Or, even worse than a delay: New MacBooks could have worse graphic than older MacBooks.

Here's the story: Starting with the unibody MacBook Pros in Oct. 2008, Apple dumped Intel's own chipset and integrated graphics for Nvidia's GeForce 9400M as a combined GPU/chipset, since it wildly outperforms the Intel's integrated garbage, which had hampered previous MacBooks. Since the 9400M is in basically every Mac now, there's a baseline of graphics performance across every Mac—nothing has crappier graphics than the 9400M. Important, because the OpenCL tech in Snow Leopard leverages your graphics card for extra processing power.

Since Oct. 2008, Intel's introduced its blazing fast Core i7 and i5 processors, which use the Nehalem microarchitecture. The problem is that Nvidia can't make compatible chipsets for it. Intel claims that Nvidia's license to make chipsets for its processors doesn't apply to any current or future processor with an integrated memory controller, which all Nehalem and Westmere—the 32nm die shrink of Nehalem—processors do. Nvidia sued and is pulling out of chipsets entirely, at least at the desktop level. (Intel's also cut them off at the Atom level, making what the Ion 2 will look like something of a mystery as well.)

Which produces a question: What are the next set of MacBook guts going to be? The Arrandale Core i5 mobile processors Intel is expected to announce at CES don't just have integrated memory controllers, they have integrated graphics, built right onto the die, too. If the MacBooks were upgraded to off-the-shelf Arrandale processors, it can't, on the face of it, use an Nvidia chipset or more to the point, Nvidia's superior integrated graphics. Intel's integrated graphics still suck. So there are a couple of possibilities from here, it looks like.

• Possibility 1: Some kind of discrete or separate graphics cards for all MacBook Pros. Pre-unibody MacBook Pros, and even the 12-inch PowerBook G4, had discrete graphics cards only. The problem is that it's more expensive, and that now-famed 6-8 hour battery life would take a hit. It's how the latest iMac got away with using a Core i7 on with an Intel chipset, though.

Or maybe Apple will put discrete graphics cards in every MacBook Pro, but use Intel integrated graphics as a battery-saving fallback. Which is sort of the way all but the low-end MacBook Pros work now, with both integrated and discrete graphics. (Though the Nvidia integrated graphics are good enough to be the default option on current MacBooks.) It would rock the boat the least.

• Possibility 2: Suffer the crappier graphics on lower end models. A problem, given that any machine using Intel integrated graphics would result in worse graphics performance than the current MacBook or MacBook Pros. Which sounds counterproductive, given Apple's obvious bet on graphics cards for processing juice with OpenCL.

• Possibility 3: A customized set of hardware of some kind from Intel, either on the processor or chipset level that would let the next MacBooks match the power consumption and graphics capabilities of current models. It wouldn't be unprecedented: Apple asked for and received essentially custom chips from Intel before, for the MacBook Air. (Though Intel later let everybody else play ball with other chips meant for really skinny laptops.)

• Possibility 4: Apple's gonna wait on something else before upgrading from Core 2 Duos. Will people have to wait longer for blazing new silicon in MacBooks than in PC notebooks? Sometimes they do, yes, but sometimes Apple gets Intel's latest first—Nehalem Xeons in Mac Pros, and the ultramobile chip in the MacBook Air.

Something else to consider is that for the first time in a long time, if Apple wants to push new guts soon, it could switch to ATI graphics (which it's using in the iMac) for notebooks because of delays in Nvidia's Fermi architecture that push their truly new graphics cards out until Spring 2010. ATI's got a solid 4 months where it's got the newest graphics silicon around.

Whatever happens, it's a mystery for now. Which is kind of a fascinating point, actually, given that Macs run on PC guts now, yet it's still trying to do something different on the hardware level.

Remember all the excitement about Penryn a couple years ago, which took the Core 2 and made it more efficient with a new manufacturing process? It's the same deal here, as the tick of Intel's tick-tock cycle. "Tock" is a whole new microarchitecture, while "tick" is a die shrink of that, which makes it more power efficient. Nehalem is the tock—it was 45 nanometers—and Westmere is the tick, shrunk to 32nm.

Arrandale is what this set of mobile Core i5 chips, based on Westmere, is called. (Here's our primer on Intel codenames.) One thing in particular about Arrandale is that it has a graphics core built right onto the main chip package, which Intel says is good to go for Blu-ray.

