Guide for New Notebook Technologies in 2006

by Dustin Sklavos, California USA

INTRODUCTION AND REFLECTING ON 2005

In this article, I'm going to be dealing in hard facts, coupled with a little bit of conjecture and speculation as to the future of mobile computing in 2006.

When we look back on 2005 we can see it as primarily a year of transition. Since the debut of Intel's Centrino platform, coupled with ATI's release of the Mobility Radeon 9600, notebooks have been maturing as a competitive computing platform at an alarming rate. While I don't really think 2005 will be remembered necessarily as a banner year in new technologies, I think it was a strong transitional period. Mobile technology is maturing at an alarming rate; in 2006, we're going to see some major milestones. I intend to cover most of them here.

2006 IN PROCESSORS

DUAL CORE AND X64

There are two major "no way back" technological milestones for processors right now: multi-core (remember, dual-core is just a stage) and 64-bit technology.

64-bit is by and large the standard for desktop processors now, with AMD finally picking up the rear by producing 64-bit Semprons. Unfortunately, the radical change in software required of 64-bit hardware, coupled with the kludge that is Windows XP Professional x64, has led to a general malaise toward the technology. While it's an inevitable transition, it's moving at a snail's pace and ultimately may not matter even after Windows Vista arrives.

More than that, AMD is currently the only producer of true mobile 64-bit chips, and their presence in the exploding mobile market is frankly just not that strong.

The transition to multi-core, on the other hand, seems to be going rather smoothly. Multi-core operation offers immediate performance benefits in the way of smoother computing, and a software base for it was pre-existing in the form of multi-threaded applications that could take advantage of multiple processors in the same computer.

You see how it goes.

The problem with this year is that the move towards multiple cores coupled with 64-bit technology is inevitable, and at the moment it's unwise to have one without the other. That makes buying a notebook right now a nightmarish task.

Which one would I choose, personally? Hard to say. But you'd better at least have one of those technologies.

THE KING IS DEAD, LONG LIVE THE KING: INTEL'S CORE DUO

Yonah in the First Half

The first half of the year is seeing the gradual introduction of Intel's Yonah dual-core processors, known to the consumer as Core Duo. Try to ignore Core Solo. I'm sure it offers fabulous battery life, but it's not worth it.

With the Core series, Intel moves away from the Pentium M naming scheme and their existing model number scheme. Dual-core Centrino platform notebooks will now be known as Centrino Duo.

The model numbers for the Core series are as follows:

• T26002.16 GHz
• T25002.00 GHz
• T24001.83 GHz
• T23001.66 GHz
• T1300 1.66 GHz*
• L24001.66 GHz
• L23001.50 GHz

T represents standard chips with a heat dissipation of 31W, up 4W over existing Dothan core chips; L represents lower voltage chips with a heat dissipation of 15W. It's a nice way of thumbing their nose at AMD by swapping the letter designations of their model numbers.

* The T1300 is a Core Solo with a heat dissipation of 27W, and should be shunned, mocked, and pantsed on the playground. It's a modern dinosaur.

With the Core Duo comes a new Intel chipset, and early reports are that not only does the new Centrino Duo platform have substantially improved performance over the old Sonoma platform, but the battery life also either meets or exceeds it.

Unfortunately, Yonah still lacks support for 64-bit technology. While that wouldn't necessarily be the death blow there, the knowledge of the next core coming makes the new Core Duo a difficult recommendation.

Merom in the Second Half

The Core Duo will see a new design in the second half of 2006 when it switches over to the new Merom core. The Merom chips will double the on-die cache of the Yonah chips in addition to bringing in 64-bit technology, thus making a complete, low power, 64-bit dual core mobile processor. Intel has even gone on record as saying they expect the power consumption of the Merom based Core Duo to be even lower than the Yonah based ones.

Excited? You should be. But it's this foreknowledge of the Merom chips that serves to really blunt enthusiasm for the current Core Duo processors, even though they're remarkable in and of themselves. Unfortunately, Merom is truly complete technology, while Yonah is more of a transitional stepping stone, and right now, we have Yonah.

ALWAYS OUTNUMBERED, NEVER OUTGUNNED: AMD'S DUAL CORE TURION

AMD hasn't released details in months, but thus far appear to have kept to their existing roadmaps. They're due to make an announcement soon, but here's the information we have so far.

AMD's roadmaps indicate that the Turion 64 will undergo a transition from being a single-channel, single core 64-bit architecture to a dual-channel, dual core 64-bit architecture, which I assume will be derivative of the upcoming Socket AM2 desktop Athlon 64 X2 processors just as the existing Turion 64 processors are derivative of Socket 754 desktop Athlon 64 processors.

