|
||
|
||
|
AMD Athlon MP 2200+ As you read all over the 'net the other day, AMD announced it's newest "certified for SMP" processor, the Athlon MP 2200+. Running at an actual clock speed of 1800mhz (or 1.8ghz), this is not only the fastest Athlon MP to date, it is also the first MP processor to be graced with the new "Thoroughbred" core. For those of you not in the know, the T-bred core was AMD's die-shrink to a .13 micron process. Other than a lower voltage requirement, smaller core and reduced heat output, the T-bred is otherwise identical to its predecessor, the "Palomino". Let me start off by saying this... The Athlon MP 2200+ is 66mhz faster than the 2100+. It is based on the "Revision A" T-bred core as I mentioned before (not the swanky new revision that the 2400/2600+ XP's are using). The voltage requirements are lower than the previous Athlons (sort of) at 1.65v. If this was a Palomino core, we'd be seeing voltages in the 1.75v range for this speed. I wish my digital camera hadn't broken because the 2200+ does look a bit different than the older MP's. All of the resistors on the bottom of the Palomino based MP's have been moved up top, leaving a nice smooth bottom (that sounded bad) on the 2200+. Those of you who are obsessed with temperature monitoring and have thermistors in the sockets on your board should be tickled as now you can have direct contact with the bottom of the core. Nothing in the way. Speaking of heat output, the T-bred does do a bit better than the Palomino. Although the actual slug is smaller, allowing for less contact area with the heatsink, the 2200+ MP's are running slightly cooler than the 2100+ that they replaced. Normally I wouldn't be so stoked about a few degrees, but I am currently using the retail Taisol coolers supplied by AMD as opposed to the highly modified Alpha coolers I was using on the 2100+. Given the same heatsinks, I would imagine that the "few degree" difference would stretch into "more than a few degree" difference. Benchmark Setup(s) In the past I have presented numbers for other, non-similar machines as a reference, not a direct comparison. I received some heat over that in my last review, so I thought I should clarify this time around. That being said, I am going to throw in a few odd systems for reference, but note that the only direct comparison can be made with the 2100+ numbers as they were benchmarked in the exact same machine with the exact same configuration.
Without further ado... The numbers please! Sisoft Sandra 2002 and Clibench >>
|
|
|
|
|