AMD Athlon MP and Intel XEON: The Long Term

For a refresher on both reviews, I am going to go over the conclusions we drew after testing both systems. We determined that for most workstation purposes, especially those involving audio/video and 3D application work, the Athlon MP system was superior to the XEON system not only in performance but also in price. I also mentioned that I would be following up that article with more in depth testing of high-end workstation and server applications, and I still intend to do that. From my findings so far it looks like the XEON’s memory throughput and very fast cache are helping it fend off the MP system in web and e-mail server situations. This is really not surprising, and in a way, fits with the pricing structure of both machines.

Figure 1.1 /// XEON Logo Figure 1.2 /// Athlon MP Logo

That leads me into the one comment and request we got more than any other – The Athlon MP (and in a sense the new XEON) systems have NOT been tested for long-term reliability and many of you are worried about that. I have received e-mails from professionals in the industry who are either wary of taking the plunge and going with an AMD based workstation or server, or who desire to do just that, but cannot because of others who are being cautious. There have even been articles on the net addressing this issue. Some sites are being very careful about handing out recommendations, and I cannot blame them. Now for you and I it is easy to scoff at this wariness, but they are right. These systems, in particular, the MP (since it is AMD’s first attempt to enter the workstation/server market) have not been tested for long-term reliability because no one has had the time to do just that.

I want to point out that in the business world stability means something entirely different that it does in the desktop market. At home in our living rooms we often shut our systems down everyday, or every week – often times we don’t flinch if we get a lockup or software failure. This isn’t true for everyone, but it is for most consumers.

On the other hand, when you talk about servers, you want them to be able to run non-stop, without constant supervision, for months and months on end. In the business world, when workstations or servers go down money is lost – a lot of it. That is why it isn’t uncommon for some businesses to spend the extra money up front for machines they know might be a little slower or a little heavier on the wallet but have proven themselves up to the tasks the business needs them to perform. You’re paying for reliability, and like I said above, we still don’t know if either of these machines can deliver that for the long-term.

This is not a criticism of any website that reviewed the AMD Athlon MP or Intel XEON – I understand that time limitations made it impossible to do this since most sites had these systems for no longer than a month before the initial reviews. I think that sometimes readers assume we get the hardware months and months before it is released. While we try and spend as much time with hardware as is possible before reviewing it, often times there are deadlines that must be met – and this causes a problem.

As I am sure you’ve all read, initial reports on the stability of both systems are very good. The Tyan Thunder K7 using the AMD762 chipset and the Iwill DX400-SN using the Intel i860 chipset, have both been rock solid in my testing and that of many others. Huge credit has to go to all parties involved on both sides in that sense. They just haven’t had ample opportunity to show long-term stability or instability because they haven’t been around long enough to do that.

Figure 1.3 /// MP ad Figure 1.4 /// Intel Ad

So, what do I intend to do about this? First I intend to follow-up both articles, like I mentioned above, with more and better testing. Hopefully this will help determine where both machines are best suited. Ultimately I want to use both these systems for an extended period of time in a server or workstation capacity to really test their reliability. As we all know a machine that can feel rock solid after the initial install of an operating system can quickly become ‘quirky’ after certain software or hardware is loaded. Obviously I will not be able to test every scenario – but over the next six months to a year I want to throw everything I can at these machines to see what, if anything, will cause them to falter. I do want to mention that these updates are not going to be MP vs. XEON articles. I will leave that for the reviews – these will be reports on the stability and usability of the machine.

So what about the specifics? As I am sure you can see Accelenation.com is still in its infant stages. We took a big leap when changing the design and name from FullOn3D, and we’re growing everyday at an exponential rate. Because of this we have to think about the long-term options available to us with regards to serving out page. We plan forums and other traffic intense ventures and as I am sure you can tell, our current server isn’t the fastest machine on the planet. What I am getting at is this – I am going to run one of these machines as our dedicated server (sadly, one is enough right now!) and the other I plan on using as an audio/video workstation that will be working non-stop, all day, everyday. With the help of some industry professionals I am going to do everything I can to try and find flaws with either machine and will report to you the progress of both regularly as I use them.

To digress for a moment, this leads me into something else we plan to do here at Accelenation.com and that is following up on our reports. Often times you’ll read a review, see some benchmarks, check out the conclusion, and that will be the end of it. What we want to do is keep up on the products we review, keep using them, get feedback, and report any changes to you readers. Much like we did at FullOn3D with the Iwill KK266 and KK266 FAQ we plan on updating you on problems, BIOS updates, drivers, fixes, and anything else that affects the owners of these products.

The thing that sparked this idea was a review I once read of a motherboard that had just been released, the infamous FIC SD11. This board was one of the very first to support the newly released AMD Athlon (heh), and was met with harsh critism for being horribly unstable, which, at the time it was! I bought one of these boards when it was first released and was disgusted. However, within 6 weeks of the release FIC has released at least 4 new BIOS’s which fixed almost every issue the board had. I didn’t see one follow-up anywhere regarding this and found it a little disappointing. The SD11 was never a great board, but FIC deserved some credit for keeping up on the issues and fixing them pretty quickly. So, how do you know, when you read a poor review of a newly released board, that that board won’t be a great buy two weeks after the review? You wouldn’t unless someone told you, and that is what we’re going to try and do.