What's key is that if AMD can deliver the dual-core Turion 64 before Intel gets Merom out the door, in this author's opinion, that would likely be the definitive choice for mobile processors, at least for a short while.

Unfortunately, AMD still doesn't have a true dedicated mobile solution the way Intel does, and Intel will likely continue to beat them up, down, and sideways with low power consumption and heat dissipation in the mobile sector. Remember that the Centrino platform isn't just the processor, it's the processor AND the chipset (yeah, and the wireless). Intel continues to produce potent, low-power chipsets to pair with their processors, but AMD doesn't have that. No one in AMD's camp seems to be really interested in producing a low-power chipset, and as a result notebooks with their processors are basically using desktop chipsets which consume a lot of power.

And that's a problem, but with any luck the problem will solve itself this year and help AMD gain a truly competitive edge in the notebook market.

Either way, dual-core Turion 64 is going to be the way to go if they can make it surface prior to Merom.

2006 IN GRAPHICS

The massive architectural and market changes in mobile processors are going to be paired not so much with change in the graphics sector as just progress, and in many ways, that's very exciting.

As I'm sure you've noticed, low-end parts don't linger in the mobile sector like they do on the desktop. While you can still go and buy a brand new Radeon 9200 for your desktop, your low end notebook is going to be running a HyperMemory X300 or a GeForce Go 6200 - both superior parts to the aging 9200.

In 2006 we're going to see a low end that rivals the existing mid-range, and this is only good news for mobile gamers. This will be the year where mobile gaming really matures - at least that's my prediction. I'll tell you why.

MOBILITY RADEON X1K SERIES

ATI is already releasing four Mobility Radeon parts for 2006, with more undoubtedly on the way. First up are the X1300 and X1400, both four-pipeline parts. The X1300 will have a low-end version running largely off of HyperMemory. The X1400 will be a higher clocked version of the X1300.

If you've seen the X1300 Pro's desktop performance, you should feel heartened. Our entry-level part is going to at least exceed the performance of a Mobility Radeon X600, which is currently really the bare minimum for solid gaming performance. X1400 is just gravy from there.

Next is the X1600 which has already been spoken about and reviewed on the forums here. The X1600 is a mighty twelve-pipeline part that has performance sitting between an X700 and a GeForce Go 6800. Another fairly impressive notebook part.

ATI also announced the Mobility Radeon X1800XT. This I can't be as forthcoming about, but word on the street is performance comparable to a desktop Radeon X1800XL.

All of these parts are Shader Model 3 ready and better, include ATI's AVIVO technology which allows the GPU to assist in decoding H.264 hi-definition video, along with substantially accelerating transcoding between video formats.

ATI is also planning on releasing a four-pipeline IGP later this year. While it hasn't been formally announced for notebooks, expect to see it. ATI claims performance comparable to an X600. Heck of a bare minimum isn't it?

ATI has also gone on record as saying they will be releasing a driver that allows the notebook to switch to an IGP part when running on battery to conserve power. If this comes to fruition, it will be a major boon for mobile gamers, as they will now be able to achieve the desired mobility when unplugged, while getting excellent performance while plugged in. Details of this technology are sketchy, but it has been announced, so look for it.

GEFORCE GO 7 SERIES

nVidia has been less forthcoming with their mobile line, and I think that's largely due to the slow rollout of their desktop 7 series. They have pages up on their site for their 7 series mobile parts, so we'll discuss them briefly. Unfortunately, though, because there is no existing information on their desktop counterparts, it's very difficult to gauge where these parts will wind up performance-wise.

In the low end, the GeForce Go 7300 and 7400 are the next generation analogues to the GeForce Go 6200 and 6400 respectively, and appear to inherit the same TurboCache shared memory architecture.

In the mid end is the GeForce Go 7600, which is an eight-pipeline part that will be competing against ATI's twelve-pipeline X1600.

In the high end is the GeForce Go 7800 and Go 7800 GTX. The GTX is on the market already and is by far the fastest notebook GPU available. nVidia really hit this one out of the park; it's pin-compatible with the Go 6800 and has the same thermal envelope, but is also a 24-pipeline part offering incredible mobile performance.

nVidia has admittedly never been much of a mobile GPU company, and that's been readily apparent in the last generation of chips. One of their major downfalls as of late has been a lack of an IGP in the desktop and mobile markets, only having just recently re-entered the desktop IGP market. Unfortunately, they have nothing to compete with ATI's Radeon Xpress 200M right now, and I expect that will continue through 2006 as well.

