One of the first things I had wanted to test in our Core i5/Core i7 desktop launch piece was power consumption. Specifically, I was hoping to see a reduction in moving from Windows Vista to Windows 7. After all, Intel was touting its involvement in getting ideal core and core parking optimized for its Nehalem architecture, and the implication was that, at the very least, we’d see more cores idling more often.
It turned out that our PCMark Vantage testing did demonstrate slightly lower Windows 7 idle numbers, but because Windows 7 would ramp into Turbo Boost more quickly, it delivered better performance and higher power consumption than the same machine under Vista. We were told to wait for Clarksfield, where battery life would show the benefit of Microsoft’s and Intel’s collaborative work.
But first I wanted to chart a couple more Vantage runs. First up: our mobile Core i7-920XM-based system running fresh installs of Windows 7 and Windows Vista. As you can see, our results here mirror what we found in looking at the desktop Lynnfield configuration. Idle power consumption is lower in Windows 7, while load numbers are higher (along with performance).
I ran the same comparison on our mobile Core 2 Extreme setup and saw the same thing. It’s also interesting to note that idle and load power consumption fairly accurately reflects the 10W TDP increase of a Clarksfield setup. In essence, it’s going to take more battery to get the same run time out of a top-of-the-line mobile Core i7 versus a flagship Core 2 Extreme.
- Arrandale: 35W, 25W, And 18W
- The Calpella Platform Update
- Centrino: Wireless Networking
- Eurocom’s Clarksfield Cougar
- Test Setup
- Benchmark Results: Synthetics
- Benchmark Results: Audio/Video
- Benchmark Results: Productivity
- Benchmark Results: Gaming
- Power Consumption: Windows 7 Versus Windows Vista
- Power Consumption: Core i7 Versus Core 2