Rambus: You're terminated *!@$ – Intel
Always wondered what those *!x£$ meant
As an innocent young bairn brought up in bonny Scotland we were always puzzled when we came across **** in a piece of text. Did adults have some dark secret they weren't sharing with their children?
Not as perplexed as when every front page of every UK national newspaper carried big white blanks when Britain and France invaded Suez in 1956. Question: "Is this called censorship, dad?" Answer: "No, it's called freedom of the press, son." But still pretty puzzled.
So a close examination of the original SEC filing which describes the nature of the relationship between Intel and Rambus Ink at first had us thinking that both companies were cursing at each other.
The document, a filing which you can find at this incredibly long URL, outlines everything you'd need to know, bar the asterisks, about the wheeler-dealing between INTC and RMBS, and thanks to a reader (thanks, David) who took up our ServerWorks challenge yesterday, it is now possible to state that if they weren't cursing then, they're certainly cursing and Intel at least, is sweating now.
We really can't do better than quote what our reader makes of this document, which you'll need to refer to, ignoring the asterisks. (Send in your invoice, David).
"Although some of the details have been removed for confidentiality, there is still some good dirt there. Some of the interesting tidbits are the "Success Determination Date", which has to come to pass before Intel gets to buy their four million shares for $2.50 a piece."
1.33 Success Determination Date means the date on which all of the following are satisfied: (i) twelve (12) months after the sale in the normal course of business, in aggregate by all manufacturers thereof, of one million (1,000,000) Rambus-D DRAMs, (ii) twelve (12) months after [***] (iii) a total of six (6) suppliers are able to ship one million (1,000,000) units per month Rambus-D DRAMs, and (iv) for at least three (3) of such six (6) suppliers, the cost of Rambus-D DRAM is within five percent (5%) of the cost of 100MHz 4Mbitx16 SDRAM manufactured on the identical process.
"I personally won't hold my breath waiting for (iv). More interesting is the "Termination" section of the contract."
(b) Rambus, at its option, may, in addition to any other remedies it may have, terminate this Agreement on written notice to Intel if (iv) at any time between January 15, 1997 and the Success Determination Date, Intel communicates to any of the then current top ten (10) DRAM manufacturers that Intel has plans to support, as the primary DRAM for PC main memory applications for the years 2000, 2001 and 2002, any New Interface other than the Rambus-D Interface Technology; or (v) at any time between January 15, 1997 and the Success Determination Date, the Intel senior member attending the quarterly Rambus/Intel executive meeting, upon Rambus' request, does not represent to the Rambus officers attending such meeting that the Rambus-D DRAM will be the primary DRAM for PC main memory applications for the years 2000, 2001 and 2002.
(i) For purposes of this Subsection (b), "New Interface" means any interface for PC main memory applications, other than main memory interfaces on Intel's chipsets shipped prior to the second calendar quarter of 1998, and evolution of such main memory interfaces extending therefrom. Any DRAM interface which provides greater than one (1) gigabyte/second/device bandwidth is considered a New Interface.
"There you have it in writing why Intel is so gung-ho on Rambus. If they don't say that Rambus is their primary memory technology, Rambus can take away their licence. Without a license, they can't ship i820 or i840 chipsets. So, Intel has to choose between being a Rambooster, or completely eliminating Rambus from their lineup.
"Personally, I can't imagine what kind of sweetheart deal Intel thought they were getting to sign up for a one-sided contract like this." ®