Still, nVidia has been providing enough of a challenge in the mobile world to at least keep ATI on its toes, but they don't have anything this time around that really competes with what AVIVO brings to the table in terms of pure features, and as I mentioned, they're sorely deficient in the IGP market, not even competing there at all.

PORTABLE MULTI-GPU

This is hazy, but expect it to surface this year in boutique notebooks. Chipsets are becoming available from both nVidia and ATI that enable SLI and Crossfire respectively in notebooks, and many manufacturers, like Clevo, are experimenting with enabling these technologies in notebooks. The problem, however, is largely with cooling two GPUs in a notebook. Still, the zany die-hard gamers with cash to burn are going to want to keep an eye out for this.

If nothing else, it's cool just to think about having a notebook running two GeForce Go 7800s in SLI, isn't it?

2006 IN OPTICAL DRIVES

HD-DVD, BLU-RAY, ETC.

Even now, the battle rages on. Toshiba demonstrated a notebook at CES with an integrated HD-DVD drive, but quite honestly, you're not going to want to worry about either of these technologies until they're done tearing one another apart. Media buffs waiting for these to materialize would do well not to bet on either camp.

Early impressions of these technologies have been disappointing, since all they offer above DVD are a higher resolution picture along with extremely consumer-unfriendly, prohibitive Digital Rights Management schemes. DVD became popular because it was a universally agreed upon standard; the schism between the Blu-Ray and HD-DVD camps may very well be the end of them both.

All that said, I can see whichever one ends up victorious (assuming one is victorious at all) leading a prosperous life in the mobile market. I don't know if you've ever watched a DVD on your notebook, but watching something encoded at 720x480 on a 1280x800 screen can leave a lot to be desired. While the average person's home TV has almost nothing to gain from either technology, his or her laptop will offer a somewhat crisper picture than it would with just a regular DVD.

Just some things to consider, but don't hold off on buying a notebook on account of either of these. Everyone is by and large happy with DVD right now and those of us that have invested in 200+ DVD collections are going to be fairly hesitant to adopt a new technology that offers so little real value to the consumer.

2006 IN HARD DISKS

PERPENDICULAR HARD DISKS

In the middle of 2005, Seagate demonstrated a 2.5" notebook hard drive that utilized a new method of storing data, called perpendicular recording. Without going into the esoteric details of the technology, it effectively doubles the usable space of a single hard disk platter. Platters in notebook drives are maxing out at 60GB right now, and at only two platters per notebook drive, we're stuck at 120GB while desktop users are enjoying having dual 500GB hard drives.

Seagate's perpendicular recording prototype drive stored 160GB - 80GB per platter. Keep in mind that greater data density on the drive contributes to improved read/write speeds, and perpendicular recording nearly doubles that density.

So what does all this mean to you? Perpendicular drives were demonstrated at CES this year, and we have every reason to believe they will begin surfacing on the market by the middle of the year. And because it's just a change in the internal technology of the drive, they will be compatible with your existing notebook.

2006 IN OPERATING SYSTEMS

THE INTEL MAC

This is a painfully complex subject, but an exciting one, too. I'll try and refine it down to its fundamental points. By now you've heard that Apple is transitioning to Intel chips, and in fact have MacBook Pro's based on Intel's Core Duo now available for sale on their site (shipping February, I'm told). To say the new MacBook Pro is sexy is the understatement of the year, and it immediately explains why Apple elected to make the switch to Intel chips from IBM's PowerPC chips. And it comes with a Mobility Radeon X1600 standard! (Of course, maybe it'd better at that price point.) So now you have a mobile dual core processor, coupled with a GPU capable of doing video transcoding in hardware, under what has long been the definitive multimedia platform. Oh, and unlike PCs (well, Windows), Macs have been designed with parallelism in mind for years.

Yeah, I'd jump ship, too.

But it's not all bread and roses. Apple has gone from a 64-bit chip, to a 32-bit dual core chip, and late this year they'll be going to a 64-bit dual core chip. We know this to be true because it's on Intel's roadmap and has been readily announced.

Also note that Rosetta technology - that's the technology that lets PowerPC apps run on the new Intel chips - is conditional. Like the emulation on Xbox360, some prior applications (and unfortunately, not old ones either) just won't work. Visit Apple's site for more information.

What it all amounts to is a massively transitory period for Apple and the Mac, and it's a period that I would strongly advise against jumping aboard in. The software written specifically for the Intel platform isn't quite ready to go yet. And Mac OS X is basically going from x64, to x32, and back to x64 again.

Be excited about the MacBook Pro and the transition. You have every right and reason to be. Hell, you know I am. But be patient and wait for it to finish moving through its growing pains.

Then we can all jump ship and be done with Windows.

