Sign in with
Sign up | Sign in

Clarkdale's Efficiency: Core i5-661 Versus Core 2, Athlon II, And Phenom II

Clarkdale's Efficiency: Core i5-661 Versus Core 2, Athlon II, And Phenom II

Intel’s new Core i3 and i5 dual-cores arrived with a bang, offering more performance than the Core 2 Duo family they’ll eventually replace. The integration of a graphics unit into the processor may be a key enabler to maximizing efficiency of LGA 1156 systems, but just how much more efficient is Intel’s new platform? We grabbed an H55-based motherboard and the Core i5-661 (centering on the Clarkdale core) to compare with Intel's Core 2 Duo E8600 and G45 chipset. We also compared against an AMD 785G-based system running both an efficient Athlon II X2 240e and the fast Phenom II X2 550 Black Edition.

Intel is working towards greater integration. The H55/H57/Q57 platforms with Core i3/i5 Clarkfield CPUs now move the graphics unit and the memory controller out of the chipset and into the processor. The memory controller was already relocated on the Core i7 (Bloomfield) in the high-end market and Core i5/i7 (Lynnfield) in the upper-mainstream. The low-end Atom platform also follows with next-generation Pine Trail, which also transitions the platform from three to only two chips. Today, though, Intel is focused on the lucrative mainstream desktop and mobile segments. For purposes here, we’ll concentrate on the desktop side.

Our technology launch article by Chris Angelini already covered all aspects of the new processors, such as the integrated HD Graphics unit, Turbo Boost in the context of integrated graphics, the additional instructions to accelerate AES encryption and decryption, and specifications and clock speeds. Now it’s time to have a closer look at system power consumption and efficiency in terms of performance per watt. Intel now has its 32nm silver bullet, which should make the dual-cores much more efficient at base clock rates. Most people expected Intel’s new dual-cores to dominate in efficiency, but we wanted to have a close look at the power characteristics.

Ask a Category Expert

Create a new thread in the UK Article comments forum about this subject

Example: Notebook, Android, SSD hard drive

Display all 7 comments.
This thread is closed for comments
  • 0 Hide
    N19h7M4r3 , 5 January 2010 17:08
    at the moment there is only one thing amd can do... cut prices a lot lol, but i think they already where expecting this... i think the next generation of CPU's is still gona be "dominated" by intel's offerings. but the one after that amd should start giving intel a run for its money, dont forget amd has been filling a financial hole for some time, but due to recent events, and because of very strong releases in the graphics department they have been that very steadly... give amd a little time... ;) 
  • 1 Hide
    Herr_Koos , 5 January 2010 19:56
    One can only hope that Bulldozer turns out to be a MAJOR win for AMD...
  • 0 Hide
    mikehunt80 , 5 January 2010 20:50
    I thought this article really missed the point comparing Intel's new $200 CPU against AMD's $95 part (regardless of core count). I read the bit about the Athlon 240e having the lowest power consumption vs the 3/4 core variants, which is only partially true according to the data in Anandtech's review of the Athlon 435.
    At idle the 605e only looses 2.4 watts to the 240e, but at load the 605e runs a full 10.5 watts leaner than the 240e. The lower clock speed will likely be the reason, which would also result in lower single threaded performance, but at the same time the doubling of cores would improve well threaded apps such as video encoding.
    When considering that the 605e comes in about $20 cheaper than the i5-661 (not including the board and more expensive DDR3), I think it borders on dishonest to not include the 605e in the comparison.
    If you add the 605e into the picture then things are much less clear cut. Still a big win for Intel, but possibly not in all situations.
    To maintain the journalistic integrity that have kept me reading Tom's for the last 13 years, you should always be looking to compare equally priced solutions. Whereas in my opinion, this article seemed to be tipped in Intel's favor.
  • 0 Hide
    mikehunt80 , 5 January 2010 20:57
    I'd also question the use of a 785G when touting Intel's low power numbers, because in your own review you observed the Nvidia 8200 runs 10 watts leaner. With the 8200 equaling and beating the G45 in your 785G review in game performance I can't see why you didn't include it, other than to tip the scales in Intel's favor for MCE use.
  • 0 Hide
    seeratlas , 5 January 2010 21:00
    Assuming that Intel abides by its latest agreements and terminates its predatory and monopolistic efforts aimed at destroying AMD's access to markets, AMD should have a chance at making a comeback. Their choice to go with a combined cpu/graphics package that pretty much dominates intel's onboard graphics offerings at substantially lower price points, appears to have been a good one. Until Intel comes up with a decent igp, opportunities for AMD will remain open-assuming the relevant watchdog governmental authorities keep a tight leash on Intel's unlawful predatory behaviors. One could argue that Intel's history in this regard reflects a deep seated fear of AMD's potential.
  • 2 Hide
    julianbautista87 , 6 January 2010 06:43
    Not fair. You can't compare Amd's processors with doble-priced intel's ones.
  • 1 Hide
    hugodaniel , 6 January 2010 07:03
    Also, don't forget you are comparing apps that were compiled to be of generic use. In the open source world things are not that way since you can compile a given app to be optimized specifically for your cpu model. I get to handle quite some computers in my work, and realized that, if optimized, i usually find amd processors to be faster than intel ones.

    I also agree that it is NOT FAIR to compare double priced intel processors with amd just because they are of different brands. Its like comparing the best car fiat has to offer, with the best car porsche has to offer... yeah... right...