Anyways, what all this means is that there's about to be a whole bunch of new laptops with faster, better Intel chips inside that won't munch your battery as hard. [Cnet]

Update: For now, IBM's just saying that the successor to the current PowerXCell 8i is canned, not all Cell development. [Fudzilla via MaxConsole, Driver Heaven]

The key feature about Arrandale, versus current Core i5/i7 processors, is that it's manufactured using a 32nm process, meaning it'll be less power hungry. Remember the jump to the Penryn Core 2 chips a couple years ago? Same speeds, better efficiency? Like that. The three coming out in the first half of 2010 are the Core i7-640UM (1.2GHz), Core i7-620UM (1.06GHz) and Core i5-520UM, all for "ultra thin" laptops.

We'll also see some 32nm chips for the desktop, codenamed Clarksdale, announced in January. Mmm, chips. [Digitimes]

Bobcat's a low power chip at Atom, like for netbooks, that can run using less than 1W of power. It'll be manufactured using a 32nm process. It's not hitting until 2011 though, and as Ars' Jon Stokes points out, by then Intel will have been at 32nm with Atom for a while, and already close to going to 22nm.

Bulldozer is AMD's new server architecture, also, slated for 2011 which uses a new design with two "tightly linked cores" as the heart. You can read about it in way more detail over at Ars, though for now, Stokes says it's still hard to tell how competitive they're going to be with Intel's 2011 lineup. Guess we'll see. [Ars, Image via Michele Eve/Flickr]

The 40-nanometer TILE-Gx will reportedly draw about 55W of power at full load, and though it's expected to cost between $400 and $1000 (depending on volume), it's more intended for use on Linux-based enterprise Web servers.

According to Wired, the chip isn't really geared for regular operating systems such as Windows 7—for that, you'll have to wait for Intel's 80-core processor which was demonstrated last year, and is expected in about 5 years. [Tilera via Wired]

Chips, Chipsets and Damned Chipsets

Okay, so the first thing to understand is that an Intel brand, like Core 2 or Core i7, actually refers to a whole bunch of different processors. Although they generally have the same basic microarchitecture (in other words, chip design), the brand envelopes both desktop and mobile chips, chips with radically different clock speeds, that use different motherboard sockets, etc.

Because of these differences, each particular chip is given a codename, chosen for obscure geographical locations (seriously, plug just about any codename into Google Maps). For instance, the original mobile Core 2 Duo processor was Merom, and it was replaced after about two years by Penryn, which was manufactured using a new 45-nanometer process to be more efficient. Quite different, these two, but Intel pimped both as Core 2 Duos nonetheless.

View Intel in a larger map
Although Intel doesn't market chips according to their codenames, the individual chip gets a model number that gives you an idea of how it compares, spec-wise (clock speed, cache size, etc.), to other chips in the same group. So, a Core i7-950 is gonna be faster than a Core i7-920, and a Core 2 Duo P8600 isn't going to quite stack up to a Core 2 Duo P9600. The difference between a P8400 and P8600 is obviously less than the difference between a P8600 and a P9600. To match a particular chip codename to a particular model number, though, you probably have to do some Googlin' (or Bingin').

In some cases, Intel pushes chips with a ULV designator for "ultra-low voltage," which doesn't mean anything in particular in terms of chip design, since it includes several brands of chips, from Core 2 to Celeron. The point is that these chips power notebooks that are almost as portable at netbooks, but are more expensive, so computer makers (and Intel) make more money.

While we're at it, I might as well explain what the hell Centrino is. It's not a single chip, it's a platform. That is, it's a combo meal for notebooks with a mobile processor, a chipset (essentially the silicon that lets the processor talk to the rest of the computer) and a wireless networking adapter. Typically, Intel releases a new combo meal every year, though they're all been called Centrino, with the most recent making the leap to being called Centrino 2.

The reason we decided to tell you all this stuff now is that Intel is gradually phasing out the Core 2 family, like Pentiums before that, and is moving Core i7, Core i5 and Core i3 up to take its place. This is how all the families relate to each other...

Nehalem Rising: Core i7, Core i5 and Core i3

Core i7 systems use a totally new microarchitecture called Nehalem, and it's badass.

The first set of Core i7 chips, codenamed Bloomfield, launched in November 2008 for high-end desktops. They're the most outrageously fast Core i7 chips, with triple-channel memory (meaning they're able to use memory sticks in triplets rather than pairs) and other blazing accoutrements.