WINDOWS VISTA: WINDOWS XP IN MAC OS X CLOTHING

Windows Vista is set to release near the end of 2006, and if I know you, I know you're keenly interested in making sure your hardware will fully support it.

Well, it won't. Deal with it.

Okay, that's a lie. By and large it will, but thanks to new Digital Rights Management "features" in Vista, some protected HD content may not play on your new machine. That's because Vista checks your hardware to ensure it's secure and certified, and if it's not, you're basically screwed. There're more irritating DRM "features" in Vista, but by and large this forced obsolescence of some hardware seems to be what rubs most users the wrong way.

There're also a lot of random, wild hardware requirements being thrown around for it and a lot of misinformation. This is rumor control: here are the facts.

You're going to be okay. The reality is that it requires about the same hardware Windows XP does. I know this, because the beta is running perfectly fine on most computers available today. The big kicker is the Aero Glass interface, which requires hardware acceleration by the system's GPU. For Aero Glass, the GMA 900 and lower users by and large won't be able to enjoy all of the eye candy. Your bare minimum is going to be a 64MB true DirectX 9 class GPU; you can refer to my GPU guide to see where you fit in, but basically it's Mobility Radeon X200M onward.

Worried about having a 64-bit processor? Don't be. It's been guaranteed there'll be a 32-bit version of Vista for those that want it; in fact, nearest I can tell the "bitness" of Vista is sort of foggy, but I'm sure a helpful forum member will fill in the blanks. Understand that 64-bit is the future, but it's not a prerequisite for enjoying Windows Vista.

Unfortunately, word of mouth regarding Vista right now is middling at best, and the consensus seems to be that it's little more than a glorified Windows XP. But it's not due out until late 2006 (think October/November), so there's still time for change.

CONCLUSION

If you haven't guessed by now, 2006 is going to be quite a year in mobile technology as multiple new standards come into place and new hardware enters the market and matures. I could speculate as to how things are going to go, but I'll bulletpoint my key ideas:

• Dual core, 64-bit is the future. But dual core is the more immediate future.
• AMD could very well be first to market with a 64-bit dual core notebook processor.
• If you're in no great rush to buy a notebook, wait until the second half of 2006 for Merom.
• It's going to be a banner year for mobile gamers, as after ATI's next IGP is released, the only notebooks that won't be worth gaming with will be ones with non-ATI IGP parts.
• We could very well all be using Macs at the end of the year, but let's wait until then to give the transition time to mature.
• Windows Vista isn't the sure thing Windows XP was. Plan accordingly and don't believe the hype.
• HD-DVD and Blu-Ray could very well die on the shelf.
• Bigger mobile hard disks! Still kinda slow, though.

Thank you, and here's to a great 2006; it's going to be an exciting year for notebook enthusiasts!

 

Best techie article I've read in a long time.

Even rivals some of those from ExtremeTech.

Amazing stuff.

I am very excited about the new ATI X1000 series. Thinking about upgrading my desktop to one (currently has a 9800XT - I'll upgrade the mobo to PCI-e to accomodate) and my laptop for sure - looking the AXIOM X1600 to come out so I can upgrade my X700. Not that my X700 is slow, just that I'd like to have the best gaming performance possible.

Once again, bravo Pulp.

 

Right on about the x1300 and x1400. In a review of a new Dual Core Acer model with an x1400 posted in NBF.. it's 3dmark05 score is around 1600-1700.. placing it above the x600.. and comparable to what my desktop Nvidia FX 5950 gets. Although I believe an x600 with 256mb dedicated, can get up to 1600.

 

I'm very interested in how the processor wars shape up this year, whether AMD will continue to gain ground on Intel -- if Dell will do anything with AMD in the mobile market (I don't think they will this year, but they might go with AMD for servers and maybe a mobile processor for end of this year or sometime next year).

I agree on the optical drive stuff, just stick with DVD technology and don't put any weight on a Blu Ray or HD-DVD drive -- waste of money IMHO and way too early to adopt either in a laptop, the mfrs are just trying to get hardware out there so they can say their technology has market share (Toshiba is backing HD-DVD and putting it in their Qosmio this year for instance, expect Sony to maybe do the same with Blu-Ray since they're the main backers of that technology).

I've also seen a couple of machines that'll be coming out in April I'm pretty excited about so outside of these internal technologies there's cool looking new hardware around the corner, keep an eye on the Dell XPS line

thanks for the great analysis and look at 2006 for notebooks!

 

One word, AMAZING! dugg!

 

Yeah, the media war is going to suck ass, besides they'll be short lived since holographic drives are already around the corner, just not for laptop. Those drives are still far from portable. But i don't see them keeping the blu-ray and hd-dvd alive once holographic drives come out, even if not on laptops. They should stick to dvds for now.

