Benchmarks: Intel's Sandy Bridge

Benchmarks: Intel's Sandy Bridge

Summary: Using a battery of benchmark tests, we take a detailed look at the performance of Intel's new Sandy Bridge processors.

SHARE:
TOPICS: CES, Hardware, Reviews
0

Intel's new Sandy Bridge architecture represents a big step in energy efficiency, as is evident in the system power consumption results. In sleep mode, the Core i7-2600K/HD Graphics system draws only 33 watts. With ATI's Radeon HD 5870 the figure rises to 53W — almost 20W less than the Core i7-870 system with the same GPU. The new Intel H67 chipset probably also makes a contribution to the more frugal energy budget.

Windows 7 SP1 RC1 64-bit, ATI Radeon HD 5870, 8GB DDR/1333; shorter bars are better

A similar pattern emerges under full load. The Core i7-2600K/HD Graphics system draws 135W, rising to 151W with the ATI Radeon HD 5870, while the Core i7-870/ATI system draws 206W. The most power-hungry system under full load is the one powered by AMD's Phenom II X6 (244W).

Windows 7 SP1 RC1 64-bit, ATI Radeon HD 5870, 8GB DDR/1333; shorter bars are better

In addition to the overall PCMark score, Futuremark's PCMark Vantage suite provides various test scenarios. The Memories scenario measures the system's performance when handling digital photos and video. This test is remarkable for the relatively weak performance of the Core i7-2600K with the integrated HD Graphics rather than the discrete ATI Radeon HD 5870.

Windows 7 SP1 RC1 64-bit, ATI Radeon HD 5870, 8GB DDR/1333; longer bars are better

Within the Memories scenario, the 'GPU image manipulation' test uses the GPU's shader units for colour correction and other image processing tasks. The discrete ATI Radeon 5870 achieved 932fps in this test, while Intel's integrated HD Graphics managed only 83fps (see below) — hence the latter's poor overall result.

Topics: CES, Hardware, Reviews

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

0 comments
Log in or register to start the discussion