Intel D865PERL - Page 6

..:: Benchmarks ::..

Well, now it is time to take a look at the performance of the Intel i865PE chipset, and the D865PERL ďRock LakeĒ motherboard. For our tests, we used a 3.00GHz 800MHz FSB Pentium 4 with Hyper-Threading on an Intel D875PBZ, and an Intel D865PERL. Each system was tested immediately after a fresh system reformat with the latest system drivers installed. For both systems, all onboard features were disabled during the benchmarking process, and identical test hardware such as graphics card and RAM were used in order to achieve the best results. This will give us an accurate ďapples to applesĒ performance difference between the two chipsets with processors running at identical internal clock speeds and system bus speeds. It is important that I note both systems had Hyper-Threading enabled for the most accurate chipset comparison results. In the graphs, we have shown the performance of both chipsets in Dual Channel mode for optimal performance.

..:: SiSoft SANDRA ::..

To start things off, letís take a look at the results we achieved. Since the only items that i865PE lacks over the i875P chipset are ECC memory support and PAT, the performance difference between these boards should be rather small with the i875P holding a slight edge. The first test we ran with SiSoft SANDRA was the Arithmetic test, and here we can see that both boards put up good numbers, but the i875P chipset comes out on top for both ALU and FPU. The same can be said for the Multimedia test, yet once again both chipsets are putting up very similar numbers as they well should. The last SANDRA test is one that any users will be interested in, that being the memory bandwidth. Here we can see where Intelís PAT comes into play. There isnít too much of a difference between i875P and i865PE, but it is enough to make a difference in applications and games.

..:: SuperPI ::..

In the SuperPI tests, we run the program four times, once at one million, once at two million, once at four million, and yes once at eight million. The numbers in the graph above show the time in seconds that it took the system to calculate pi to the set number of digits. In this benchmark, we can clearly see that the i875P chipset holds an advantage over the i865PE chipset, although once again this advantage is quite small in the great scheme of things. Both chipsets put up identical numbers in the one million digit calculation, however, i865PE starts to fall slightly behind by three seconds in the two million digit calculation, then by two seconds in the four million digit calculation. Finally, in the last eight million digit calculation, the i865PE board has fallen behind by a mere six seconds. This is nothing like the differences we were seeing when i875P was placed against the older i845PE. There we saw up to 59 second differences. Needless to say, i865PE is shaping up nicely so far

..:: Specviewperf 7.0 ::..

So far, the i865PE chipset has held up well when placed against the i875P chipset. Both of these boards put up some very nice numbers, especially when placed against older chipsets such as i845PE. In the SPECviewperf 7.0 benchmarks, we can see that, as expected, the i865PE and i875P chipsets put up nearly identical numbers in each and every performance benchmark. In the UGS01 test, the i875P holds a lead of 3.5%, while the Proe01, Light05, Dx07, Drv08, and 3dsmax01 tests show performance leads of 1.4%, .7%, .7%, .5%, and .04% respectively.

..:: FutureMark 3DMark2001SE ::..

FutureMarkís 3DMark2001 SE shows us something we have seen throughout our testing so far, the i875P chipset comes out on top. In these tests, and all those following, we ran the benchmarks in 640 x 480 resolutions in order to avoid the graphics card bottleneck that we had been experiencing in high resolutions. Well, for 16-bit performance, the i875P chipset holds a lead over the i865PE chipset of 1.5%, while in the 32-bit benchmarks, we see similar results with the i875P coming out on top by roughly 1.0%.

..:: Quake III Arena ::..

Well, hereís a big surprise, once again we see both chipsets putting up some nice numbers when placed head to head. In the 16-bit color benchmarks, we are seeing a performance gain over the i865PE chipset of 1.7%, while when the benchmark was run in 32-bit mode, we see a performance lead for the i875P of roughly 1.3%. With the kind of numbers the i865PE chipset is putting up, itís becoming harder to justify any reason as to why a normal user should purchase an i875P board, but Iíll delve deeper into that topic in a minute.

..:: Unreal Tournament 2003 ::..

Last up for today we have yet another benchmark for our motherboard reviews, Unreal Tournament 2003. We are using the built-in benchmarking utility with custom .ini files to make sure all settings are at high quality to allow for optimal benchmarking results for comparison. Each of the UT2003 benchmarks once again shows the i875P chipset coming out on top, albeit by a small margin. In these benchmarks, we are seeing a performance gain of roughly 2.8% in the Flyby and 3.3% for the BotMatch tests. These are the largest performance gains we have seen all day between i875P and i865PE.

Well, it seems that the i865PE chipset holds its own very nicely when placed head to head against the i875P chipset. Since the i865PE is geared more towards the enthusiast market, while i875P is geared more towards a workstation environment, the minor performance losses we are seeing here do not justify heading out and throwing down the extra green on an i875P board unless you want the best of the best. From our benchmarks, we can see that Intelís Performance Acceleration Technology does indeed have some effect on the overall performance of the system, but in the end it turned out to be fairly small. This performance difference did seem to grow as we started to utilize the more modern gaming benchmarks such as Unreal Tournament 2003, and I suspect it would continue to grow with more demanding games. Now that weíve seen everything there is to see with the D865PERL, letís wrap everything up and get on to the next review!