 

Very interesting article. I don't see either blu-ray or HD surviving in the end. I read that Toshiba is relying on their notebooks to push HD-DVD and Sony will be using the Playstation 3 as their engine. It will be interesting to watch the war.

 

Interesting article pulp.

 

Very interesting and very nice. Good job Pulp!

 

Does anyone know whether pcmcia will be phased out in the new platforms. I have some good (and costy) pcmcia cards that i don't want to become worthless.

 

amazing article pulp. i learned a lot.

price is another thing that should be considered. i'm pretty sure once Core Duo comes out on more computers, it's going to be expensive (at least hundreds of dollars more than a good notebook now). so, I do think it's good to wait for now cause once Core Duo is widely available, notebooks considered good now will be more affordable. Correct me if I'm wrong. although notebooks seen as good now will also be seen as slightly obsolete.

as for the new optical drives, they're just idiotic. it's the whole vhs/beta war all over. i understand why they need it, but it will be hard to convince normal consumers to adapt.

 

oh. forgot to add this.

whether you need or don't need a notebook now, do you think it's wise to buy one now?

 

If you need one, buy a Core Duo. If not, don't buy one until September.

Everyone's worried about price and doesn't understand: Core Duo is replacing the Pentium M. It's the next generation. I don't expect it to cost any more than existing Centrino notebooks do.

 

Well said, sir. When're they gonna git them thar fancy wrist watches - the ones with the television inside'm?

 

April?!? crap....I've been waiting to see what Dell would do with the XPS line expecting something in February. If I waited until April then I might as wait for the 64-bit systems/OS in the fall.

I could always go cheap now then expensive later.

 

so what are the chances of microsoft going back on its plans to include all of this secrity crap in vista, when no one buys it?

 

This was a very interesting article, a great read! I'm really curious as to what AMD is going to release next and when. AMD has been quiet about their upcomming processors while Intel has been screaming and kicking about theirs. It kind of makes me wonder if AMD some some ace up its sleeve.

 

so even if i dont need a dual core, just because the prices are bound to go down, i should at least wait til the dual cores come out right? (btw i'm waiting for the tablets' prices to go down.. they just seem to stay the same for the longest time)

and any opinion on this?
http://www.newscientist.com/channel/info-tech/dn8370

 

Thanks for the good article. I am convinced to wait for Meron to come out, before I buy another notebook.
Maybe the descrete graphics cards will migrate down to the 13" and 12" laptops.

 

FYI:

http://www.dailytech.com/article.aspx?newsid=283

today Seagate becomes the first manufacturer to announce an entire line of perpendicular hard drives for notebooks. Called the Momentus series, Seagate is manufacturing them in 2.5" form factor, which is standard notebook size. What's not so standard however is that the Momentus series will be introduced with a flagship unit sporting a capacity of 160GB -- the largest capacity for notebook computers ever.
...
Currently, perpendicular technology will only be offered in Seagate's Momentus 5400.3 160GB model, but the company plans to transition all models and release drives ranging from 40GB to 160GB capacities in 20GB increments by the end of March. Seagate also indicates that by June, it will have perpendicular SATA notebook drives as well. Although Seagate did not specifically indicate when desktop drives will make the transition, it did say that we can expect perpendicular desktop drives to arrive later in 2006.

 

Thanks for the article. Very nice job of condensing and addressing alot of FAQ that have been floating around these forums. I have to say that I agree with your recommendation of waiting a bit were Apple's new architecture is concerned. Not so much due to due legacy software compatibility, but more due to the fact that personal experience has shown me that it's always best policy to wait for a Rev 2 of anything new Apple releases, (as with most vendor's new hardware) the iBook G4 being a good example. On the software side I believe the picture may actually be a little better than anticipated. The majority of Apple's own core apps run natively already (iLife, Safari, Mail etc.). Unfortunately their more expensive pro apps which have been converted to intel (Final Cut, Motion, Soundtrack, DVD Studio Pro, Aperture, Logic, and Shake) will require the user to shell out a cross grade fee. A decent number of the most relevant third party vendors that matter most to Mac users seem to be either on track or ahead of schedule, like Quark, Adobe and MS Office (of which the latter two incidentally appear to run fine under Rosetta). Out of necessity, due to a dying iBook G3, my wife will be purchasing a MacBook Pro in the next couple of months. I hope to be able to run it through it's paces and have something resembling a review in March.

 

I understand that Core Duo is replacing the Pentium M. But what I'm pointing out is the price of current Pentium M. Wouldn't they lower in price due to Core Duo and therefore, lower the price of current notebooks that has Pentium M?

