Friday, April 06, 2007

Oh Intel... guess who you are two years behind... again...

Over on Mepis Lovers somebody has a problem with their AMD based laptop freezing randomly. By keeping a program running though, the system doesn't freeze. It sounds like a C-State error, the processor goes into Halt or C3 and just doesn't restart. While researching the C-States on an Athlon64... I found this interesting...

Photo Sharing and Video Hosting at Photobucket

The processor I had grabbed for the example in the post was the TMSMT32BQX4LD.

This processor is the Turion64 MT-32. It was released in March 2005, has a maximum thermal envelope of 25watts, has a 1.8ghz clock, 512k cache, and is a Socket 754 processor. The processor is also of the original Lancaster family of Turion processors, built on a 90nm process.

As it downclocks, the wattage drops. Given the about 50% drop in Thermal wattage to processor speed, (about 20w @ 1.6ghz vs. 8w @ 800mhz) it is not unreasonable to assume that if the processor were clocked at 1ghz it would be putting out less than or just about 10watts?

Okay... fast forward from March 2005 to April 5th 2007... and you have this little tidbit from Cnet :

The two new Core 2 Duo chips only consume a maximum of 10 watts of power when running full speed

Intel will offer the chips in two speeds, the U7600 runs at 1.2GHz and the U7500 runs at 1.06GHz.


Now, just to put this in perspective, these are Memron based processors. Memron is built on a 65nm process.

Is it clicking yet? It's taken Intel an entirely new die size to replicate what AMD was doing literally over 2 years ago. Is it clearer about how much of a non-issue Core2 performance and power usage is? Intel is literal years behind AMD in all areas, period, end of story.


All Core2 has done is bring Intel's products in line with AMD's products, and these new "low power" chips are just further proof of how far back Intel's technology is.

No comments: