Competition British and American scientists have succeeded in discovering one of the most elusive technical challenges in semiconductor science: the Gelsinger co-efficient. The Gelsinger co-efficient is the point at which Intel's VP Pat Gelsinger overheats. Usually in mid-keynote.

The breakthrough could pave the way for cooler, more reliable keynotes in the future.

The scientists are all Register readers, and responded to our appeal to solve the mystery once and for all. As an added incentive - as if being remembered alongside Fermat and Newton isn't enough - we offered a Reg baseball cap to the most creative formula.

The standard of research here was impressive. Pete Freeman from Leeds thought this formula could explain the thermals:-

E=MC^2

Where:-

E = Energy (heat)
M = MegaHurtz
C = Copyright Protection

Right formula, wrong variables Peter, so no cigar, but a consolation prize awaits if you want to get in touch. Matt Collins actually provided the proof:-

Given E=MC^2, it is easy to calculate G:

C is the speed of light. This is well known to be produced by the Sun (S). Intel (I) is in the same business as Sun. Therefore S must be equal to I, which is equal to C.

M is the amount of matter in the Universe. By definition, this must be greater than the amount of matter in Intel (M > I). G, therfore, is be the point when I grows to reach M, causing the heat death of the Universe (by sucking all the energy from the Universe into the first Pentium XVII no doubt), leaving S in its wake.

So, to conclude:

If E=MC^2, then G = S/(MC^2) if S/I is >= E

Brilliant. The prize is shared Dr John Moffett who dates the problem back to the earlier Gelsinger Paradox. Take it away:-

The problem of heat dissipation in microprocessors, known as the Gelsinger Paradox, has resisted analysis due to a lack of formal treatment. Here I present a formal analysis of the Gelsinger Paradox (GP). In brief, the GP states that as microprocessors get bigger, faster and hotter, the companies profits are reduced at a corresponding geometric rate. The solution was provided by a modification of the Gibbs Free Energy equation which is shown below.

^G = ^H - T^S

The change in free energy of a system (delta G or ^G) is equal to the change in heat content (^H) minus the entropy component (T^S).

Modification of this equation provides us with the Gelsinger Free Enterprise Equation:
^G = ^M - C^P

Where ^M is the change in MHz the chip can support during it's expected lifetime, C is the cost of producing the chip, and ^P is the change in price of AMD chips after Intel releases the new processor.

Therefore, the change in Gelsinger Free Enterprise (delta G or ^G ) is equal to the change in MHz minus the cost of the chip, multiplied by the drop in AMD prices.

Gelsinger Free Enterprise is thus a measure of the profits Intel can expect from a particular line of microprocessors. If this equation seems to paint a dim picture for Intel, so be it. Numbers don't lie, people do.

John Rodney Moffett, Ph.D.

Caps are on their way to you as soon as you tell us where you live, gentlemen. ®

Related Stories

Project Jackson - why SMT is the joker in the the chip pack
Chip designers vow to cool overheating Gelsinger
Intel touts Alpha, IBM designs to beat 'hotter than reactor'chips