 

Nice article!!!! I've changed my mind. I don't enter to the school till august so I'll wait till I got more options to chose from (not only core duo on acer laptops). Thanks

 

This is a very nice article, it hit the spot for me

Two cents:

First, I totally agree with Pulp that Mac fans (or would be converts) should wait till Apple releases dual core 64 bit powerbooks (MacBooks if you prefer). I think their OS and software bundle will be much much more robust under a 64bit architecture.

Second, I am not quite so sure whether Windows user's will get all the benefit of a 64bit architecture within the time frame of 2006. It's true that there is Xp 64 and Vista is knocking on the door, but what about other windows applications which are currently running on native 32bit platforms and have gone through several revisions. Will a user a see a distinct performance difference ?? It is nice to have the latest and greatest, but I think a 64bit Windows LAPTOP will be an overkill, atleast till more applications mature.

 

Super well written ,

Of course I knew most of it

But seriously it confirms my suspicion that ATI has no real answear to Nvidia as the x1600 doesn't match even the old 6800 ( which is the minimum to run bf2 in my opinion ) not to mention the 7800 .

It would be nice to add an article comparing duo core performence compared to old P-M so we will know what gain is there .

 

i kinda like xp64 i think it is faster, and 64 bit apps are starting to appear, like counter strike and half life jus got a 64 bit upgrade, so geting a 64bit laptop now isnt such a bad idea especially if you want your laptop to last 2 to 3 years.

 

"
Oh, and unlike PCs (well, Windows), Macs have been designed with parallelism in mind for years.
"

I don't quite understand that statement. MacOS was the last OS to dump the "concurrent multitasking" thingy well after Windows did. In fact, that transition was done in 2002, so not so long ago.

So to say that Macs were designed for parallelism for years is not quite true. Windows NT, with the first version going out in 1993, has always been design with parallelism in mind since the very begining.

 

Pulp,

Great article !

I have few questions:

-Do you think GMA950 will be sufficient for the Aero Glass interface ?
What is missing in the GMA900 to qualify for Vista's Aero Glass ?

-Also, why is 64 bits important ? The way I see it, 64 bits
give you more addressable memory, so IF your application
needs more than 4 GBytes of RAM (maximum address space
of 32 bits), then you'll benefit 64 bits. I can see this on our
larger Oracle server we have here at work. However most
PC's and Notebook's motherboards do not even support
4 GBytes of physical memory, and yet end user's application
run just fine. So why the rush to 64 bits ?

Keep up the excellent work !

 

ATI cant compete with nvidia when it comes to performance but ATI cards are still found in the majority of laptops because they consume much less power and produce less heat. The ability to put a card comparable to the 6800 in a 14'' or smaller laptop is what makes ATI and the x1600 so great. The smallest laptop with a 6800 is the 17'' dell 9300 while there are a couple 15.4'' laptops with the x1600 and I think there will be some smaller ones in the near future.

 

The X1600 isn't designed to compete with the Go6800 - it's supposed to be ATI's new mid-range offering, which will replace the current Radeon X700.

The X1600 is quite an impressive card for notebooks from what I'm seeing - over 3,800 in 3DMark05 (and that's only the 128MB version), plus very low power consumption, potentially making it usable in 13" notebooks.

 

Sorry to insist... but any plan on MXM/Axiom for dell,toshiba,hp,acer...... (sager and asus not plz, they arent available here in mexico)

 

I don't think that mainstream manufacturers will have those new technologies - they don't want the consumer messing around with the internals of the laptop, and on the other hand, they would rather have you buy a new laptop rather than upgrade the graphics card.

 

very interesting! i read it all nice article.

 

Man, this was a great read. It cleared a few things up for me. This was straight and to the point, like the rest of your articles.

1) GMA950 as I understand it is just an enhanced GMA900 with a max of 224 MB of VRAM that it can take from the system RAM while GMA900 can only take 128 MB of RAM. So if the bare minimum to run Vista with the Aero Glass effects is the ATI X200, then the GMA950 will struggle if it would be able to run at all with the Aero Glass Effect in Vista.

2) You are absolutely right, since there is not enough addressable memory, the complete 64-bit power will not be utilized. I think the push for 64-bit is more of a push for more people to utilize and evolve technology into the 64-bit realm. The first 64-bit processors came years before the first release of a 64-bit Windows OS, and even then most people didn't feel the need to use the 64-bit Windows. If more of a push were made for more people to adopt a 64-bit Windows in conjunction with a 64-bit processor, a famous one, then the evolution of technology into the 64-realm would be able to begin. Larger addressable bandwith memory, right now, really only benefits the people who would need it and , such as graphic designers, gamers, server systems admins, and other people involved who would need the large space to be able to calculate large projects. However, if more of a push were made to make this mainstream then even more people could benefit the 64-bit and create a need for developers to create more software that was more in tune to 64-bits rather than 32-bits.

