The Last POWER1 CPU on Mars Is Dead • Hacker News Board

Re: The Last POWER1 CPU on Mars Is Dead

#41

Post by wlesieutre » Mon, Feb 25, 2019, 9:17 PM UTC

teraflop wrote: post #32

wlesieutre wrote: post #18

Earlier quoted context omitted.

At a certain point of accuracy you run into issues with this - seconds don't pass at the same rate due to relativistic effects (both speed and gravitational). An example we already deal with is GPS satellites orbiting Earth. Their clocks tick a little bit slower on account of how fast they're orbiting, and tick a bit faster on account of being further up Earth's gravity well. The gravitational effect is stronger, and…

I don't think the relativistic effects would be a problem in practice. You can synchronize clocks on other planets to the UTC or TAI reference time, which are already specified in terms of the Earth's reference frame. Because of relativity, clocks on other planets would very slowly drift relative to UTC on Earth. But the drift is on the order of a few parts per billion (see e.g. [1]) which is comparable to what you'd…

If I did my math right that's 0.178 seconds per year. As long as everyone's handling time the same way you shouldn't have a problem, but it's enough to be noticeable if someone forgets about it.

I suppose the more noticeable oddness for most people would be that the speed-of-light delay in communications from Earth to Mars varies so much. Around 3 light minutes up to 22 depending on where they are in their orbits.

Re: The Last POWER1 CPU on Mars Is Dead

#42

Post by wbl » Mon, Feb 25, 2019, 9:24 PM UTC

echlebek wrote: post #34

twodave wrote: post #24

Earlier quoted context omitted.

Well the problem is a "second" is just a unit of rotation of the Earth on its axis. So you'd either need to use some other reference material than a rotating body in space or else use some arbitrary one (like Earth, since that's where we are from!) and choose a reference date (why not use one we're all familiar with? I know! We could use the UNIX epoch!). Even if you did this, your universal time would just be a sync…

That's actually not true if you're using SI units for seconds. "The SI unit of time is the second (s): The second is the duration of 9 192 631 770 periods of the radiation corresponding to the transition between the two hyperfine levels of the ground state of the caesium 133 atom." I'm not a physicist, but I think that's a universal definition - the rate of the periods should be constant everywhere in the universe. S…

You are forgetting relativistic effects.

Re: The Last POWER1 CPU on Mars Is Dead

#43

Post by yingw787 » Mon, Feb 25, 2019, 9:24 PM UTC

DannyB2 wrote: post #40

yingw787 wrote: post #5

Might be O/T, but I found it really interesting how they mention "5,352 Martian solar days". Do we humans have plans in the pipeline for a universal time standard we can use wherever we are? Something that is a little more useful than "seconds since UNIX epoch" but something a little less terrestrial than UTC? Or is time going to be based wherever we land or travel (e.g. divide up Mars into time zones upon colonizati…

According to relativity, you cannot get two different observers to agree on any kind of universal time, or even the order in which two events occurred. Observer A says that event X happened before Y. Observer B says that event Y happened before X. And they're both right from their POV.

Hmm...would this only be relevant on the scale of nano/picoseconds, or seconds or higher granularity, given a frequent synchronization schedule?

Re: The Last POWER1 CPU on Mars Is Dead

#44

Post by greiskul » Mon, Feb 25, 2019, 9:26 PM UTC

DannyB2 wrote: post #40

yingw787 wrote: post #5

Might be O/T, but I found it really interesting how they mention "5,352 Martian solar days". Do we humans have plans in the pipeline for a universal time standard we can use wherever we are? Something that is a little more useful than "seconds since UNIX epoch" but something a little less terrestrial than UTC? Or is time going to be based wherever we land or travel (e.g. divide up Mars into time zones upon colonizati…

According to relativity, you cannot get two different observers to agree on any kind of universal time, or even the order in which two events occurred. Observer A says that event X happened before Y. Observer B says that event Y happened before X. And they're both right from their POV.

You are correct, but thankfully relativity tells us that in this case both events are independent, so neither can be the cause of the other, so it's not that bad of a disagreement.

It's even useful as a consistency model for some distributed systems.

Re: The Last POWER1 CPU on Mars Is Dead

#45

Post by gmueckl » Mon, Feb 25, 2019, 9:27 PM UTC

tomp wrote: post #23

hermitdev wrote: post #14

Earlier quoted context omitted.

The problem comes in due to relativity. Mars and Earth (and other planetoids) travel at different speeds. The very length of a second will drift over time.

Really? Can't you just anchor to e.g. one revolution of the planet? That should be synchronizable...

No. You would have to observe the planet from your position and that observation changes based on how you move relative to the planet. The whole of SRT and GRT work in such a way that no general clock synchronization can ever exist for observers in different frames of reference.

If you take a reference with you on a journey through the solar system, that reference will stay accurate for an observer travelling with it and drift as seen from an observer that stayed behind. Both views are accurate. Locally, the amount of time that has passed for each observer was different. If you would correct for any observed drift, you would mess with your fundamental unit of time in your local frame of reference and get incorrect measures of time passed in your own frame.

