5. DHCAT Results For Recent Builds

Alas, we can only report results for five out of our six most recent HTPC builds. For the PC described in (Media PCs): Day 2 of our recent System Builder Marathon, we were unable to run DHCAT successfully in time to include those results with this story. That said, we're still working with Intel to try to figure out what's up, and will post a report to the forum entry for this story as soon as that information becomes available. Right now, all we know is that something in that configuration is causing Intel's MPEG2toWMV.exe transcoding procedure to crash whenever it runs on that machine. We have to wonder if something about the 8800 drivers is involved, because that's the major differentiating factor for that build.

The other results are best summarized in a table, which also includes links to the detailed results from each assessment run (from the entries in column 1, except also on the ViiV results for the two Intel processors as well, those being the only CPUs for which ViiV results are available). That also explains why we report results separately for the ViiV and Nero Media Home media servers only for Intel processors.

Processor Nero Home ViiV Remarks

4600+ 76 N/A This is one of the lowest DHCAT scores we've ever seen, underscoring the unsuitability of using the X1250 on-board graphics for media PC graphics. 4800+ 93 N/A Our experience in using this machine to view high-definition Blu-ray and HD-DVD content contradicts this relatively low score. 6000+ 116 N/A The 6000+ bested even the T7600 in PCMark05 and 3DMark05 comparisons, but comes out behind both Intel models. T7200 103 122 This is the first PC in our handful to earn a passing score in all scenarios except one (it does best in Enjoy and Share, and worst in Record, where it managed a Good/Acceptable but not an Excellent/Pass). T7600 111 132 Interestingly, the most powerful Intel processor in this bunch outdoes all the other contenders.

Please note that both of the T7600 processors are missing the summary graphics for that benchmark (it also appears in a file named PCM.png, which was missing from the DHCAT results for those two runs for some reason we don't fully understand). You can get access to this information by scrolling down to examine the detailed results section near the end of this long tabular data display. Also, the T7200 Nero Home media server results include some strange visual artifacts in the same section, again for reasons we can't fathom. To learn how to interpret DHCAT results, see sidebar to this story entitled "Reading the DHCAT Results Report."

Given that the same Intel hardware scores 15-16% better when running the ViiV media server as opposed to the Nero version, it's hard not to speculate about competitive advantages. Intel's counter would surely include language to the effect that AMD should try to get media server software written that works as well as theirs does, but perhaps that isn't necessary. By either adding the delta from the Intel side to the AMD builds, or comparing only the Nero Media Home based results to one another, we get a much better sense of how the field stacks up. Not coincidentally, this mostly echoes our earlier benchmarks on these platforms, except that here the T7200 edges out the 4800+ by just over 10 percent, whereas the advantage goes the other way in those other measurements.

When we ran multiple iterations of the DHCAT assessments - the runtime environment makes this very easy, and in fact, three iterations is the default - results varied by as much as two or three points across repeats. Interestingly, results appeared most likely to decline by one or two points from the first through the third repetition. This means that those scores for machines that fall within three points of one another are more or less the same, and makes a more detailed item-by-item comparison all the more interesting.

Our final observation apparently represents a quirk within DHCAT's device detection code. Though the program will happily detect either a Blu-ray or an HD-DVD drive by itself, when faced with both types of drives, it reports the HD-DVD drive as present and the Blu-ray drive as absent. We don't know what happens when the program is presented with the LG combo drive, either, because we haven't yet had a chance to test one. According to Intel representatives, there is no difference in the DHCAT score between a system that has one type of high-definition DVD player or another, and one that has both (or a combo drive). Nevertheless, we hope they'll fix this in their next release.