A timely platform primer

On the surface of it pre-defined PC platforms such as Intel's ‘V8'  and AMD's recently launched ‘Spider' are a great idea, as you can be sure that they've been extensively tested and that the CPU, chipset and graphics card will play nicely with each other. But there are still plenty of things the budding power user should know...

  • E-Mail
By  Jason Saundalkar Published  November 19, 2007

Intel's V8- and AMD's Spider platforms are two very different beasts. The first is aimed at those interested strictly in content generation and editing work. Thus it offers support for two multi-core CPUs, which will speed up encoding jobs as well as photo- and video-editing. (Note: Intel's own game-friendly platform, codenamed Skulltrail, should be out by quarter one of 2008.)

AMD's platform however primarily targets gamers and thus offers multi-GPU support, built-in automatic overclocking features and more (read the January 2008 issue of Windows Middle East - on-sale December - for our full Spider report).

Like pretty much everything in computer-land however, things are never quite as simple as two plus two. You see, while the components that make up the platform may run perfectly with each other, there's no saying how they'll react to other components, such as memory (RAM), add-on cards such as RAID or SCSI controllers, and their respective drivers.

So, point number one to understand is not to expect absolutely everything to go well, component-wise, just because you've based your new system on a combined technology platform. (It never hurts to do a fair amount of research - on ITP.net for example - to see what publications such as Windows Middle East are saying about a particular component or technology.)

Another point to consider with platforms, particularly from the perspective of the tech-savvy power user, is that they do limit compatibility; platforms often only work with similarly branded products and technologies. The company most notorious for this is GPU maker nVidia, as you simply cannot use the firm's multi-GPU SLI (Scalable Link Interface) technology with any other core-logic chipset other than its own.

As far as I can see, there isn't any technical reason why this technology couldn't be adopted to run on AMD or Intel-based core-logic chipsets. (An AMD engineer I chatted with recently at the Spider platform launch in Poland said it is in fact possible but, of course, AMD would need nVidia's consent before going ahead, which is where the trouble lies.)

Intel offers support for competitor AMD's CrossFire multi-GPU technology on its present mid- and high-end chipsets (as well as its predecessors), so why can't nVidia do the same? It seems to me that this boils down to the fact that nVidia wants to push more of its own-designed core-logic chipsets, rather than let its customers choose exactly what they want.

Sure nVidia can argue that SLI performs best on its own chipsets but I'd personally rather it let the people who are actually forking out the cash decide. If I see that Intel's Core 2 Extreme CPU combined with an Intel X38 chipset and a pair of 8800 Ultras running in SLI are the best for my gaming experience, you can bet your last penny that's the spec I'll want. I - and I suspect you as well - don't want to be forced into using SLI on a nVidia chipset just because the company wants to make more cash pushing its own platform. Bottom line: it's my money and I want to spend it in a manner that I deem fit.

Platforms are perhaps the best bet then for less techie users who want to build a focused machine without too much guess work (particularly when it comes to component selection). For hardcore techies who don't mind getting their hands dirty however, picking out components individually remains the best way to go.

Add a Comment

Your display name This field is mandatory

Your e-mail address This field is mandatory (Your e-mail address won't be published)

Security code