The Test:

Santa Rosa Test Bed
Performance Test Configuration
Processor Intel Core 2 Duo T7700
(2.4GHz, 4MB Unified Cache)
Chipset Intel PM965 (800MHz FSB)
RAM DDR2-667 (2x1GB), 5-5-5-15
Networking Intel 82566MM Gigabit, Intel 4965AGN WiFi (802.11n)
Hard Drive Hitachi 7K100 80GB
System Platform Drivers Intel - 8.2.0.1014
Video System NVIDIA GeForce 8600M
Video Drivers NVIDIA 158.18
Audio Intel HD Audio
Operating System Windows Vista Ultimate 32-bit
.

NAPA Test Bed
Performance Test Configuration
Processor Intel Core 2 Duo T7600
(2.33GHz, 4MB Unified Cache)
Chipset Intel 945PM (667MHz FSB)
RAM DDR2-667 (2x1GB), 5-5-5-15
Networking Integrated Gigabit LAN, Intel Pro Wireless 3945 (802.11g)
Hard Drive Hitachi 7K100 80GB
System Platform Drivers Intel - 8.2.0.1014
Video System ATI Radeon X1600
Video Drivers ATI V8.352.1
Audio Intel HD Audio
Operating System Windows Vista Ultimate 32-bit
.

For our test Intel sent us an unbranded OEM whitebook built around its new Santa Rosa platform. While serving as a fine platform to test Santa Rosa on, it didn't give us much opportunity for direct comparison to other Centrino designs. Given that the Santa Rosa Merom CPU and the Napa Merom CPU aren't interchangeable due to different pinouts, developing an apples to apples comparison became rather difficult.

We turned to our original Napa test platfrom from our first Merom article, the ASUS Z96j. Outfitted with a Core 2 Duo T7600 (2.33GHz), and the same DDR2-667 memory and hard drive from our Santa Rosa test bed, it was the closest competitor we could find. Of course the Z96j used a mobility Radeon X1600 while our Santa Rosa test platform used a GeForce 8600M, so we couldn't perform any 3D gaming tests comparing the two (there will be a follow-on article looking at the performance of the GeForce 8M series of GPUs).

Using two very different notebooks made battery life comparisons meaningless: they had different screens, different GPUs and different batteries. But we did look at average power consumption between the two notebooks to give us some indication of the battery life of this new Centrino platform compared to its predecessor.

Honestly there's much more testing that has to be done in order to give us a true verdict on the Santa Rosa Centrino platform, the majority of which will force us to wait for a better test platform. The verdict is still not out on Turbo Memory as you've already seen and we do need to look at the range and power consumption of Intel's 802.11n solution compared to its competition in the market. Once better Santa Rosa platforms become available for us to test, you can expect a follow-up to this article focusing on what we've left out.

Note that in our charts we have two Santa Rosa bars, one listed as w/ TM, indicating that Turbo Memory was enabled for that configuration.

Turbo Memory Testing General Performance
Comments Locked

22 Comments

View All Comments

  • avaughan - Thursday, May 10, 2007 - link

    I would have thought that the real benefit from Robson would have been the power savings from the ability for the OS to write to
    disk/save something/do a small amount of swapping without needing to spin up the hard disk.

    Regarding performance expectations, what's the read/write throughput on the flash, and is this a fast hard disk? I would expect
    reading/writing a large contiguous file to/from a fast hard to be faster than reading/writing to cheap (= slow) flash.

    If Vista can store part of the data on the hard disk, and the rest on flash, and read/write both chunks simultaneously
    or the flash has throughput as fast/faster as the hard disk then I would hope to see performance benefits.
  • smn198 - Thursday, May 10, 2007 - link

    Maybe it was an interesting article but I cannot read the begining of each line as there are Intel vPro dogs in the way.

Log in

Don't have an account? Sign up now