4 ms·

Why does the demo title within the image say "Gigabit Ethernet"? Is that a marketing lie of what they had hoped to eventually achieve, with the actual demo runn

Why does the demo title within the image say "Gigabit Ethernet"? Is that a marketing lie of what they had hoped to eventually achieve, with the actual demo running 100mbit? Or was the same demo repeated for (what would become) 1000Base-T?

exmadscientist 2y ago
I think those are two different demos, years apart (1995 vs 1998)? You're right that it's sure not clear.
- mindslight 2y ago
  Ah, I had missed the author's caption on that photo dating it at 1998. Looking at it again, it's even more odd because the article was written in 2001 - meaning that the 100Base-T4 anecdote was twice as old as the gigabit demo photo. The part that stands out to me is the author writing off the high frequency loss because each pair is only carrying 25 mbit, yet Gigabit is ten times the bandwidth which is around where I'd think things would start to get a bit wonky. But maybe my intuition is still just the result of single conductor flapping around in free space, rather than a "controlled" impedance balanced pair.
  - exmadscientist 2y ago
    Gigabit brings in all four pairs, so it's only at 125 MHz, five times faster (which is in fact exactly why it does that). That's not all that much worse than 25 MHz. And as for intuition... Howard Johnson is something of the authority on signal propagation, so if he says that's what your intuition ought to be, I listen!
    - mindslight 2y ago
      I'm just saying, scopes have a 20Mhz limiting switch for a reason and that's also around where I start to think something might be better off not on a breadboard. I recognize neither of those things is a controlled impedance pair. It just hits differently to write off 25MHz vs 125MHz as "low frequency".
K0balt 2y ago
I’m not certain, but I believe the demo photo is from WideBand (Roger E Billings CEO). I got certified as a “wideband network administrator “ at their (really) underground bunker headquarters at the International Academy of Science. It was a little bit surreal, but the tech was really cool. We were pushing 1gbps over cat3 to 150 meters, so pretty respectable even today. I think they still are building ultra-low latency switches that are favored by HFT and others that need nanosecond latency switches. 1000basetx eclipsed WideBand and my cert was basically useless, but the experience was really cool and it was great to hang out with a science cult for a while in their vast subterranean lair.

exmadscientist 2y ago

I think those are two different demos, years apart (1995 vs 1998)? You're right that it's sure not clear.

mindslight 2y ago
Ah, I had missed the author's caption on that photo dating it at 1998. Looking at it again, it's even more odd because the article was written in 2001 - meaning that the 100Base-T4 anecdote was twice as old as the gigabit demo photo. The part that stands out to me is the author writing off the high frequency loss because each pair is only carrying 25 mbit, yet Gigabit is ten times the bandwidth which is around where I'd think things would start to get a bit wonky. But maybe my intuition is still just the result of single conductor flapping around in free space, rather than a "controlled" impedance balanced pair.
- exmadscientist 2y ago
  Gigabit brings in all four pairs, so it's only at 125 MHz, five times faster (which is in fact exactly why it does that). That's not all that much worse than 25 MHz. And as for intuition... Howard Johnson is something of the authority on signal propagation, so if he says that's what your intuition ought to be, I listen!
  - mindslight 2y ago
    I'm just saying, scopes have a 20Mhz limiting switch for a reason and that's also around where I start to think something might be better off not on a breadboard. I recognize neither of those things is a controlled impedance pair. It just hits differently to write off 25MHz vs 125MHz as "low frequency".

K0balt 2y ago

I’m not certain, but I believe the demo photo is from WideBand (Roger E Billings CEO). I got certified as a “wideband network administrator “ at their (really) underground bunker headquarters at the International Academy of Science. It was a little bit surreal, but the tech was really cool. We were pushing 1gbps over cat3 to 150 meters, so pretty respectable even today. I think they still are building ultra-low latency switches that are favored by HFT and others that need nanosecond latency switches. 1000basetx eclipsed WideBand and my cert was basically useless, but the experience was really cool and it was great to hang out with a science cult for a while in their vast subterranean lair.