Overclocking with more clocks

  • Thread starter Thread starter Curious
  • Start date Start date
C

Curious

Is it possible to use a billion 1Hz clocks to make a clock rate of 1 GHz?

How would this be done?


Thanks
 
Curious said:
Is it possible to use a billion 1Hz clocks to make a clock rate of 1 GHz?

How would this be done?
Click to expand...

Most 1Hz "clocks" use a higher frequency standard, and divide down from
there. If the frequency standard was 1MHz for example, one could
multiply it 1000 times to get 1GHz at the same standard frequency base
as the 1Hz "clock". There are many kinds of multiply and divide circuits.

Virg Wall
 
No.

Not really possible, you could make lower frequencies eg 1/2, 1/4, 1/8 Hz
but not higher ones (well you might manage 2 and 4 Hz possibly)
 
half_pint said:
No.

Not really possible, you could make lower frequencies eg 1/2, 1/4, 1/8 Hz
but not higher ones (well you might manage 2 and 4 Hz possibly)
Click to expand...

Nerves can fire up to 1 KHz max. At pitches above 1 KHz our cochlea
makes-up for this by firing different lines of neurons at succesive
cycles.


http://www.cns.nyu.edu/~msl/courses/0022/lecturenotes/pitch/pitch.html

"Volley Principle: The volley principle reconciles the fact that the
cochlear microphonic mimics the sound pressure waves with the
implausibility of the temporal code. Wever suggested that while one
neuron alone could not carry the temporal code for a 20,000 Hz tone,
20 neurons, with staggered firing rates, could. Each neuron would
respond on average to every 20th cycle of the pure tone, and the
pooled neural responses would jointly contain the information that a
20,000 hz tone was being presented."

"Phase Locking is an empirical observation that supports the volley
principle. When 8th nerve neurons fire action potentials, they tend
to respond at times corresponding to a peak in the sound pressure
waveform, i.e., when the basilar membrane moves up. The result of this
is that there are a bunch of neurons firing near the peak of each and
every cycle of a pure tone. No individual neuron can respond to every
cycle of a sound signal, so there must be different neurons firing on
successive cycles. Nonetheless, when they do respond they tend to fire
together."


Is it possible design an overclocking scheme similar to that of the
cochlea?
 
Curious said:
"half_pint" <[email protected]> wrote in message

Nerves can fire up to 1 KHz max. At pitches above 1 KHz our cochlea
makes-up for this by firing different lines of neurons at succesive
cycles.


http://www.cns.nyu.edu/~msl/courses/0022/lecturenotes/pitch/pitch.html

"Volley Principle: The volley principle reconciles the fact that the
cochlear microphonic mimics the sound pressure waves with the
implausibility of the temporal code. Wever suggested that while one
neuron alone could not carry the temporal code for a 20,000 Hz tone,
20 neurons, with staggered firing rates, could. Each neuron would
respond on average to every 20th cycle of the pure tone, and the
pooled neural responses would jointly contain the information that a
20,000 hz tone was being presented."

"Phase Locking is an empirical observation that supports the volley
principle. When 8th nerve neurons fire action potentials, they tend
to respond at times corresponding to a peak in the sound pressure
waveform, i.e., when the basilar membrane moves up. The result of this
is that there are a bunch of neurons firing near the peak of each and
every cycle of a pure tone. No individual neuron can respond to every
cycle of a sound signal, so there must be different neurons firing on
successive cycles. Nonetheless, when they do respond they tend to fire
together."


Is it possible design an overclocking scheme similar to that of the
cochlea?
Click to expand...

I think the trouble is you need a 1 ghz clock to syncronise the waves.
So yes, you can make a 1 ghz clock, *if* you already have one.
 
Yes, it is the offset that is the bear... Plus, I assume you are referring
to the cheap xtal oscillators...they DO need to be run for around 168 hours
(aged) before they become fairly stable.
 
You'd basically be time division multiplexing them (i.e., each would have
it's own time slot for it's pulse...it isn't practical, but theoretically
possible). You'd also have to shorten the pulses appropriately.
 
George said:
You'd basically be time division multiplexing them (i.e., each would have
it's own time slot for it's pulse...it isn't practical, but theoretically
possible). You'd also have to shorten the pulses appropriately.
Click to expand...


Yes and TMD would required a 1GHz clock, so your using a 1GHz clock
and billion 1Hz clocks to produce a 1GGz clock.
Obviously the billion 1 hz clocks are surplus to requirements.
 
half_pint said:
I think the trouble is you need a 1 ghz clock to syncronise the waves.
So yes, you can make a 1 ghz clock, *if* you already have one.
Click to expand...

Huh? I can make one if I already have one?
 
half_pint said:
Yes and TMD would required a 1GHz clock, so your using a 1GHz clock
and billion 1Hz clocks to produce a 1GGz clock.
Obviously the billion 1 hz clocks are surplus to requirements.
Click to expand...

