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.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: Petz who wrote (26411)	12/4/1997 12:05:00 AM
From: Yousef	Read Replies (1) \| Respond to of 1572472

John, Re: "Tell me, how many 3.5 volt Pentiums do you know of that failed due to gate oxide failure?" John, you had better stop while you have "some" credibility left ... you are way out of your technical experience here. Intel's process for 3.5V Pentiums uses a much thicker gate oxide and thus there is no leakage problem. Also, this leakage issue has nothing to do with how much current is drawn by the chip ... again, don't embarrass yourself!! Make It So, Yousef

To: Petz who wrote (26411)	12/4/1997 12:24:00 AM
From: Elmer	Respond to of 1572472

<Tell me, how many 3.5 volt Pentiums do you know of that failed due to gate oxide failure?> Petz, Intel's 3.5v parts were made on a .6u process with thicher gate oxides. They could take the higher voltage. AMD is risking failure because they are in such a desperate position that have nothing to lose. I am still waiting for your apology, but that would take integrety on your part, I guess I shouldn't hold my breath. EP

To: Petz who wrote (26411)	12/5/1997 5:07:00 AM
From: Jeff Fox	Respond to of 1572472

John, re: What, no such thing as Flash ???? "current through insulator" - total double talk nonsense Too bad you weren't around 25 years ago to tell Dov Frohman about this. Perhaps he would have taken your council and not invented the EPROM. Then we could have forgone the EEPROM and Flash memory as well and saved AMD a lot of money along the way. All the above products rely on charge storage on floating polysilicon gates. The charge is applied or removed through the gate oxide. Charge moving is called current in the trade, thus there is current through gate oxide. If you check out EPROM literature you will find papers describing limited write cycles. This was as little as 10,000 cycles in past days and is still usually spec'ed at under one million cycles. Why? Energetic electrons in the programming currents damage the gate oxide. Eventually a short will develop rendering the chip useless. The same physics is true for processors, but .35 the micron logic process has gate oxide much thinner than an EPROM process, and processor gates toggle millions of times per second. For .35 micron and lower gate oxide integrity has overtaken source-drain punch-through as the principle limiter on supply voltage. The nominal "safe" Vcc was set for .25 volts (nominal) for .35 generation. With top quality manufacturing (i.e. defect free oxides), this was pushed to .28 volts. Anything greater results in "stepping over the cliff", with greatly shortened operating life. Like I said before, I don't have privy to AMD's stellar processing expertise. Perhaps they have overcome this physics. I am just saying that if I were Compaq I would want to see lots details before putting my rep on the line. If I were a small customer without the resources to qualify the part I would be just very nervous. Tell me, what is Intel's warranty on its non-boxed processors? One year I believe. Tell me, how many 3.5 volt Pentiums do you know of that failed due to gate oxide failure? None that I know of. These Pentiums were made with .6 micron technology with much thicker gate oxides. If the problem is current, you better worry about the Pentium II, which draws 30% more current AT THE SAME CLOCK SPEED as a K6. You might be confusing this with operating current which is not at all related. The gate leakage currents are extremely small. It is just that these currents should be zero to avoid gate oxide damage. The PII supply current consumption isn't a problem as long as their is sufficient heat removal. I understand that the SLOT 1 cartridge does a great job at heat dissipation. Jeff