 

now when you say second half doesnt that mean 6 months from now
because that is a long time and would i reall be worth it?

 

On the 64-bit side of things, it isn't just about addressable memory.

There's also the more fundamental side of things - you can do calculations using 64-bit numbers which are much larger than 32-bit numbers. For numbers greater than 32-bit and less than 64-bit a 64-bit processor will be quicker because it will be able to treat the calc as a single operation rather than multiple operations with a 32-bit processor. It's not just the size of the numbers either, it is the precision (in the case of floating point) - here the 64-bit processor running a 64-bit program will be able to give you either greater precision than a 32-bit system, or give you faster results (because a 32-bit system will have to do more work to do the same calc).

 

Kudos Dustin Sklavos,

Well written and organized article. Bit long on hyperbole for my taste, but if that's your style, so be it.

You left out a huge detail in all your facts, 65nm. Intel has it, Yonah is produced on it, Merom will be produced on 65nm and AMD won't have 65nm for roughly 12 months. Neither I nor you no exactly when.

Why is 65nm huge? Two reasons: cost of production and power consumption. So even if T64 X2 shows up in a couple of months, it will be a 90nm power hog relative to Yonah/Merom until AMD can transition their fabs to 65nm. That and for the equivalent transistor count [CPU or cache] AMD will have much larger die sizes, which as you well know, dramatically increases AMD's unit cost. Its all about profit so you can build those multi-billion dollar fabs.

One small area of disagreement with one of your conclusions about Yonah. Your basic premise is that Yonah is but a stepping stone on the way to 64bit Merom. It is my firm belief the the two will coexist for a couple of years. Why? Cost!!!

As you correctly stated, Yonah is no lightweight or insignificant CPU. Since the combination of the 64bit functions and the doubling of the cache, the Merom die will necessarily be ~50% larger than Yonah, since both will be on 65nm. This will effectively double Intel's cost to produce Merom, its not linear. Therefore, Yonah will have a rightful place as a lower cost lappy CPU. You or I may may not own Yonah, but the bulk of the sales will go to Yonah.

BTW, early feedback on Merom is that its running significantly higher power than expected. Article at either Geeks or Anandtech.

Again thanks for your article, we all benefit from the knowledge and discussion.

Hermit

 

even if amd has a larger overhead on the unit cost i bet they still wont be as expensive as intels cpus, they just love to make as much profit as possible (like microsoft).

 

Thats exactly my point

I'd love to see a small form gaming notebook like the joybook 7000 that for its time was superb option in a small stylish package .

However the x1600 is no match to the 7800 gtx or even go version and since i dont see any 1800 offering that means that ATI has lost in the gaming section for now .

On a note I dont consider myself heavy gamer or one that must have the latest hardware but I will not buy a laptop with a relative weak card knowing that games demands will turn it useless in a very short while .

When I bought the i9300 with 6800 less then a year ago I was sure it will handle all game with full settings at ease into a year or two , I was wrong .

Games demands have grown much faster then I expected and the 6800 can barely run BF2 medium settings with no shadow , lights or any aa af , so I expect the 7800 will be the bar setter for games demands , anything below just wouldnt do .

 

It's doubtful that it will matter until a *very* long time after Vista. We are far from needing to access that much memory in a laptop (or even being able to put it in a laptop), and there is no reason for software not to fallback gracefully to work on 32-bit - I think you can go at least 3 years before you get software that won't run (acceptably) on a 32-bit processor. Although it may depend on precisely what you use it for.

Simple - out of the two, dual core... unless 64-bit only software arrives, the dual core will give you greater use and performance throughout it's entire life than a single core 64-bit processor (at the same clock speed).

There is - and may well continue to be - a significant problem. The cost. The MacBookPro is hugely expensive compared to a PC laptop of the same spec (bit of trouble finding something to match, but from evidence that exists, you are paying something like a 50% premium - $800 - for the MBP).

That's a whole lot of cash, and yes, it's better engineered than the cheap PC, but is it enough? And add to that the cost of transitioning software. Ouch.

Even OS X will lose a lot of it's benefits when Vista comes out - the proposed security and Aero interface (on supported hardware) will really close the gap. There are also other practical problems with choice on OS X - take for example iSync... great little utility, but the device support is woeful, and Apple have been terrible at updating it.

I own two Macs, use them constantly and enjoy them - but between lack of choice, expensive upgrades, and diminishing benefits, I just don't see any reason to remain on OS X. The Apple apps are great, but it's not like there isn't other choices on Windows...

