Reducing Global Emissions through ICT

Last week an important report was published entitled Smart 2020: Enabling the low carbon economy in the information age. Coming from The Carbon Group on behalf of GeSI (Global eSustainability Initiative), the study has important data on information and communication technology (ICT) and CO2 emissions.

Many of us have become accustomed to headlines about data center energy consumption and power/cooling concerns. And this report does detail the direct energy consumption of ICT including forecast data with clearly documented assumptions. As seen elsewhere, it’s estimated that ICT accounts for about 2% of global CO2 emissions. But this study goes further. Through enabling other sectors to reduce their emissions, the ICT industry could reduce global emissions by as much as 15% by 2020 – a volume of CO2e five times its own footprint in 2020.

Now this is not to say that we shouldn’t worry about the 2% and growing direct footprint of ICT. Intel is strongly committed to continue to use the benefits of Moore’s Law to deliver improvements in performance and in energy efficiency – in fact through our “go managed, go mobile” analysis, we can demonstrate the dramatic power savings of latest generation desktop and notebook systems as well as the benefits of power management technologies. All told, today’s Intel Centrino processor-based notebooks are 17X more efficient than Pentium D based desktops (see footnote for details). Similarly, our server processors are delivered unparalleled performance and reductions in power consumption. Intel along with the industry is taking steps to improve the 2% beyond what any one company can do alone. We have formed the Climate Savers Computing Initiative to improve power delivery efficiency and the adoption of power management – these two steps will reduce computer energy consumption 50% by 2010. And the Green Grid is focused on holistically improving data center efficiency – both through facility design and IT best practices.

But the coverage in this report from GeSI on the benefits of ICT on specific areas of global emissions is breakthrough. It makes specific recommendations for focus. Through our leadership in the high performance computing market, we’ve long seen the benefits of compute cycles on demanding application areas such as oil and gas exploration and medical imaging. The report encourages us to expand our thinking to include the permeation of communicating and programmable devices into the electric grids worldwide and to make building smarter as examples. With the Atom family of processors, we’re well-positioned to do our part in this transformation.

Going forward, I’m eager to engage in the dialog that will naturally result from this thought-provoking study and formulate new efforts within Intel, across the industry and in cooperation with governments globally to continue to maximize what energy ICT consumes – and also to effectively apply it to reduce CO2 much more broadly.


Footnote (details on Centrino vs. Pentium D efficiency comparison):

Intel Core 2 Duo processor E6550 (2.33GHz, 1333MHz FSB, 4MB L2 Cache), Intel DQ35JO motherboard with Intel Q35 Express chipset; 1333MHz FSB, Intel integrated graphics, 2x1GB Micron DDR2-667 5-5-5-15, Seagate 320GB/16MB cache/7200rpm, Windows Vista Enterprise.

Intel Pentium D Processor 945 (3.4GHz, 800MHz FSB, 2x2MB L2 Cache), with Intel 945G Chipset on Intel D945GPM board, Intel Chipset Software Installation File, Dual Channel Micron PC2-5300U 2x1GB of DDR2 667 5-5-5-15, Seagate Barracuda 320GB NCQ SATA2 7200RPM, Windows Vista Ultimate RTM Build 6000 NTFS.

ACER 5720 Intel Core 2 Duo T7700 (2400MHz/800FSB L2 4MB) with Integrated Graphic,2x1GB Hynix DDR2-667, Hitachi 160GB 5400rpm HDD

Display power draw values gathered from spec sheets of 10 models of each type of display (CRT and LCD).

Intel EEP methodology is described here:

And now, for some legalese: Performance tests/ratings are provided assuming specific computer systems and/or components and reflect the approximate performance of Intel products as measured by those tests. Any difference in system hardware or software design or configuration may affect actual performance. This data may vary from other material generated for specific marketing requests.

Comments are closed.