The new Core i7 chips, launched last month, are for desktop and mobile. The desktop variant is codenamed Lynnfield, and it more closely resembles its mobile equivalent, codenamed Clarksfield, than it does the Bloomfield monster—dual-channel memory, not triple, for instance.

You'll be seeing a lot more Clarksfield in the next couple weeks, like in the HP Envy 15, since most computer makers were holding off for Windows 7 to drop their new laptops. All of the Core i7 processors are quad-core, even the mobile Clarksfield, so you're not gonna see it in anything like Dell's skinny Adamo.

Core i5 is going to be Intel's more mainstream Nehalem-microarchitecture chip brand, and as a broader brand, the chip differentiation gets a little more confusing. Core i5 actually includes some, but not all, of the desktop Lynnfield processors. For now, the only Core i5 chip is quad-core, but you're going to start seeing dual-core Core i5 chips, and soon enough they will make up the bulk of Intel's mainstream processors. In English: Unless you're looking for a crazyfast new computer, your next machine will probably run an Intel Core i5 CPU.

Eventually, dual-core Core i3 chips will come out, and as you can guess by the number, they won't be quite as fast—or expensive—as the Core i5 or i7 chips.

Netbook's Best Friend: Atom N and Z

Atom is probably the Intel chip you hear about second only to Core 2 Duo: It's essentially the CPU that goes inside of netbooks. There are a couple of different variations out now, the N series (codename Diamondville) and the Z series (codename Silverthorne). The Diamondville chips are for nettops and netbooks (though as pointed out, nettop don't use the N prefix, just the chip number), and can handle full versions of Windows Vista and 7. Silverthrone is used in netbooks but was designed for smaller connected devices like UMPCs and MIDs. (This is why Sony shoving an underpowered Atom Z in the Vaio P, and trying to run Windows Vista on top of it, was retarded.)

The next generation of Atom is more interesting, and more confusing, in a way. The CPU is codenamed Pineview, and it's actually got the graphics processor integrated right onto the same chip, precluding the need for a separate GPU tucked into the netbook's overall chipset. The benefit is longer battery life, since it'll take less energy to crunch the same visuals. We'll start seeing Pineview netbooks sometime early next year, most likely.

Oldies But Goodies: Core 2 Duo, Quad and Extreme

Intel's Core 2 chips have been out three years now, an eternity in computer years. Because of this, and because they're the main ones used in most personal desktop and laptop systems, there is a metric shitton of different Core 2 chips.

It's also more confusing because there are way more codenames to wade through. Let's start from the top: Core 2 Solo has one core, Core 2 Duo two, and Quad has four (as does Extreme). From there, you have two distinct generations of chips within the Core 2 family.

In the first generation of Core 2 Duos, the main desktop chip was Conroe (with a cheaper variant called Allendale), while the main mobile one was called Merom. There was also a branch of Core 2 Quads called Kentsfield.

The next generation (that is, the current generation, unless you're already on the Core i7 bandwagon) arrived with a new process for making chips with even smaller transistors. Among other more technical differences, they were more energy efficient than their predecessors. With this generation of Core 2s, the mainstream desktop chips are Wolfdale, the desktop quad-cores are called Yorkfield, and the mobile chips are Penryn—if you've bought a decent notebook in the last two years, it's probably got a Penryn Core 2 inside of it.

Ancient History: Pentium and Celeron

Pentium is dead, except it's not, living on as a zombie brand for chips that aren't as good as Core chips, but aren't as crappy as Intel's low-end Celeron processors. If you see a machine with a sticker for Pentium or Celeron, run.

Okay, I hope that helps, at least a little—you should probably thank me for staying away from clock speeds and other small variations, like individual permutations of Core i7 Bloomfield processors, to hopefully give you a broader overview of what all's going on. Intel told me it'll all make more sense once their entire road map for the year is out on the market, but I have a feeling it's not gonna help my mom understand this crap one bit better.

Top image via soleiletoile/Flickr

Thanks to Intel for helping us sort all this out!

Still something you wanna know? Send questions about sweet potato chips, pumpkin pie or turduckens to tips@gizmodo.com, with "Giz Explains" in the subject line.

Two ARM cores, one running at 600MHz (for apps) and another at 400MHz, with a 200MHz 3D graphics core that supports OpenGL 2.0 (like the iPhone 3GS, which is actually an advtange over the Pre).

It's hard to know how much different it's gonna feel like in practice versus the Pre until we got our hands on a final unit—there's plenty of time for optimizations and other plumbing work this far out. [Palm InfoCenter]

Now that is (or was) expected to halt at some point, meaning our computers and gadgets would start remaining the same size. Yep, no more "Honey, I Shrunk the Processors." But work by IBM could keep the sequels coming.

