You clearly don't travel much if you think a 1hr clock shift in any way resembles jet lag. The common joke when we travel between Sydney and Perth is the problems associated with the "jet lag" from the 2hr time change.
GPUs have small memory footprints. SKA will be processing HUGE images and data sets. And the image creation cannot be broken up into discrete independent chunks. So the I/O between GPUs is a real problem. Obviously CPUs have the same problem as the on chip memory is (relatively) tiny, but they are designed to pull on the much larger system memory which should be adequate.
Image analysis may well be a different kettle of fish.
Too much television will "rot" your brain - there is no question about this. Studies have clearly show that, for example, children under 2 there is a steep correlation between hours of tv watched per week and vocabulary (tv watching decreases your small kids vocabulary).
My 7yo daughter is reading years ahead of her age - I'm sure this is down to a very small amount of television watching - and she does not miss tv at all. She would much rather play with her sister or read a book.
This is all true, but remember the "holes" between the mirrors don't gain you much. You mostly care about the collecting area. In this case the equivalent collecting area is 22.5m^2
Obviously you have never done basic mathematics.... 8.5 x sqrt(7) is 22 (close enough). Why sqrt? because a telescope is (essentially) 2 dimensional - we care about the collecting area not the diameter (directly). A 20m telescope has 4 times the collecting area of a 10m telescope. This is really, really simple stuff.
This is hardly the largest telescope being built. At seven 8.5m mirrors, it is equivalent of a 22m telescope. Just last week I was using the Parkes 64m telescope which was build 50 years ago (and it is hardly the biggest telescope in the world).
Hell, why do we even allow calculators to be used in ANY exam? What's the point in "teaching" math if you let the calculator do 90% of the work?.
You obviously have never done any sort of serious math, or are still in high school. Math is not about adding up numbers - thats just some of the raw ingredients needed at the start.
NB: In Australia only very basic casio style scientific calculators ($20 jobs) are allowed in exams. There is a list and your calculator must be on the list.
I find this hard to believe. In Sydney (Australia) the Google stuff gets if wrong ALL the time. U-turns on major highways, not knowing about no left or no right turns and missing some connecting roads. So far (knock on wood), my TomTom has NEVER instructed me to do something impossible. Now comparing which gives the best route is almost impossible in Sydney. The traffic is so unpredictable the only sure way to device which was the best route would be to send out two cars simultaneously, with drivers of similar "style".
I needed to uses some digital oscilloscopes for this first time for about 10 years a couple of years back - they were certainly a big set up from the old analog CROs we use in prac at Uni. Features I really liked are:
- Multiple inputs (at least 2) and ability to do sum/difference etc on the two signals. Vital it looking at differential signals
- Ability to save output as a bitmap and some way of accessing the dump - the ones I was using used floppies but I guess that is rapidly becoming unusable
- Remote (http) access. Not useful if these are for playing with directly but great if you want to setup some experiment and monitoring it in comfort.
- Persistent mode - Basically the screen does not blank bits. This is great for looking at long term stability of a signal. You can monitor a clock signal, say, and check it has no glitches over a few hours.
Maybe all digital oscilloscopes do this - I just wander down to the digital lab and pinch the closest one off the shelf...
I was in Amsterdam for five days in May and found your people to be really open, charming, and tolerant. I actually really liked the culture; everyone just goes about their lives doing whatever they damn well please and if someone's being an asshole people tell them to knock it off. American culture, on the other hand, can be VERY judgmental and VERY conformist, so I can see how such openness would cause some of us to become very, very paranoid:) The Dutch are not tolerant particularly. It however is important to been seen to be tolerant (ie the tolerance is a fairly surface thing). And they are definitely judgmental! Almost all expats who have lived there for more than 5 days agree.
Actually the statement about Nyquist's theorem is poppycock. This a mathematical fact, not some weird subjective result open to interpretation. Saying that Nyquist's theorem is wrong is equivalent to stating that the value of pi is really 6.
To get really technical there is some truth in the statement. Nyquist theorem is about sampling. It is says nothing about digitization. If you sample an arbitrary (band limited) waveform at twice the rate of the highest frequency present you can reproduce the original waveform exactly. But you need infinite precision (ie "analog" sampling). As soon as you add quantization into the mix (ie digitization) you cannot get back the original wave. By increasing the number of bits or oversampling (sampling faster that Nyquist) you get closer to the original waveform.
But this is a really minor point - the rest of what this guy was trying to say is stupid
CSIRO is required to be partially self funded via external revenue. How is it expected to develop new technologies if it cannot then raise revenue on exciting "stuff" it has developed? Given CSIRO is not in the business of mass production of widgets, selling its IP (ie patents) is really the only approach it has.
CSIRO owns patents on pretty much all common wireless technology deployed today. It owns these patents because it developed the technologies. Currently CSIRO is actively trying to get other companies to actually honor its legitimate patents not just for "n" but the existing 802.11x standards as well.
These are real patents that lay the mathematical/technological foundation for how pretty much all wireless networking works currently. Saying that CSIRO has no right to defend these patents would be like saying CISCO is morally obliged to give away all their routers because we all depend on the internet and we need routers for the internet to work.
Slashdot Mirror
You clearly don't travel much if you think a 1hr clock shift in any way resembles jet lag. The common joke when we travel between Sydney and Perth is the problems associated with the "jet lag" from the 2hr time change.
