Murtaza's Technoblog

Today I thought I would talk about the advantages which smaller fabrication processes offer. All of the next gen chips are made on smaller and smaller processes. Intel has recently announced their 32nm fabrication process for the next generation processors. Nvidia has started manufacturing their latest GPUs on 55nm processes. So what does all this mean?

Smaller fabrication processes not only benefit the manufacturer but also the end user. They run much cooler and are more power efficient than older processes. If this isn’t enough, even the costs come down significantly. An example of this is Intel’s Wolfdale. Previously, an all round processsor would have been 2.4GHz and cost around $300. Intel released a 3.0GHz processor for $250 which was not only faster but it could be overclocked like no other processor and still run relatively cool. An Intel E8400 at 3.0GHz could be overclocked to 3.6GHz with no problems whereas previous Intel processors struggled to make such a speed jump without extra cooling.

This means that a 32nm processor will outshine a 45nm processor in everything. Speed, overclocking ability and cooling. Not only this but it will mean that not only gamers will have access to incredibly fast processors due to their affordable price. Anyone will be able to get a processor with a speed greater than 3.0GHz. The effect on Video Cards will be similar. Discrete graphics cards will not be limited to just users needing high performance. Not only this, but it adds the potential to easily build powerful graphics chips into motherboards as cooling will be less of a problem and it will take up less space. Currently Nvidia’s latest graphics card, the GTX 295 is fast but it runs incredibly hot. It is impossible to run it in triple SLI with only air cooling. Part of the problem is the 65nm chip. This needs more power and runs very hot. In the future Graphics card manufacturers can fit similar performance in a smaller more accessible package. A smaller process also means more can be crammed onto the chip. As you can see from the above image, Intel is able to fit a graphics chip and memory controller into the processor now as well!

This is what these smaller fabrication processes mean. The advantages they give us are huge. Computers could be put in places you never thought possible. IBM is experimenting with chips right now using structures similar to DNA. This allows manufacturing processes as small as 22nm.

The new chip has graphics capability built into the core processor, a first for Intel, and has new Advanced Encryption Standard (AES) instructions also built in for faster encryption and decryption. The new chip will go into production in the fourth quarter of the year.

“The rapidly increasing number of transistors and processor instructions we add have made possible the integration of more and more capabilities and features within our processors,” he said.

“This has driven an incredible amount of innovation throughout the industry.”

Maloney will also be showing the Westmere-EP server platform and said that it would provide a performance boost greater than that seen with the introduction of the Xeon 5500 series of server chips over the previous generation.

Intel is also working on an ultra-low power server chip in the Xeon 3000 range, which will operate at just 30 watts.

[ad#Adsense]

As a follow up to Westmere the next generation chip, codenamed Sandy Bridge, will see an acceleration in graphics capabilities for the chip, with Intel’s sixth generation graphics core built into the die. This will accelerate floating point, video and processor intensive software applications.

To help IT administrators Intel has developed a new PC management tool called Keyboard Video Mouse Remote Control, which allows managers to view problems directly as the user sees them, which it hopes will speed up fault solving.

Read Full Story

SAN FRANCISCO (Reuters) – International Business Machines Corp is looking to the building blocks of our bodies — DNA — to be the structure of next-generation microchips.

As chipmakers compete to develop ever-smaller chips at cheaper prices, designers are struggling to cut costs.

Artificial DNA nanostructures, or “DNA origami” may provide a cheap framework on which to build tiny microchips, according to a paper published on Sunday in the journal Nature Nanotechnology.

Microchips are used in computers, cell phones and other electronic devices.

“This is the first demonstration of using biological molecules to help with processing in the semiconductor industry,” IBM research manager Spike Narayan said in an interview with Reuters.

“Basically, this is telling us that biological structures like DNA actually offer some very reproducible, repetitive kinds of patterns that we can actually leverage in semiconductor processes,” he said.

Read Full Story
[ad#Adsense]