If you didn't know Intel's processors and transistors are about to hit 32 nanometers in size (fun fact: a single hair strand is roughly 80,000 nanometers in width). Now that is pretty darn small, but if we want things to get even smaller, like Zoolander phone small, it is said that the physical constraints in the silicon in these transistors can only go so tiny. Apparently, they have even been playing tricks with the silicon even since 90 nanometers.

The New York Times reports some seriously advanced solutions which are being worked on by Dr. Ross of IBM (not George Clooney's ER character who had the same name). FinFETs are one type of transistor and are the basis of 22-nanometer technology which we may see by 2012. These transistors are vertically tipped, offer greater density and better insulating properties. She is also concentrating on constructing FinFET and silicon nanowire switches in a whole new process.

It is a kind of nanofarming. Dr. Ross sprinkles gold particles as small as 10 nanometers in diameter on a substrate and then suffuses them in a silicon gas at a temperature of about 1,100 degrees Fahrenheit. This causes the particles to become "supersaturated" with silicon from the gas, which will then precipitate into a solid, forming a wire that grows vertically.

Complicated and extremely intricate stuff which is all apparently riddled with significant challenges, but Dr. Ross and her IBM team have got to keep at it. It means the continuation of us getting thinner and smaller electronics in our hands (and lost in my bag). [The New York Times]

The spaser works about a thousand times faster than the fastest transistor, while having the same nanoscale size. This opens up the possibility to build ultrafast amplifiers, logic elements, and microprocessors working about a thousand times faster than conventional silicon-based microprocessors.

This new method treats light in a different way than traditional optical CPUs, which are "difficult to miniaturize because you can't contain photons in areas smaller than half their wavelenght." In other words: Optical CPUs won't fit in current electronics. Plasmonic devices, on the other side, can concentrate these nanoparticles in spaces similar to current processors. [Technology Review]

Following the original 8086, it cost $100 when it was released in June 1979, which is about $300, adjusted for inflation. Today, $300 will buy you a Core i7 processor with 731 million transistors. How much more powerful is that than the 8088?

Immeasurably. Even Intel couldn't tell us. Meaning if you took a Core i7 back in time to 1979, and Miles Bennett Dyson was an Intel employee, he would fuck the human race in ways you can't even imagine. Or, you know, we just would've had Xbox in 1983.

With the ascendance of Windows, the x86-based PC would eventually take over the world in its own way.

[Intel, Top image: Wikipedia, Thanks to Intel for their help!]

Gizmodo '79 is a week-long celebration of gadgets and geekdom 30 years ago, as the analog age gave way to the digital, and most of our favorite toys were just being born.

The tech is applied to things like USB ports, which in 3.0, will go from polling (clock based, always checking) devices to being managed via events, so they can sleep whenever not being used. And graphics, when the page isn't changing, can be run out of a frame buffer so the GPU and video RAM can sleep. When I say more sleep, I mean for additional milliseconds or longer. This adds up, over the course of a day when people stop to read or step away from their computers. In the past, the OS controlled the power savings, and that required power to process in turn, so you were using the system's power to manage power, keeping those other components from ever really turning off. By doing power management with more granularity, in hardware and software together, you can switching things on/off fast enough to fit in lots of "naps" and you can also do it with less processing overhead.

I'm excited for this tech to go everywhere where there's a chip.

Though the two phones share the same CPU and GPU, they run on different chipsets (iPhone: Samsung SoC S5PC100, Pre: TI OMAP 3430). However, this Samsung chipset is also capable of handling 720p video recording, streaming vid conferencing and even Dolby 5.1 audio processing, giving this hardware some serious future potential.

AnandTech also posted an interesting theory regarding this particular chipset and battery life last week after the 3GS announce. They noticed that the CortexA8 processor can use up to 3x more power than the previous ARM 11 in the first two iPhones, yet battery life managed to improve with the 3GS. At the time, they thought the hardware might be using a process similar to Intel's Quick Start technology, which saves power by quickly ramping the processor up to full power when in use, and then dropping it down to an idle state when not in use actually averages out saves more power than running it at more constant power levels.

This might explain the improved battery life. Then again, the increased-capacity battery (1219 mAh, compared to the previous 1150 mAh), might also have something to do with it. [AnandTech and Rapid Repair via Engadget]