bit-tech.net

The PCI-Expressway to Hardware Hell

It’s not often that I feel the need to sit down and rant in public about something, but today, you’re going to hear it full on from me. Some of you may know that I’ve not had the greatest of times with hardware over the last few months – you’ll remember we looked at a Radeon X700 XT which was nothing more than an engineering sample, and we remember SLI and GeForce 6600GT also being announced many moons before we actually had the chance to see products in our hands. These ‘paper’ launches, coupled with the last 14 days of sheer bad luck, have really taken me to the end of my tether.

As you know, it wasn’t long ago when Intel made quite a significant step forwards with their LGA775 processors and 915/925 chipsets, which brought PCI-Express in to the market place. You can read many, many reviews around the web (and in print), that say how good or bad PCI-Express is. We have also featured a few PCI-Express reviews, where our thoughts have been quite positive, and certainly not bad. On the whole, I do think PCI-Express is the way forwards in the long run - it is a fantastic development that should enable many technologies, most importantly 3D, to move on at an even greater pace than they are already. However, after the ongoing problems that I've had with PCI-Express – yes, it's almost as if the problems were part of the bundled extras included with the product - I don't think that it is quite a way of life just yet. I’ll get to why in due course, but for now, you will just have to hear me out…

"... it is a fantastic development that should enable many technologies, most importantly 3D, to move on at an even greater pace than they are already."

Many of you will know that NVIDIA finally released their SLI technology a week ago Tuesday after announcing it back in August. ATI also released a new video card this Wednesday just past - the Radeon X850 XT Platinum Edition, and many of you are also wondering where our reviews of these technologies are, right? We were sent kit to evaluate NVIDIA’s SLI technology, albeit not quite all of what we would have liked – it would have been nice to receive all three video card configurations to examine. The video cards that we received were damaged in the post, and so far, no other cards have become available to replace our broken pair. Supplies are tighter than a ducks behind.

We were also offered the chance to evaluate the X850 XT Platinum Edition over the weekend - an overnight stay in a hotel with the video card (… oooer missus). Depending on which way you look at it - either fortunately, or unfortunately - we didn’t make it down to London for our weekend of mad red-hot passion in a hotel room with ATI’s latest 3D development. I was sat at home in something that couldn’t be described as anything other than Hardware Hell.

Now, I don’t normally spend a massive portion of my weekend working on articles because there is a life outside of bit-tech (so I'm told...) and one needs time to get away every now and again. Instead, I spent all Saturday and most of Sunday morning troubleshooting three different reference PCI-Express motherboards that were playing funny games with me. You all know how frustrating it is when something works in one motherboard, but doesn’t in another - I swear if they had fingers, they’d be pointing and laughing at me.

Back to top

Tim Smalley


In fact, since the release of the first PCI-Express motherboards, I’ve had my hands on seven different PCI-Express boards, of which one has functioned 100% the way that it should do. This isn’t just limited to Intel’s chipsets - I’ve experienced it with PCI-Express based motherboards for Athlon 64 too. In fact, I’m yet to see a fully working PCI-Express motherboard for Athlon 64 – quite amazing when you think about the publicity around nForce 4 and K8T890. Maybe I’m just the unluckiest guy on earth, because I know that many of my fellow hacks haven’t suffered from the same consistent issues that I have.

"I swear if they had fingers, they’d be pointing and laughing at me."

You may think I’ve got a bad taste in my mouth, or that I’m doing something drastically wrong when I’m building systems – I wish it was the latter, because I could fix that with a bit more care. The fact is that every component that has been tried on these boards, work in other motherboards, and have been verified as ‘good’ components. The fact is that two of the motherboards are reference samples – in other words, they’re not final products as they are designed for the purposes of demonstrating new technology. The final motherboard is a ‘product’ that has come from a board partner, but I’ve been informed that the design I have isn’t a final product, despite it being a ‘Rev. 1.02’ board. Apparently, more bugs have been found with this design, which has forced the engineers to have a slight design re-think in order to come up with yet another board revision – it should still be an engineering sample, not a 1.xx revision. So, do I have a bad taste in my mouth? Yes, partially…

If these reference motherboards aren’t up to the job, should they be here to review, or preview? Will these (p)reviews actually mean anything to you, the consumer, if you can’t go out and buy them, because they’re not finished yet? Would reviewers be better waiting until the products are ready for public consumption before choosing to review them? Or should these technological previews really be set off under a different heading all together? I get the feeling that the constant paper announcements that have been a telling feature of 2004, do nothing but create hype; hype for something that could never make it on to the shelves in any considerable volume to enable you, the consumer, to go out and purchase the product. Perhaps 2005 will be the year that the hype bubble bursts - we can only hope.

Now, if you'll excuse me, I've got some more motherboards to troubleshoot...