WTB: 2.1GHz Dothan (SL7V3) For a ThinkPad T40 Upgrade

[svet-am]

AND FYI, your RAM IC analogy is completely flawed: 99.9999% of the time what fails in RAM DIMMs is the PCB and soldered resistors, NOT the ICs. This is why several guys (some even on this forum) make an ABSOLUTE FORTUNE buying "bad" DIMMs, removing the chips, and soldering them on to new boards. Occasionally you get a bad chip, but most of the time it is either a PCB or solder problem. This is also why the failure rate of memory for large OEM companies like Micron is much lower than say Kingston - - - EVEN when both companies use the EXACT same chips. The OEM companies invariably have much better PCBs.

before you start throwing out statistics like 99.9999% of the time, please buy a copy of the book "The Art of Electronics" (or "Fundamentals of Electric Circuits" by Alexander and Sadiku) and read up on how transistors actually work at the silicon level. Particularly the specifics of the proper operating range for CMOS transistors and their sensitivity to voltage spikes damaging their gates nodes.

My example didn't say "in the absence of voltage or spikes" (as you did), my example was DIRECTLY related to voltage spikes (largely from a static electricity discharge). A lot of people handle ICs without antistatic straps or without storing them in a proper anti-static environment. This is bad for ASICs and other VLSI designs. And, sadly, most people don't even realize they ever damaged the ICs.

In your example of perfect conditions, then you're absolutely right. A chip just sitting in a computer and not being turned on will last practically forever. If we had an original IBM XT that had been sitting in a box since the early 80s and we decided to de-box it and plug it in then it'd work just fine. But, that's not what I was discussing. I was discussing specifically the case of a chip being damaged by (accidental) electrical discharge.

As to the differences in Micron and Kingston RAM, that is NOT directly related to the quality of the PCB or the resistors that they choose (arguably, the routing on the board and the actual circuit layout is more to blame for issues than PCB type or passive components). While the chips may be labeled the same, they're NOT the same. Micron (Crucial, really) and other high end vendors pay more to buy ICs that have passed manufacturability tests more rugged and stringent than the ICs that get binned for lower tier OEMs like Kingston.

[svet-am]

Look, we've had some lively exchanges but the more I think about it the more I think that you were referring to one thing while I was referring to another. That is, we were talking past each other.

Bottom line is that this is a FS/WTB thread. I'd be more than happy to continue this conversation in an appropriate thread in the appropriate sub-forum on this site. Start the thread and I'll follow.

Let's just not clog up the FS forum with our disagreement.

[brentpresley]

AND FYI, your RAM IC analogy is completely flawed: 99.9999% of the time what fails in RAM DIMMs is the PCB and soldered resistors, NOT the ICs. This is why several guys (some even on this forum) make an ABSOLUTE FORTUNE buying "bad" DIMMs, removing the chips, and soldering them on to new boards. Occasionally you get a bad chip, but most of the time it is either a PCB or solder problem. This is also why the failure rate of memory for large OEM companies like Micron is much lower than say Kingston - - - EVEN when both companies use the EXACT same chips. The OEM companies invariably have much better PCBs.

before you start throwing out statistics like 99.9999% of the time, please buy a copy of the book "The Art of Electronics" (or "Fundamentals of Electric Circuits" by Alexander and Sadiku) and read up on how transistors actually work at the silicon level. Particularly the specifics of the proper operating range for CMOS transistors and their sensitivity to voltage spikes damaging their gates nodes.

My example didn't say "in the absence of voltage or spikes" (as you did), my example was DIRECTLY related to voltage spikes (largely from a static electricity discharge). A lot of people handle ICs without antistatic straps or without storing them in a proper anti-static environment. This is bad for ASICs and other VLSI designs. And, sadly, most people don't even realize they ever damaged the ICs.

In your example of perfect conditions, then you're absolutely right. A chip just sitting in a computer and not being turned on will last practically forever. If we had an original IBM XT that had been sitting in a box since the early 80s and we decided to de-box it and plug it in then it'd work just fine. But, that's not what I was discussing. I was discussing specifically the case of a chip being damaged by (accidental) electrical discharge.

As to the differences in Micron and Kingston RAM, that is NOT directly related to the quality of the PCB or the resistors that they choose (arguably, the routing on the board and the actual circuit layout is more to blame for issues than PCB type or passive components). While the chips may be labeled the same, they're NOT the same. Micron (Crucial, really) and other high end vendors pay more to buy ICs that have passed manufacturability tests more rugged and stringent than the ICs that get binned for lower tier OEMs like Kingston.

ROFL.

Dude - Micron MAKES their own ICs. And the use them ON THEIR OWN branded memory.

They also SELL the EXACT SAME ICs to Kingston and other memory PACKAGERS (NOT MANUFACTURERS - they don't MAKE memory, they simply PACKAGE it onto a PCB).

And if you think the PCBs that Micron uses are of the same quality as those that Kingston and other packagers use, then:
1) you didn't read the book you are trying to get us to read
2) you might as well be trying to sell us OCEANFRONT property in KANSAS


There is a HUGE difference in memory quality and the VAST majority of it is due to PCB differences (i.e. 4-layer vs. 6-layer boards, and post packaging testing).

[brentpresley]

Look, we've had some lively exchanges but the more I think about it the more I think that you were referring to one thing while I was referring to another. That is, we were talking past each other.

Bottom line is that this is a FS/WTB thread. I'd be more than happy to continue this conversation in an appropriate thread in the appropriate sub-forum on this site. Start the thread and I'll follow.

Let's just not clog up the FS forum with our disagreement.

Agreed. If we are going to talk about the financial aspects of your purchase, I think you would have been better saving $100-$150 and going for the 2.0.

As someone that had a 2.1:
1) you can't tell the difference
2) the thermals of the 2.1 are disproportionately higher - to the point that they often don't fit within the cooling envelope of the T4x series (particularly noticeable on a HOT day).

[wearetheborg]

Agreed. If we are going to talk about the financial aspects of your purchase, I think you would have been better saving $100-$150 and going for the 2.0.

Heck, the OP might have been better off getting a 2Ghz T42, instead of upgrading...The increased bus speed might iteslf have made it faster than a 2.1 Ghz @ 400Mhz

[brentpresley]

Agreed. If we are going to talk about the financial aspects of your purchase, I think you would have been better saving $100-$150 and going for the 2.0.

Heck, the OP might have been better off getting a 2Ghz T42, instead of upgrading...The increased bus speed might iteslf have made it faster than a 2.1 Ghz @ 400Mhz

FSB is the same on the T40 and T42 (400MHz).

Only the memory bus is different (266MHz vs. 333MHz). And honestly, you are rarely memory bus constrained on a laptop. The difference is hard to notice.

[ccotenj]

he probably would have gotten more functional "speed" by going to a 7200 rpm drive... for less than 200 bucks, he could have done that and gotten a 1.8, and been pretty much set (not to mention avoided potential frying issues)...

but he wanted a 2.1, and now he has one... the decimal equivalent of lighting your cigar with 100's imo, but it's his money, and it was burning a hole in his pocket... of course, he may burn a hole in his thinkpad as well...