* Posts by JeffyPoooh

4286 publicly visible posts • joined 28 Jun 2013

Boffins unwrap bargain-basement processor that talks light and current

JeffyPoooh
Pint

Re: I wonder what the latency of this is?

Colin Tree found one example Tx <---> Rx pair 65ns...

It works the other way too. Assuming that the example device that you found isn't physically capable of storing bits beyond just being in one state (On/Off), then the latency is a hint about the maximum speed of the device.

65 ns of latency hints that it's very slow device, operating in the single-digit MHz range. Even 10 MHz is likely to be a horrific phase-shifted sinewave. There are alternate explanations, but they're perhaps less likely and just as low performance.

If a device with 65 ns latency was used (for example) to transmit a 1 Gbps signal, then it needs to hold at least 65 of those bits within itself physically, which may clearly and obviously be impossible.

This is all a very useful technique to detect tech-nonsense instantly.

Happy Boxing Day.

JeffyPoooh
Pint

Re: I wonder what the latency of this is?

In this thought experiment, it's more like 'reductio ad absurdum' memory.

Which is to say, if it were, it would be, but it obviously isn't. Which is the point of course.

Merry Xmas!

JeffyPoooh

Re: I wonder what the latency of this is?

There's a simple thought experiment that can reveal the approximate latency of any such system.

This: "How many bits can it hold?"

Let's say a transducer or subsystem or media is passing 100 Gbits/s (100 bits/ns)..

Somebody is worried that it may impose, for example, 100 ns of latency.

So... How many bits can it hold?

Can the structure store 10,000 individual bits? (100 bits/ns x 100 ns = 10,000 bits)

Have they accidentally invented 'The Fastest Cache Memory....In The World'?

If it can store 10,000 bits and at 100 Gbits/s (!!), wow!! A Nobel Prize awaits.

If the structure can only be in one state (0 or 1) over its entire length (ref Speed of Propagation if required; c x velocity factor), then the latency must be less than the inverse of the data rate. In this example, less than 0.01 ns.

It all becomes perfectly obvious once you consider this thought experiment approach.

Merry Xmas.

(I'm just waiting for the family to awaken on Xmas morning.)

JeffyPoooh
Pint

IC package with FO interface

Next: Fiber optics to interconnect ICs.

Little grooves on the package, aligning the interchip fibers and bringing them close to the on-chip photonics, that are behind a transparent window. A hinged cover then snaps down to hold them in place.

Replace overly complicated PCBs with simpler PCBs and a handful of fibers.

Better, Faster, Cheaper.

Merry Xmas.

Getting metal hunks into orbit used to cost a bomb. Then SpaceX's Falcon 9 landed

JeffyPoooh
Pint

Re: FFS!!!

"This doesn't need to be a competition."

An exquisitely incorrect statement.

JeffyPoooh
Pint

Re: Party Trick

NeilPost "...having to carry....20% more fuel..."

Mush is quoted as it's a 30% ***reduction in payload***. Not 20-30% "more fuel".

Payload to fuel mass ratio is a small fraction.

So extra fuel (and including vanes and legs) is 30% x a small fraction.

Cheers.

JeffyPoooh
Pint

Re: Party Trick

NeilPost "Think of all the extra fuel needed to be carried all the way into space and back to allow this."

Much less fuel than your instincts are suggesting.

Musk has stated that the Return to Land approach (fuel and extra gear) imposes 30% reduction on payload. NOT 30% extra fuel mass, but 30% less payload. Payload mass is a very small fraction of initial fuel mass. So that's 30% multiplied by some very small fraction.

Although the 1st stage looks much the same landing as it did at launch, it's then nearly empty of fuel and is thus vastly less mass.

People are visual creatures, and one can't see that it's essentially empty at landing.

They should make it transparent so that everyone can more easily understand.

Merry Xmas!!

JeffyPoooh
Pint

Re: Real numbers would be interesting (space debris)

I noticed that the shroud panels were jettisoned surprisingly (to me) early in the ascent. It made me wonder if that was intentional, in order to ensure that the panels were low enough and slow enough to ensure that they will definitely reenter quickly.

I've seen plenty of old NASA launches (videos) where the shroud panel were finally jettisoned after reaching the parking orbit. Meaning that those shroud panels will certainly become space junk for months or years.

If this detail is as I suspect, then kudos.

JeffyPoooh
Pint

Re. "...the first footprints on Mars be made..."

Prediction. The first footprints on Mars be made by somebody from the NHK/Discovery Channel joint venture video crew that lands a few days before 'The Big Event' to set up. Actual first person will probably be some previously unknown caterer running outside to set up the snacks table.

It'll be months later before anyone twigs, "Hey, who videoed this 'first step' anyway? Oh. Oh my gawd..."

JeffyPoooh
Pint

Re: Real numbers would be interesting

Credas: "...has to leave fuel unburnt to power it's descent. ...mass taken off the rest of the payload. You end up with a much bigger, and hence more expensive, launcher than a non-reusable one for the same payload. ..."