Now, Apple could seriously help their business out with some gutsy moves. The hardware could be made competitive with PC hardware - by selling it on a equal basis. Actively sell it to the Windows/Linux crowd, and take a slightly reduce margin but much larger market.

Then, with the apps not needing to sell the OS/hardware, you can take one of two decisions - the most profitable would be to dump OS X, and just sell the software on Windows (so Apple would be a hardware and software company - just without the bit in the middle).

Or, free OS X - let it run on any hardware, maybe even make greater utilisation of open source to keep costs down and sell it cheap. Your hardware business is still intact from virtue of selling to a huge market. Your OS share increases due to not forcing people to buy your limited choice of hardware, and app sales increase thanks to larger install base. (Note that this is distinctly different from the earlier licensing scenario - remember that then Apple could only sell hardware to people that wanted to run their OS, hence why third party manufacturers took away their sales. By selling to the entire Windows/Linux market, they can actually increase their hardware sales regardless of whether their OS base increases).

But can Apple be that bold?

Doesn't matter unless you are buying HD-DVD or Blu-Ray, which you won't for some time.

I think it *should* be. The security improvements and additional apps (calendar, sync, etc.) - all the things that Mac users take for granted - ought to make it a big deal. No, it may not change your 'everyday when things are working well' experience significantly, but your world will be a happier, more comfortable, and friendly place. As someone that has been using OS X for 3 years can testify, that should mean a great deal.

 

whoaa..
the whole article,
and the last post
just totally blew my mind
i have some soulsearching to do..

anybody can guess how long the current duo core (32 bit)can last us? what i infer is that it'll be the mainstream choice for sometime, even after the 64 bit duo core comes up. .. septmber is an awfully long time away..

anw.. thanks for the info flood .. woohoo *high*

 

I hate computers.

 

2008. 17" macbook pro descendant. multi-core 64-bit. 4+GB Ram. OSXI. goodbye windows.
now i just need to accept dropping 10K for it and the complete software switch

hmmmmmm. maybe it's time to switch to linux and full-on open source after all...

most excellent article BTW

 

Sticky FTW!!!!

 

Okay, here's a question that's been bugging me for some time now:what exactly is the aero glass interface, and what is so special about it?

 

I can't fathom the "wait for Merom" thinking. INTEL has been steadily losing marketshare to AMD in all market segments. Why? Price and performance. So INTEL has taken lately to blowing their horn about products that don't yet exist. Turions are cheaper for the same performance and have 64 bit already without much of a performance penalty. And if you get the thin and light Turion, you get great battery life.

64 bits is important; I'm old enough to have discounted the move from 8 to 16 bits in the CP/M computer days when IBM announced the first PC and we all later wondered when 32 bit software would flourish. Dual core is somewhat important, but only for running multiple apps simultaneously. Most won't see that much of a difference in either for alot of years.

This all said, you can get great deals on computers during the period of change. If one needs a notebook for general needs, the $499 special is 1/4 the cost of the bleeding edge computer. 3 years from now, a dual/64 bit will be the standard on low cost computers. Buy a cheapy Sempron or Celeron notebook and spend the excess on a high performance desktop if you need to do gaming or high resolution 3D processing or want to show off to your friends. I've chased the "gaming" do it all notebook market before; I finally bought a cheap desktop with upgradeable video and motherboards to keep up with performance needs.

My recommendation -- if money matters, buy cheap notebooks. They are a bargain and last just as long as the high end ones.

Rich

 

Great Article Pulp, well done again. It's clear and simple and has some very interesting facts. The fact that Vista uses so much system RAM as well as video RAM may be a bit of an annoyance though. However, I do think the Aero glass GUI will look very nice.

The Graphics and Processor market is moving at an alarming rate, that you need to buy a notebook at just the correct time.

I've one question though. What exactly is the difference between 32 bit and 64 bit processors. Do 32 bit CPU's work with 32 bits oif information per clock cycle, and 64 bit CPU's work with 64 bits of info per clock cycle? If this is the case, are 64 bit CPU's twice as fast?

 

Excellent article! By no means irrelevant to today's notebook buyers' concerns, even though you wrote it in January! Good work & thanks for the very helpful information.

 

No, they both perform the same...unless it was written for 64 bit hardware which almost no consumer programs are at the moment. even when there will be, the difference is quite minute, nothing like "double the speed".

And I just want to comment on the original post...turns out Merom isn't the "complete" solution after all...as TDP on battery power suffers from the new merom....turns out it is faster...but at a cost of higher temperatures....not what Intel had claimed to be even cooler than its predecessor.