[MadVikingGod]

So the findings here do make sense. For sub 5m cables directly connecting two machines is going to be faster then having some PHY in between that has to resignal. I'm surprised that fiber is only 0.4ns/m worse then their direct copper cables, that is pretty incredible.

What I would actually like to see is how this performs in a more real world situation. Like does this increase line error rates, causing the transport or application to have to resend at a higher rate, which would erase all savings by having lower latency. Also if they are really signaling these in the multi GHz are these passive cables acting like antenna, and having a cabinet full of them just killing itself on crosstalk?

[Palomides]

high speed links all have forward error correction now (even PCIe); nothing in my small rack full of 40Gbe devices connected with DACs has any link level errors reported

[kazinator]

They looked at the medium itself, not the attached data link hardware.

Look at the graphs. The fiber has a higher slope; each meter adds more latency than a meter of copper.

This is simply due to the speed of electromagnetic wave propgation in the different media.

https://networkengineering.stackexchange.com/questions/16438...

Both the propagation of light in fiber and signal propagation in copper are much slower than the speed of lightin vaccuum, but they are not equal.

[jauntywundrkind]

There's also hollow core fiber, which is pretty close to speed of light in a vacuum. 2.0e8 m/s for fiber, 2.3e8 m/s for copper, pretty close to the full 3e8 m/s for hollow core.

No glass, just some reflective coating on the inside of a waveguide (hollow tube).

https://azure.microsoft.com/en-us/blog/how-hollow-core-fiber...

[timewizard]

> are these passive cables acting like antenna

With both ends connected to a device? No.

Aside from that you've got a linear scrambler into balanced drivers into twisted pair. It's about as noise immune as you can get. Unless you put the noise right up next to the cable itself.

[Hilift]

Storage over copper used to be sub optimal but not necessarily due to the cable. UDP QUIC is much closer to wire speed. so 10 GB copper and 10 GB fiber are probably the same, but 40+ GB fiber is quite common now.

[laurencerowe]

> So the findings here do make sense. For sub 5m cables directly connecting two machines is going to be faster then having some PHY in between that has to resignal. I'm surprised that fiber is only 0.4ns/m worse then their direct copper cables, that is pretty incredible.

Surely resignaling should be the fixed cost they calculate at about 1ns? Why does it also incur a 0.4ns/m cost?

[cenamus]

Light speed is ~3ns per metre, so maybe the lowered speed through the fibre?

Speed of electricity in wire should be pretty close to c (at least the front)

[myself248]

Velocity factor in most cables is between 0.6 and 0.8 of what it is in a vacuum. Depends on the dielectric material and cable construction.

This is why point-to-point microwave links took over the HFT market -- they're covering miles with free space, not fiber.

[jcims]

I always thought it was about reduced path length. Interesting.

[cycomanic]

It's both. Those links try to minimise deviation from the straight link (and invest significant money to get antenna locations to do that), but they also use copper/coax cables for connecting radios as well as hollow core fibre for other connections to the modems.

[laurencerowe]

I misremembered the speed of electrical signal propagation from high school physics. It's around 2/3rds the speed of light in a vacuum not 1/3rd. The speed of light in an optical fibre is also around 2/3rds the speed in a vacuum.

It seems there is quite a wide range for different types of cables so some will be faster and others slower than optical fibre. https://en.wikipedia.org/wiki/Velocity_factor

But the resignalling must surely be unrelated?

[throw0101b]

> Light speed is ~3ns per metre, so maybe the lowered speed through the fibre?

Obligatory Adm. Grace Hopper nanosecond reference:

* https://www.youtube.com/watch?v=si9iqF5uTFk&t=40m10s

[b3orn]

It's c, but not the same c as in air or vacuum. The same applies in optic fibers. They're both around two thirds of the speed of light in vacuum.

[GuB-42]

c is constant, the speed of light is not.

c is the speed of light in a vacuum, but it is not really about light, it is a property of spacetime itself, and light just happens to be carried by a massless particle, which, according to Einstein's equations, make it go at c (when undisturbed by the medium). Gravity also goes at c.

[bigfishrunning]

I've always considered C the speed of light and gravity goes at the speed of light, not that light and gravity both go C, which is a property of spacetime. This is a much simpler mental model, thanks for the simple explanation!

[Sniffnoy]

You can think of c as the conversion rate between space and time; then, light (and anything else without mass, such as gravity or gluons) travels at a speed of 1. Everything else travels at a speed of less than 1.

(Physicists will in fact use the c=1 convention when keeping track of the distinction between distance units and time units is not important. A related convention is hbar=1.)

You can tell that c is fundamental, rather than just a property of light, from how it appears in the equations for Lorentz boosts (length contraction and time dilation).

[Eldt]

I've always thought of c as the speed limit of causality

[Sesse__]

c is the speed of light in vacuum.

EM signals move at about 0,66c in fiber, and about 0,98c in copper.

[BenjiWiebe]

More like 0.6c to 0.75c in Cat6 Ethernet cable.

The insulation slows it down.

[colanderman]

Don't know why you were downvoted, this is true. RF energy is carried primarily (solely?) by the dielectric, not the copper itself, simply by virtue of the fact that this is where the E and M fields (and therefore Poynting vector) are nonzero. It's therefore the velocity factor of the dielectric which is relevant.

[bhaney]

> I'm surprised that fiber is only 0.4ns/m worse then their direct copper cables

Especially since physics imposes a ~1.67ns/m penalty on fiber. The best-case inverse speed of light in copper is ~3.3ns/m, while it's ~5ns/m in fiber optics.

[p_l]

DACs don't cause problems, but twisted pair at 10Gig is a PITA due to power and thermals

[somanyphotons]

What allows DACs to avoid the power/thermal issues that twisted pair has?

(My naive view is that they're both 'just copper'?)

[kijiki]

DACs are usually twin-ax, which is just 2 coax cables bundled. The shielding matters a lot, compared to unshielded twisted pairs.

Faster parallel DACs require more pairs of coax, and thus are thicker and more expensive.

[p_l]

Another reason is that they are shorter range, and the better shielding also means that interference effects are smaller.

In comparison, twisted pair sending 10Gbit over 8P8C cable (popular "RJ-45") requires complex modulation schemes to provide solid signal over any meaningful distance, in a much less shielded cable, and with need to support longer distances.

[tcdent]

PHYs are going away and fiber is going straight to the chip now, so while the article is correct, in the near future this will not be the case.

[sophacles]

The chip has a phy built into it on-die you mean. This affects timing for getting the signal from memory to the phy, but not necessary the switching times of transistors in the phy, nor the timings of turning the light on and off.