And for the first time, we are benchmarking workstation power consumption. To do this, I used the P3 Kill-A-Watt Electricity Usage Meter, which I highly recommend. It's inexpensive, it's accurate, and it gives all sorts of electrical data. I measured power consumption at rest (processor at 0 per cent operation, with the system not hibernating) and at peak (100 per cent operation for all processor cores). To obtain a single power rating for a machine, it's useful to employ the same power-consumption measure published in the recently released EnergyStar 4.0 specification. Its formula is:
Rating (in watts) = 0.35 (Pmax + (HDD x 5))
Pmax is power consumption when the system is running at 100 per cent, and HDD represents the number of hard drives on the system. This mark is given in the benchmark results table for each workstation. Between it and the other two figures plus the handy power meter, you can see how your current systems compare with those reviewed here. I suspect you'll find the reviewed workstations are surprisingly efficient. All four rely on 80+ power supplies (meaning more than 80 per cent of the incoming power is transformed into power for the workstation), and both HP machines have been given EPEAT Gold certification for meeting a series of environmental criteria.
The middling midrange
Both midrange workstations, the Dell Precision T5400 and the HP xw6600, are impressive in several respects. They are small -- smaller even than standard desktop PCs. They are also almost completely silent. Save for the Dell startup routine in which it races its fans, there is no difficulty hearing a clock tick over the faint hum of these systems. Their heat dissipation is also remarkably good, considering their 0 per cent load consumption is a healthy 189 watts. Most of the time, it is barely noticeable, even in a small office. At full tilt, however, the heat they throw off becomes quickly apparent.
We asked Dell and HP to use the same models of quad-processors (Intel Xeon E5430 chips running at 2.66GHz) and to load the systems with Microsoft Windows XP Professional SP2. We then asked them to deck out the workstations with any additional hardware they wanted that would still put the price in the roughly US$5,000 range. As can be seen from the features table, both companies made essentially the same choices: Nvidia Quadro FX 4600 graphics, 4GB of system RAM, and motherboards using the same chip set. Because of this, their benchmark results are approximately even.
The Dell T5400 model is slightly faster on graphics, but has a slower and smaller disk (80GB). The HP xw6600 had a faster, more capacious hard drive (250GB), but is about 7 per cent slower on the graphics. The Dell system has slightly better expandability as regards the PCI Express slots. It has two more slots, both of which are PCI-X 64-bit. It also has a more capable power supply. For these reasons, I could see giving a slight edge to the Dell Precision T5400, but considering that their prices are nearly identical and pricing varies widely from week to week due to sales promotions, I suggest that if these models appeal to you, you should buy from your preferred vendor or base your decision on price at the moment of purchase.
When two models are so close, it is customary for reviewers to suggest that both models should be evaluated before a decision is made. While that advice is certainly applicable, it's not my first recommendation. Rather, I think you should examine the comparative benefits of the other two machines: the value-oriented HP xw4600 and the high-end Dell Precision T7400. In both cases, I think you get more value for the dollar.