It is definitely GBytes, not Gbits. I have benchmarked CPU Memory->GPU transfers of > 100 Gbps with PCI-E 3.0
cat /proc/cpuinfo | grep name | sort -u
GPUs have small memory footprints. SKA will be processing HUGE images and data sets. And the image creation cannot be broken up into discrete independent chunks. So the I/O between GPUs is a real problem. Obviously CPUs have the same problem as the on chip memory is (relatively) tiny, but they are designed to pull on the much larger system memory which should be adequate.
Image analysis may well be a different kettle of fish.
More than just this. A bus is approx 160x the mass of a person. A rain drop is approx 30x the mass of a mosquito. (source google!)
Obviously depends on the size of the bus and person.....
Too much television will "rot" your brain - there is no question about this. Studies have clearly show that, for example, children under 2 there is a steep correlation between hours of tv watched per week and vocabulary (tv watching decreases your small kids vocabulary).
My 7yo daughter is reading years ahead of her age - I'm sure this is down to a very small amount of television watching - and she does not miss tv at all. She would much rather play with her sister or read a book.
This is all true, but remember the "holes" between the mirrors don't gain you much. You mostly care about the collecting area. In this case the equivalent collecting area is 22.5m^2
Obviously you have never done basic mathematics.... 8.5 x sqrt(7) is 22 (close enough). Why sqrt? because a telescope is (essentially) 2 dimensional - we care about the collecting area not the diameter (directly). A 20m telescope has 4 times the collecting area of a 10m telescope. This is really, really simple stuff.
So who is the idiot?
This is hardly the largest telescope being built. At seven 8.5m mirrors, it is equivalent of a 22m telescope. Just last week I was using the Parkes 64m telescope which was build 50 years ago (and it is hardly the biggest telescope in the world).
Oh did they mean largest optical telescope....
Actually it *is* a rumor. There is no official statement. Who knows where the "leak" came from.
Hell, why do we even allow calculators to be used in ANY exam? What's the point in "teaching" math if you let the calculator do 90% of the work? .
You obviously have never done any sort of serious math, or are still in high school. Math is not about adding up numbers - thats just some of the raw ingredients needed at the start.
NB: In Australia only very basic casio style scientific calculators ($20 jobs) are allowed in exams. There is a list and your calculator must be on the list.
Can in run Linux???
Driving to skiing last week, I found this feature *really* useful.
I find this hard to believe. In Sydney (Australia) the Google stuff gets if wrong ALL the time. U-turns on major highways, not knowing about no left or no right turns and missing some connecting roads. So far (knock on wood), my TomTom has NEVER instructed me to do something impossible. Now comparing which gives the best route is almost impossible in Sydney. The traffic is so unpredictable the only sure way to device which was the best route would be to send out two cars simultaneously, with drivers of similar "style".
Maybe you should patent the idea of suing companies for filing invalid patents....
But they only have a 200 kHz input bandwidth!
I needed to uses some digital oscilloscopes for this first time for about 10 years a couple of years back - they were certainly a big set up from the old analog CROs we use in prac at Uni. Features I really liked are:
- Multiple inputs (at least 2) and ability to do sum/difference etc on the two signals. Vital it looking at differential signals
- Ability to save output as a bitmap and some way of accessing the dump - the ones I was using used floppies but I guess that is rapidly becoming unusable
- Remote (http) access. Not useful if these are for playing with directly but great if you want to setup some experiment and monitoring it in comfort.
- Persistent mode - Basically the screen does not blank bits. This is great for looking at long term stability of a signal. You can monitor a clock signal, say, and check it has no glitches over a few hours.
Maybe all digital oscilloscopes do this - I just wander down to the digital lab and pinch the closest one off the shelf...
Yes!
Those results are ancient! Its for kernel 2.4.26. So for 99% of us totally meaningless. 2.6 improved i/o performance significantly.
This should lead itself to a whole new form of hacking - buy the 10 core system and tweak it to use all 100
LOFAR is much smaller than SKA - in size bandwidth sensitivity etc. LOFAR is basically a technology demonstrator for SKA
Actually the statement about Nyquist's theorem is poppycock. This a mathematical fact, not some weird subjective result open to interpretation. Saying that Nyquist's theorem is wrong is equivalent to stating that the value of pi is really 6.
To get really technical there is some truth in the statement. Nyquist theorem is about sampling. It is says nothing about digitization. If you sample an arbitrary (band limited) waveform at twice the rate of the highest frequency present you can reproduce the original waveform exactly. But you need infinite precision (ie "analog" sampling). As soon as you add quantization into the mix (ie digitization) you cannot get back the original wave. By increasing the number of bits or oversampling (sampling faster that Nyquist) you get closer to the original waveform.
But this is a really minor point - the rest of what this guy was trying to say is stupid
CSIRO is required to be partially self funded via external revenue. How is it expected to develop new technologies if it cannot then raise revenue on exciting "stuff" it has developed? Given CSIRO is not in the business of mass production of widgets, selling its IP (ie patents) is really the only approach it has.
CSIRO owns patents on pretty much all common wireless technology deployed today. It owns these patents because it developed the technologies. Currently CSIRO is actively trying to get other companies to actually honor its legitimate patents not just for "n" but the existing 802.11x standards as well.
These are real patents that lay the mathematical/technological foundation for how pretty much all wireless networking works currently. Saying that CSIRO has no right to defend these patents would be like saying CISCO is morally obliged to give away all their routers because we all depend on the internet and we need routers for the internet to work.