nm is well underway, and work on 32nm has already begun. The program just need to be accelerated.


Under such circumstances, contrary information via word of mouth (or keyboard) in the form of hearsay from a forum is rather shaky ground to make factual statements about Intel's decisions and intentions.

 

O I C .......

 

I've heard through the wire that the T9900 has been canceled by Intel. This is, of course, unconfirmed, but it's not speculation either.

Also ES versions of the T9800 and P9600 are now hitting eBay.

T9800 = $450, P9600 = $320.

 

Man, I'm really late. Nonetheless, thanks for the information, Chaz.

 

I am glad I found this. I was looking at buying a laptop soon but will wait a few months for the refresh or till Q3 when the cool stuff is out.

 

Heheh. This is an old thread that's been necro'd....everything from the original post has already come to pass and is actually approaching being obsolete.

 

Um, the Centrino 2 refresh is out, and has been for a while. And October is likely is the earliest at when the cool stuff will be out.

 

I believe Intel delayed Calpella until Q4 2009? In any case, I don't think you'll see Calpella in July/August.

 

So is the refresh the GM47?