Can We Afford a Revolution?

By innovating withing the framework of the existing infrastructure.

Exactly!

Or maybe stop writing applictions so badly that they constantly need faster and faster processors with more and more memory!

Some one should have noticed by now that bigger and faster computers are not necessarilly producing any better results.

Slaps furry head with paw!

Del

As badly written software or protocols can indeed be attributed to part of the constant need for greater speed and memory, some need does arrise from the ever growing demand on those chips.

Costs come down not because the market dictates but because there is a manufacturer that supplies for cheaper. Why would this supplier rule the market if it were not up to the job at hand. The consumers or oem, will then respond and set the new price as a target, that is natural and goes for any industry.

Personally I do not see any different from any other industry where prices drive some out of the market, I just experienced that myself where the influx of skilled cheap labour from the eastern block countries fill the markets with products or services. I am now employed again after I have been a director of my own company for close to 15 years, forced to close it down as I did not get the level of work needed to sustain it. That is capitalism and free market. It works, or so they tell me.

P.s. I was not aware that a trend can become a law! I always thought that a law had to listen to some universal -mathematical equasion that proves it al the time no matter how or where you repeat the test. This is just a trend right, why call it a law?

Me confused, never!

The inhibiting factor is facilities cost.

I am the tool install coordinator for Photolithography and Implant for one of the worlds largest chip makers. I have worked on these projects since '96. I have been involved with some projects where there was just a grass patch and some heavy equipment ready to break ground to where there are 2 running fabs with a process refit/retool taking place.

The types and amounts of chemicals required (both aqueous and gaseous) is staggering. The water and power consumption at these factories is second to NONE. The water filtration and purification systems would rival any seen elsewhere. The purity of the chems and gases is extremely hard to maintain as well. Purity levels of 99.99999%. Fewer than 1 particle @ 1 micron per 1 scfh @ turbulent flow per line (and there are many thousand lines), trace gas concentrations (O2 & H2O) of less than 10 parts per trillion. Every line tested as by pressure decay test and most tested by helium leak detection inboard method.

The gases and chems themselves are some of the most toxic, corrosive, flammable, pyrophoric, and potentially lethal in use by any industry and number, again, in the many hundreds if not thousands. Gases such as phosphine and silane and chems such as Hydroflouric acid and tetra-ethyl-orthosilicate are legendary for health and hazard, but they are just a few of a great number. These are piped through miles of ultra-high purity piping systems utilize GTAW (orbital) welding of 316L SS with highly controlled chemical alloy composition and internal finishes controlled at 5ra surface roughness or less as well as minimal allowable wall thickness deviations. Then there are materials and joining processes such as PVDF and IR fusion, double contained PFA and PTFE. Monitoring system atop monitoring system.

This all takes place under (basically) one roof. One very, very clean roof. The internal environment of a chip factory is one of the most controlled environments in the world - especially when considering the square footage involved.

A completely independent power substation is required even for a modest facility. Now a fully capable 100% power back up system, double fail safe. Hundreds of thousands and thousands of Volts constantly being drawn.

The permitting required for a company to use/control/contain the materials used is more involved than a declaration of war - seriously.

It is a very delicate set of systems. Having to be engineered as a whole and in part to withstand weather and disaster (hurricanes and earthquakes especially).

The tools that I am responsible for make up appx 1/8 of the entire fab tool set (yet are the most expensive). 1 of the tools costs $50+ million dollars on down to .8 million and there are a bunch of 'em at these prices and between. This cost is for the tool itself. In my work area I am managing a budget with a labor and consumables cost in excess of $50,000/day for the installation. Then there are operational costs once the product hits.

A modest failure, let alone a significant one, can have a negative cost impact (loss of $) that equals more than some states annual health care costs and can result from any one of a thousand foreseen or otherwise unforeseen events.

So here is the point. The cost of building the facility is in the billions because that is how much it costs to acquire, support and utilize the technologies involved in reliable chip manufacture. It is not going to go down much quickly. The company I work for now, has done more than any other to reduce these costs but it is still a limiting factor for 99% of businesses. The safety and health liabilities are the second driving factors and then of course personnel. Not just any company can come in and build a multi-billion dollar super-clean facility in the amount of time that the proven contractors can and that is gonna cost 'ya as well.

So a consortium or similar is the only way I see it happening. There are facilities where ideas, processes and technologies are tested but they are few and far between and are typically an odd arrangement between competitors, universities and others - again deep pockets.

From where I sit it seems that what we have is status-quo for a while to come.