Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Rambus (RMBS) - Eagle or Penguin -- Ignore unavailable to you. Want to Upgrade?

To: cellhigh who wrote (44023)	6/10/2000 10:31:00 PM
From: Ali Chen	Respond to of 93625

celhi, >>>Rambus interface wasn't robust enough<<<. <please, will you define your idea of robust.> Robust is when you do not need to replace 4-6-chip RIMMs on i840 boards with 16-chip RIMMS, or drop their frequency of operation to 700 or 600 MHz to make the test pass, or use another brand, or.... etc. etc. See examples: Message 13647760 Message 13652178

To: cellhigh who wrote (44023)	6/10/2000 10:46:00 PM
From: multicollinearity	Read Replies (1) \| Respond to of 93625

re: <please, will you define your idea of robust> Hello Cellhigh, I will take a stab at this since I am familiar with the use of the term robustness in mathematical statistics and it seems to be consistent with Bilow's useage. A statistical test is said to be robust if it is not greatly affected by minor deviations from the assumptions upon which the test is based. For instance, a t-test concerning a population mean assumes for its validity that the sampled population has a normal distribution. However, there is little effect on the validity of the test when the sample size is large, and hence for large samples the test is robust with regard to the normality assumption. A second assumption for the test's validity is that the sample is random. This is a critical assumption, and consequently the test is not robust in this regard. Best, Multico

To: cellhigh who wrote (44023)	6/11/2000 12:09:00 PM
From: Dan3	Respond to of 93625

Re: please, will you define your idea of robust. Intel has had a long history of being the most reliable supplier of chipsets and motherboards. They weren't always the fastest, but they were always stable. Since INTC started to incorporate the Rambus interface into its chipsets they've had a series of problems with chipsets and motherboards. This could be coincidence (One of their recent motherboard recalls involved the Intel profusion chipset which doesn't even support Rambus). So maybe Intel just "forgot" how to design motherboards about the same time they started supporting Rambus, but I can't help but think that the two are related. My definition of robust is that the object (CPU, chipset, steel I-Beam, statistical methodology, etc.) will perform satisfactorily even when the external demands placed upon it are somewhat more challenging than expected. For electronic parts, that would include temperature, voltage, uniformity of trace dimensions etc. (I'm not an EE, but I think those are some of the relevant parameters). A robust steel I-Beam has a good sized safety margin built into it. A robust statistical methodology won't give widely varying results with minor changes in the data being analyzed. I wouldn't want to live in a house that will collapse if one more person than expected stands over the I-Beam in the middle of the floor and I don't want to buy a computer that corrupts data if the power supply is 1% out of spec. I wouldn't have paid Rambus prices for memory regardless (we've moved to Win2K and are now specing 256 meg on most new orders), but I'm glad that I've been avoiding the 8XX series chipsets from Intel in recent workstation and server procurements. They don't seem to be robust (at least, not enough for me) Regards, Dan