Re: The Last POWER1 CPU on Mars Is Dead

#46

Post by hermitdev » Mon, Feb 25, 2019, 9:27 PM UTC

tomp wrote: post #23

hermitdev wrote: post #14

Earlier quoted context omitted.

The problem comes in due to relativity. Mars and Earth (and other planetoids) travel at different speeds. The very length of a second will drift over time.

Really? Can't you just anchor to e.g. one revolution of the planet? That should be synchronizable...

No, you can't. Atomic clocks measure the oscillation of atoms (I think Cesium is currently used), and though constant at a fixed point, vary when theyre moving at different speeds. GPS wouldn't be possible without taking this into account, for example. A second on earth will never be the same as a second on Mars.

Re: The Last POWER1 CPU on Mars Is Dead

#47

Post by simcop2387 » Mon, Feb 25, 2019, 9:27 PM UTC

yingw787 wrote: post #43

DannyB2 wrote: post #40

Earlier quoted context omitted.

According to relativity, you cannot get two different observers to agree on any kind of universal time, or even the order in which two events occurred. Observer A says that event X happened before Y. Observer B says that event Y happened before X. And they're both right from their POV.

Hmm...would this only be relevant on the scale of nano/picoseconds, or seconds or higher granularity, given a frequent synchronization schedule?

For any speeds we'd be likely to be traveling, in system, it'd likely only be a problem for milliseconds if that. We'd still be able to correct for where things are as we approach them or plan to approach them.

Re: The Last POWER1 CPU on Mars Is Dead

#48

Post by hackbinary » Mon, Feb 25, 2019, 9:29 PM UTC

JumpCrisscross wrote: post #9

yingw787 wrote: post #5

Might be O/T, but I found it really interesting how they mention "5,352 Martian solar days". Do we humans have plans in the pipeline for a universal time standard we can use wherever we are? Something that is a little more useful than "seconds since UNIX epoch" but something a little less terrestrial than UTC? Or is time going to be based wherever we land or travel (e.g. divide up Mars into time zones upon colonizati…

> Do we humans have plans in the pipeline for a universal time standard we can use wherever we are? This fundamentally cannot exist. Any "universal" system would drift for observers on different planets. Having different times in New York and New Delhi isn't the worst right now; doing a calendar look-up for the time on Mars isn't as neat. But given we won't have real-time communication, it isn't as big of a problem e…

I thought a second was a SI base unit: (from Wikipedia)

The second, symbol s, is the SI unit of time. It is defined by taking the fixed numerical value of the caesium frequency {\displaystyle \Delta \nu _{\text{Cs}}} {\displaystyle \Delta \nu _{\text{Cs}}}, the unperturbed ground-state hyperfine transition frequency of the caesium-133 atom, to be 9192631770 when expressed in the unit Hz, which is equal to s−1.

Re: The Last POWER1 CPU on Mars Is Dead

#49

Post by dooglius » Mon, Feb 25, 2019, 9:30 PM UTC

mattkrause wrote: post #38

dooglius wrote: post #15

Earlier quoted context omitted.

There is BCT [0] which corrects for gravitational time dilation on Earth relative to the solar system, but I'm not aware of anything on a universal scale. It seems conceptually difficult to do synchronization if the communication round trip time, bounded below by the speed of light, is significant. [0] https://en.wikipedia.org/wiki/Barycentric_Coordinate_Time

I think NTP uses some statistical tricks to get better synchronization than you'd expect over slow, laggy internet links. Possibly something like that could work.

If we're talking interstellar travel, we're likely talking about spacecraft undergoing very significant acceleration and velocity. I'm not a physics expert, but from my understanding it seems that a starship would need to have a very accurate idea of its own acceleration vs reference time curve (relative to some reference frame) in order to calibrate its own ticking clock to the reference frame. For NTP, this isn't so much an issue because on Earth all the endpoints' clocks are all ticking at about the same rate.

Re: The Last POWER1 CPU on Mars Is Dead

#50

Post by duckerude » Mon, Feb 25, 2019, 9:31 PM UTC

DannyB2 wrote: post #40

yingw787 wrote: post #5

Might be O/T, but I found it really interesting how they mention "5,352 Martian solar days". Do we humans have plans in the pipeline for a universal time standard we can use wherever we are? Something that is a little more useful than "seconds since UNIX epoch" but something a little less terrestrial than UTC? Or is time going to be based wherever we land or travel (e.g. divide up Mars into time zones upon colonizati…

According to relativity, you cannot get two different observers to agree on any kind of universal time, or even the order in which two events occurred. Observer A says that event X happened before Y. Observer B says that event Y happened before X. And they're both right from their POV.

My understanding is that it's just that the order can depend on the frame of reference. If you agree on a frame of reference then you can agree on the time. If you both try to use your own frame of reference then you run into trouble.

There's no preferred frame of reference, meaning that any frame of reference could work, but you can derive one from the cosmic microwave background if you don't like picking something too arbitrary.