Who are you?

[giza]

Looking through my "Friend Of" list, I see a whole bunch of folks who have me listed, and I have no idea who some of you are. In particular, I don't believe I know any of the following people: faekitty71, felislunae, firestormsix, hiromi_kitsune, jfbat, keokawa, ketrava, kimono_skunk, marnanel, pardouncia, snaury, thorfinn, tigerman, turrentwolfie, and wolfylee. I was wondering if perhaps y'all would be kind enough to leave a comment telling me how it was that you came across my LJ? (And, if we've met before and this silly leopard just doesn't remember!)

On an unrelated note, here are some articles I found of interest over the last few days:

Using BGP Anycast so that multiple physical servers can respond to the same IP address - This technique is used for global services such as UltraDNS and some of the root nameservers.

Jamie Zawinski's Tent of Doom - How to make a simple office cubicle into a more exciting place.

Tough Computers - A bit of a fluff piece that explains a little about how/why special computers with radiation hardened components are put in interstellar and extraterrestial probes, such as the Mars Lander.

Advanced Vehicle Automation and Computers Aboard the Shuttle - This is a much more detailed piece that goes into lots of detail about both the hardware and software aboard the Space Shuttle. Here's a hint: the computer in any car made in the late 90's or later is probably more powerful than the computer that is in the Space Shuttle.

Comments

Radiation hardening microprocessors.

[cjthomas.livejournal.com]

Here's a bit more info, in case you're interested. I'm a university-grade chip geek, so I hear about this kind of thing =^.^=.

• Two broad classes of radiation failure exist - sudden events that cause catastrophic damage, and effects that build up over time and make the chip stop working.
  
  
• Sudden events are all caused by an ion or high-energy photon knocking electrons free within the chip substrate, which in turn cause a shower of secondary electrons. When all is said and done, you get a considerable amount of free charge in a small volume of the chip, and conduction where there might have been an insulating barrier previously (free charge can move). Spot-conduction across barriers can cause latch-up, and dumping a lot of charge in one place can cause voltage gradients and currents large enough to damage things.
  
  
• Latch-up is when a pair of parasitic bipolar transistors formed by CMOS structures activate and enhance each others' current flow. This causes currents large enough to damage the affected region of the chip, and also pegs voltages at strange values in the area.
  
  This is avoided in commercial chips by putting substrate contacts that nail down voltages to what they're supposed to be (preventing the latch-up reaction from happening or stopping it if it starts), and in many families of rad-hard chips by using silicon-on-insulator technology, which doesn't have the parasitic bipolars in the pattern needed to cause latch-up.
  
  
• You can reduce the effects of charge generation by sprinkling contacts around also, to suck away excess charge generated and reduce voltage gradients within the substrate (which can cause circuit strangeness and trigger latch-up even when the currents aren't large enough to damage things).
  
  
• Damage that builds up over time usually stems from dislocated atoms within the substrate's lattice. This can be the result of either direct strikes or from secondary electrons. These dislocations are areas that can pin charge or otherwise interrupt current flow. The most important form of damage occurs when these flaws are generated on the surface of (or even in) the oxide layer under the chip gate.
  
  
• Buildup of charge in the gate oxide, either due to charge being dumped into the insulator directly or fixing itself to flaws in the oxide/substrate interface, shields the channel from the voltage on the gate. This makes a transistor less responsive to changes in gate voltage, usually raising the threshold voltage. Once this shifts enough, the transistor can no longer turn on.
  
  I believe you can get transistors pegged on instead of off as well, but it's been a while since I read up on the details.
  
  This effect is mainly just reduced by making the gates a smaller target, which process shrinks do for us for free.
  
  
• Lastly, as the article said, chips intended to function in hostile environments for extended periods are usually made redundant (and used redundantly), so that any single failure won't take down the system.
  

And that's my pointless geeking for today =^.^=. Enjoy.

[turrentwolfie.livejournal.com]

I'm a friend of Qs and i thoguth i'd read your jounal after a while of you posting :D im also frank firestars friend

[rebelsheart]

hiromi_kitsune was once known as RobFox. kimono_skunk used to be a roomate to cerisewolf and vesuvius.

me

[ketrava.livejournal.com]

This is ketrava. I met you at Feral one year. I am pretty sure you were there this most recent year I am the biker kitty. well its time for sleep here. night.

[tigerman.livejournal.com]

or .

But I contacted you on IM about running a yahoo spirituality group. blackfeathertanfur referred me to you.

[tigerman.livejournal.com]

[faekitty71.livejournal.com]

Friend of Q, Foxiecrumz, Sprocket and three time Anthrocon attender. Meow! JFbat is my main squeeze.

[firestormsix.livejournal.com]

Hi Giza, I met you somwhere,...i can't remember if it was AC, Feral, or where, but i remember you from somwhere anyhow. Visit my main art site to see if anything looks familliar, if not, i guess you know alittle about me after you visit
Stormfire Studio.ca. here: http://www.catbox.com/firestorm/
or Northern Furs.ca home forum, here: http://www.northernfurs.ca/ff/index.php

[marnanel.livejournal.com]

I forget, tbh. Maybe a PLUG thing?

Re:

[giza.livejournal.com]

Philly actually HAS a Linux Users group? Holy crap! Where/when?

I was involved with a Python Users Group (http://emergent.brynmawr.edu/index.cgi/PythonUsersGroup) while ago, but that sorta died out... :-\

Re:

[marnanel.livejournal.com]

http://www.phillylinux.com

It was probably the Python thing then. I remember looking at a Philly Python wiki.

[marnanel.livejournal.com]

oops, dot org. But it redirects anyway