You ***just*** <VBG> need the 1GHz clock to initially get the phase offsets
right. After that, you've got the billion 1Hz clocks running asynchronously
and everything will be fine (because we all know how stable crystal
oscillators are). C'mon now, we gotta help this guy clear out the 1 billion
oscillators in his closet!
 
Curious said:
Huh? I can make one if I already have one?
Click to expand...

half-inch is a troll and posts shite. I thought you should know this as you
don't seem to have realised that he doesn't in fact know anything but he
sure does like to 'talk'.
 
~misfit~ said:
half-inch is a troll and posts shite. I thought you should know this as you
don't seem to have realised that he doesn't in fact know anything but he
sure does like to 'talk'.
Click to expand...

You are talking bollocks as usual, nothing to say apart from unfounded
personal
attacks, why don't you get a life?

 
George said:
You'd basically be time division multiplexing them
Click to expand...

http://iroi.seu.edu.cn/books/ee_dic/whatis/tdm.htm

TDM doesn't seem to fit the definition of "adding n 1hz clocks in
parallel to increase the clock rate to n hz". It seems to relate to
bits more than hz.
(i.e., each would have
it's own time slot for it's pulse

...it isn't practical
Click to expand...

Why not? Wouldn't each cycle get its own processing? If 1 clock has a
rate of 1hz then wouldn't placing a billion 1 Hz clocks in parallel
[each sample getting its own sampling] to gain a 1 GHz rate be safer
than 1 clock running at 1 GHz? The advantage I see is less heat
generated by the CPU. A CPU with one 1 GHz clock clock will certainly
get hotter than a CPU with a billion 1 Hz clocks. Higher-frequency
generates more heat given the same voltage. A billion 1 GHz clocks
each doing the work in parallel generates much less heat with same
effiency as one 1 GHz clock doing all the work in serial.
, but theoretically
possible).
Click to expand...
You'd also have to shorten the pulses appropriately.
Click to expand...

How would this affect the given signal? Audio for example?
 
Why not? Wouldn't each cycle get its own processing? If 1 clock has a
rate of 1hz then wouldn't placing a billion 1 Hz clocks in parallel
[each sample getting its own sampling] to gain a 1 GHz rate be safer
than 1 clock running at 1 GHz? The advantage I see is less heat
generated by the CPU. A CPU with one 1 GHz clock clock will certainly
get hotter than a CPU with a billion 1 Hz clocks. Higher-frequency
generates more heat given the same voltage. A billion 1 GHz clocks
each doing the work in parallel generates much less heat with same
effiency as one 1 GHz clock doing all the work in serial.
Click to expand...

I just saw this thread and am jumping in with a comment... Please ignore if
it doesn't seem relevant.


It's easier to clock a single bit serial stream than a parallel stream. This
is one of the reasons that SATA can go faster than PATA.

With the parallel stream, you have to ensure that you get all your data bits
to a usable state and hold that state long enough for the clock to fire.
 
Noozer said:
It's easier to clock a single bit serial stream than a parallel stream. This
is one of the reasons that SATA can go faster than PATA.
Click to expand...

Parallel *bit stream* is possible. Is parallel *clocking* possible?
 
Curious said:
"Noozer" <[email protected]> wrote in message

Parallel *bit stream* is possible. Is parallel *clocking* possible?
Click to expand...

Every bit you add makes it that much harder to sync up the clocking...

Consider even the tiniest error in a 100Mhz circuit. That error gets
multiplied 100,000,000 times each second. That would really skew the data.
I'm impressed that PATA drives can move data so quickly.


P.s.
The wife says if you have sync problems, try Draino. : )
 
Noozer said:
Every bit you add makes it that much harder to sync up the clocking...
Click to expand...

True. But I was talking about *cycles* and not *bits*. Ignore the # of
bits per sample.

Take 1 GHz for example. A serial 1GHz clock system would have one
clock "revving" at a rate of a billion cycles per second. A parallel
1GHz clock system would have 1 billion clocks each "revving" at a rate
of 1 cycle per second -- this would somehow result in an overall clock
rate of 1GHz with much less heat.
 
Noozer said:
Every bit you add makes it that much harder to sync up the clocking...
Click to expand...

True. But I was talking about *cycles* and not *bits*. Ignore the # of
bits per sample.

Take 1 GHz for example. A serial 1GHz clock system would have one
clock "revving" at a rate of a billion cycles per second. A parallel
1GHz clock system would have 1 billion clocks each "revving" at a rate
of 1 cycle per second -- this would somehow result in an overall clock
rate of 1GHz with much less heat.

This is a 1Hz per channel. Each channel should have its own 1 Hz
clock. Each cycle is given its own cycle.
 
Back
Top