It's a bit unobvious, until you think about it, that the amount of extra fuel required is quite small when compared to the initial mass of fuel required for launch. Because the first stage mass is mostly fuel, once it's reached initial MECO, then it's much less mass. So the retro-burn and re-entry is dealing with a very low mass vehicle. Surprisingly so.

Also, even if the 1st Stage was twice the cost for the reason you've listed (but it clearly isn't twice), if it can be reused even just three times, then you're ahead on the deal. If it can be reused dozens of time, well - you do the math.

In summary, it's a*obviously* a very clever strategy. Your objections are all true, but are obviously swamped out by the basic math.

JeffyPoooh
Pint

Bezos' naming it "New Shepard"

America's 1st man in space was Alan Shepard. He went on a short *sub-orbital* 15-minute hop (Alan Shepard himself went on to walk on the Moon during Apollo 14). It wasn't until John Glenn's flight (The USA's 3rd manned space launch) that the USA put a man into orbit.

Bezos naming his ship the "New Shepard" seems to be a subtle nod to the sub-orbital capabilities.

.: My conclusion is that Bezos hasn't missed the distinction (w.r.t. suborbital); seeing how he's made a (very) subtle joke about it.

Christmas comes early at US Patent office after massive IT outage

JeffyPoooh
Pint

Good. More time for goofing off, writing scientific papers unrelated to work

It's happened before, in 1905.

Drivers? Where we’re going, we don’t need drivers…

JeffyPoooh
Pint

@Nom^3

You are correct.

I'm going to hold out for a 3D Printed, Self-Driving, fusion-powered (too cheap to meter), Flying Car.

For a start,perhaps someone can explain lane-keeping on snow covered roads. Talk about 'strong AI'. Or a whole new navigation system, supplanting GPS.

Too much child-like naivety on this topic. In the long run, yes. But much further out than anyone is expecting.

JeffyPoooh
Pint

Re: Let's just say that...

I posted an obvious point about driverless cars vice congestion, and AceRimmer replied with some assumptions about "ride sharing" and essentially driverless taxi cabs. Ride Sharing with taxi cabs is as feasible now as it would be with driverless cars. Which is to say, not really (for many reasons); else it would be commonplace (which it isn't).

Plenty of muddled thinking on this topic.

JeffyPoooh
Pint

Re: Let's just say that...

Cynic_999: on replacing private cars with driverless taxis

Replacing parked private cars with empty driverless cars driving back for more passengers points towards a DOUBLING of traffic congestion during rush hours.

Yes, the parking garages will be much less full. But the roads will be twice as jammed. By cars wandering around twice as much.

PS. Wasted fuel too.

If you doubt this common sense prediction, then go to Manhattan NYC and look at the swarming masses of yellow cabs, many empty of passengers, on their way somewhere.

Too much unthinking hype on driverless cars.

JeffyPoooh
Pint

Fair Weather technology

Much of the development is in SoCal. Bring it up to Canada in February, when the roads are covered by snow (no visible lines), and let me know how well it works then.

Rebuttals mentioning ill-defined "GPS" magic as a solution to one foot accurate (!) lane-keeping will be ignored.

The ball's in your court, Bezos: Falcon 9 lands after launching satellites

JeffyPoooh
Pint

"...how they decided on this method, despite its obvious challenges."

The Why and How is obvious in the final design.

Obvious requirements:

1) a little bit extra fuel (not as much as you think since it's nearly empty and thus low mass)

2) those steering vane thingies

3) deployable legs

4) software

Non-intuitive requirements:

5) ITAR grade GPS (due to altitude and speed)

6) Engines that don't mind operating close to a surface

7) Multiple redundant C&C, self-destruct and abort systems (due to aiming back at the Homeland)

All you have to do is think about it, and you'll soon see that any other solution is less optimal.

The only tricky bit is the software, and that's not THAT difficult if you give the coder drones and physics/math geeks sufficient time to think it through.

JeffyPoooh
Pint

Re: Returned to Earth

ZSn offered "With a heavy load of fuel in the tanks..."

At launch, yes.

But at retroburn, much much less.

At landing, hardly any.

JeffyPoooh
Pint

Re: Return to Earth

Chris: "...cargo of 11 satellites (i'm guessing tiny ones)..."

Their *smallest* dimension ('height') is reportedly about 0.5m. 380 lbs each.

So smallish, but certainly not "tiny".

JeffyPoooh
Pint

"...spaceships land like this... ...Sci-i..."

Alister: "Isn't it curious how we've seen spaceships land like this for decades in Sci-Fi films and TV shows, but to see it actually happen for real somehow is just so much more awesome."

Space vehicles landing vertically has been done before.

At least six times.

On the Moon.

Starting in the 1960s.

JeffyPoooh
Pint

Re: That's stage 1 sorted

Peter Ford asked about Stage 2.

It flames through the atmosphere at 17,000 mph and what's left splashes into the ocean in a traditional manner.

