Penryn Goes Public: Benchmark Scrutiny
Continued from ‘Intel’s Penryn Goes Public: Performance And Pricing Statistics‘…
Now for the results. Note that the numbers below come from pre-conference testing done by Intel; at IDF, I re-ran the tests to confirm the data. My results matched those below to within a ‘noise’ margin of error; unfortunately there wasn’t time during the benchmark session to run tests for which Intel hadn’t provided pre-determined results:
|
Benchmark Or Software App |
System A |
System B |
System C |
System D |
|
3DMark 06 v1.1.0 Pro – CPU (score) |
6359 |
4569 |
NR |
NR |
|
3DMark 06 v1.1.0 Pro – Overall (score) |
17006 |
11899 |
NR |
NR |
|
Cinebench R10 Beta (CPU Benchmark) |
21521 |
11810 |
NR |
NR |
|
TMPGEnc Express v4.3.3.9999 Beta (sec) |
53 |
72 |
151 |
239 |
|
Divx v6.6.1 w/VirtualDub v1.7.2 (sec) |
NR |
NR |
20.5 |
38.9 |
NR = No Results Provided
If you look at the two desktop systems’ specifications, you’ll quickly discover that unlike past Intel Developer Forums, Intel pulled an apples-to-oranges setup this time around. HDDs and the operating systems installed on them were the same. But in addition to the inevitable CPU differential between the two PCs (which actually wasn’t as vast as I would have hoped; a 45 nm ‘Penryn’ vs 65 nm ‘Core 2′ CPU comparison would have been useful), the CPUs’ front side bus speeds were different, as were the numbers of GPUs, the amount and speed of DRAM, and even the foundation motherboards.
To get a fuller sense of the systems’ absolute and relative performance capabilities, at least until my own finer-grained study is completed and published, I encourage you to check out the following links for the results of testing done by my peers at other online and print publications:
- Intel Core 2 Extreme QX9650 - Penryn Ticks Ahead (and Motherboard Compatibility for Wolfdale and Yorkfield Processors)
- Penryn Arrives: Core 2 Extreme QX9650 Review
- Intel’s 45nm Penryn/Yorkfield architecture packs serious punch
The mobile performance benchmarks from IDF are more meaningful to me, because Intel in effect took two identical systems and replaced the 65 nm Core 2 processor in one of them with a 45 nm Penryn upgrade, leaving everything else unchanged (including the front side bus speed). Review my previously published Penryn-vs-Core 2 feature set comparison summary to refresh your memory on the chip improvements that factored in the Penryn performance boost. Predictably, of course, Intel focused its attention on video encoding applications that would benefit from Penryn’s larger L2 cache, additional SSE instructions and faster Radix-16 divider function; your mileage may vary on, say, Solitaire ![]()
Continue reading with ‘Penryn Goes Public: ASPs‘…















