Quad vs. Dual Duel: Intel Core 2 Quad vs. Intel Xeon Core 2

Technology: Page (1) of 1 - 11/21/06

Quad vs. Dual Duel: Intel Core 2 Quad vs. Intel Xeon Core 2 Two HP Machines face off with the latest processors on board By Charlie White

When Intel released its 5300 series quad-core Intel Xeon processors earlier this month, we wondered how they would compare to their dual-core Intel Xeon 5160 predecessors. So, we pitted the two directly against each other, comparing our benchmark results of the dual-core 3GHz Intel Xeon 5160 HP xw8400 machine we reviewed earlier this month head-to-head against a similarly-prepared HP xw8400 workstation equipped with two of these new quad-core Intel Xeon X5355 processors running at 2.66 GHz which Intel is officially naming Core 2 Quad. Was there a big difference between the two?

First off, these machines were not quite identically prepared, but almost. The dual-core Woodcrest machine had a slight disadvantage in that its hard disks weren't nearly as fast as the speedy array on our quad-core Clovertown machine, where the Woodcrest's read and write speed were 114MB per second and 81MB per second respectively, compared to the faster read and write speed of the Clovertown at 173MB and 165MB per second. But the Woodcrest machine had a slight advantage, too, running Windows XP x64 edition compared to the 32-bit version of XP running on the quad core machine. In previous tests, we've noticed that the 64-bit Windows offers a slight speed advantage, but it's only equal to a percentage point or two.

These Core 2 Quad processors represent the current pinnacle of Intel engineering, with 8MB of total cache (4MB per die) communicating with the frontside bus at 1333 MHz. The chips are aimed at workstation users who are probably willing to swallow its higher power requirements, with a 120-watt thermal design power (TDP) compared to 80 watts of the 2.33GHz chip that's one step down from it. As usual, these highest-end processors aren't cheap, either, each costing $1172 at their introduction this month.

Starting up the machine, the first place I went to was the Performance Monitor in the Task Manager, which showed up eight windows in the CPU usage history area. All were humming along and ready to accept our benchmarks, where we expected to determine our latest king of the hill here the Midwest Test Facility. First, we wanted to load and run the PCMark 2005 and 3DMark 2006 benchmarks, the newest members of our stable of tests. Unfortunately, the PC Mark wouldn't run on the software we had installed on the new Clovertown machine, and we suspect it's because the benchmarks required Windows Media Player 10, and when it called for that, Windows Media Player 11 was automatically installed as is Microsoft's wont. Undaunted, we went ahead with the 3DMark 2006 benchmarks, and not surprisingly, the quad core processors blew by anything we've ever tested by a longshot, showing us the fastest CPU benchmark we've ever seen, 5571, making the dual core machine's previous high score of 4653 look rather lame by comparison.

The most impressive performance in our group of benchmarks was in the Maxon CineBench rendering score, where the quad core processors and whipped the dual core processors soundly by a score of 1926 to 1558. However, a revealing factor: even though there are eight cores at work with this configuration, the CineBench rendering score only showed a 4.46 asked multiprocessor speedup. That reveals the extent of work yet to be done writing software that can take advantage of the enormous power present with all these multiple cores.

Results in minutes: seconds
Boldface indicates winner. Dell Precision Workstation 380, single Intel Dual Core Pentium EE 3.2GHz processor, 2GB DDR2 RAM
Dell Precision Workstation 390, Single Intel Core 2 Extreme 2.93GHz, 2GB DDR2 533MHz NECC RAM
HP xw9400 Workstation, two dual-core Opteron 2220SE processors at 2.8GHz, 4GB RAM, running Windows XP x64 HP xw8400 Workstation, two dual-core Xeon 5160 (Woodcrest) processors at 3GHz, 4GB RAM, running Windows XP x64 HP xw8400 Workstation, two quad-core Xeon 5300 (Clovertown) processors at 2.66GHz, 3GB RAM, running Windows XP

1. After Effects: Video Composite :38 :20 :39 :23 :24

2. After Effects: Data Project 1:12 :40 :54 :38 :34

3. After Effects: Gambler :19 :10 :23 :19 :20

4. After Effects: Source Shapes 2:29 1:21 1:40 1:07 :57

5. After Effects: Virtual Set 1:53 1:13 1:20 :59 :52

Maxon CineBench Rendering (CB-CPU score -- higher is better) 715 (2.06 multiprocessor speedup) 903 (1.86x multiprocessor speedup) 1554 (3.04x multiprocessor speedup) 1558 (3.07x multiprocessor speedup) 1926 (4.46x multiprocessor speedup)

TotalBenchmark comp 1 87 sec. 50 sec. :52 sec. :48 sec. :46 sec.

TotalBenchmark comp 2 992 sec. 781 sec. 637 sec. 482 sec. 443 sec.

Night Flight 35:01 (2101 seconds) 50:04 (3604 sec.) 30:07 (1807 sec.) 21:10 (1270 sec.) 14:23

Hard Disk Speed Read 121MB/sec.
Write 97MB/sec. Read 178MB/sec.
Write 160MB/sec. Read 159MB/.sec.
Write 163MB/sec Read 114MB/sec.
Write
81MB/sec Read 173MB/sec.
Write 165MB/sec.

3DMark 2006 SM: 1672
HDR: 1653
CPU: 2551 SM: 1655
HDR: 1635
CPU: 3697 SM: 1655
HDR: 1637
CPU: 4653 SM: 1670
HDR: 1658
CPU: 5571

PCMark 7247 8144 9303 n/a, see text

With our After Effects benchmarks, the differences between the two processors were not quite as dramatic. In our first group of five After Effects comps, the quad core machine beat the dual core in every single test, but wasn't able to best two of the benchmark results of the Dell single-processor Core 2 Extreme machine running at 2.93GHz we tested a couple of months ago. But after those small defeats, the Clovertown machine couldn't be touched again by any of our contenders, especially with the Nightflight After Effects composition, which was devised to show the power of the Macintosh over the PC using extensive vector graphics rendering. The quad core machine crunched through that in 14:23, besting its closest competitor by more than 25%.

Summing up, these first benchmarks show a faster processor under the hood, but we can only expect these numbers to get better as software writers learn how to write tighter and more-efficient code for this multithreaded hardware. We're also expecting prices to come down, and that this kind of speed will be obtainable without the power-hungriness associated with these current chips. Even so, it's always nice to see faster processors, even if they're not fully twice as quick as their predecessors. So was there a big difference between the two? No. But that depends on what software you're running.

Page: 1

Related Sites: Digital Producer , Media Workstation , Digital Video Editing , Animation Artist , Audio Video Producer , Digital Post Production , Film Imaging , HD Issues , Oceania , BN - NAB , BN - IBC
Related Newsletter: DMN Newsletter , Timeline Newsletter , Digital Media Net , BN - HD Weekly , DMNForums