They mention that they intentionally deorbit it (retro-burn), which is nice since it was some 630 km up. Otherwise it'd be space junk for nearly forever.

JeffyPoooh
Pint

El Reg. "The rocket launched at 12:28, Cape Canaveral time."

That sentence is not true.

I watched it 'live' via the 'net, and I wasn't up that late.

Canadian live route map highlights vulnerabilities to NSA spying efforts

JeffyPoooh
Pint

The Traveling Salesman Problem

Same thing. Which is interesting. Might be a Clay NP prize in there somewhere.

Perhaps the routing algorithm is distance-unaware and is optimizing based on other, similar parameters. Perhaps they're mindlessly following a trivial 'least number of steps' or 'least cost' algorithm, and any pattern of seeming routing intelligence is just an emergent property. Perhaps all the NSA needs to do is provide subsidized, high bandwidth routes to attract traffic.

Lettuce-nibbling veggies menace Mother Earth

JeffyPoooh
Pint

Daft nonsense...

Given that celery and lettuce have next to no calories, it's nearly a divide-by-zero error trying to calculate their embodied resources per calorie.

Repeat the calculation for a nice glass of water. Zero calories, and thus AN INFINITE WASTE OF OUR PRECIOUS RESOURCES (per calorie). Daft nonsense.

Frankly, the issue here is that some have fallen for this nonsensical, elementary math trick.

Ashley Madison blackmailers try again with snail mail

JeffyPoooh
Pint

"... relatively anonymous digital currency..."

I'm glad that you included the word 'relatively'.

Brit 'naut Tim Peake thunders aloft

JeffyPoooh
Pint

Has the Flat Earth Society ever explained this?

Within about a minute or so, the rocket - which is going up - is clearly pointed downwards on the video screen.

Predictable: How AV flaw hit Microsoft's Windows defences

JeffyPoooh
Pint

Y'all should have stuck with the Harvard architecture...

As soon as you put data and executable code into the same memory space, you've set yourself up - inherently - for computer insecurity. And there's NOTHING you can do about it.. ..until you separate these two spaces in hardware.

Within 5-10 years, hardware will be fast enough that you'll see malware that brings along a virtual environment for the stooge OS to live. Scanning a computer will be like sitting a lying criminal in a chair and asking him if he's innocent or guilty. USELESS.

Next issue is that hardware is being replaced with software. E.g. USB controllers are now ARM processors, and malware can hide in the controller, invisible to "security scanning" the flash. This is only going to get worse and worse. You can NEVER certify a system as secure. Inherently impossible.

HOPELESS. By design. Sensitive systems need to be Harvard architecture with certified code in hardware write-protected or write-once ROM.

Most of this was known by the late 1930s, perfectly obvious from Turing equivalence.

Volkswagen blames emissions cheating on 'chain of errors'

JeffyPoooh
Pint

Author conflating 'NOx scandal' with 'CO2 issue'

"...tried to put the scandal behind them. An accompanying press release was headlined 'CO2 issue largely concluded.' "

The author of this has obviously conflated 'the scandal', which is almost entirely about NOx, with a minor subsidiary 'CO2 issue', which is obviously about CO2.

NOx .NE. CO2.

Those at VW trying to communicate their point must be head-thumping-on-desk-ly frustrated with such confused reporting.

Uber fined $150,000 and forced to embarrass itself by French court

JeffyPoooh
Pint

"...governments can't really stop this kind of thing."

ShadowPenguin6858 wrote: "...governments can't really stop this kind of thing."

Uber is the simplest thing on Earth to stop: Make sure the laws are in place to make it clearly illegal. Government enforcers select another burner phone from the box, ask the next driver to stop by, and request to be driven straight to the Impound Yard. (You get the idea...)

The fines and auctioned-off cars pay for the burner phones and running costs. The only limitation is the size of the Impound Yard, and the speed of the legal process leading to the auction.

Trivial. Few other illegal activities have such an easy method to beckon the wrong-doers.

PETA monkey selfie lawsuit threatens wildlife photography, warns snapper at heart of row

JeffyPoooh
Pint

Re: The justification for copyright on photos can be weak at the boundaries

If I see another damn 'Leaf In Puddle with Out-of-focus Background' image I may vomit.

Software defined? No no no, it's poorly defined storage (and why Primary Data is different)

JeffyPoooh
Pint

Duh...

Even little USB sticks have migrated from ASIC based controllers to ARM processor based controllers. The IC (or embedded) between the Flash Memory and the USB pins. This is related to the news about the malware hidden in the controller, not in the flash and thus invisible.

So yeah, software is everywhere.

'Dear Daddy...' Max Zuckerberg’s Letter back to her Father

JeffyPoooh
Pint

Re: Perhaps, just perhaps...

DH listed "...heating..."

Travel much?

(Hint: The poorest areas of Earth often need Air Conditioning, not heating.)

JeffyPoooh
Pint

How about we get them modern electricity grids and cheap reliable energy first?

That'd be those 2600+ Coal Fired power stations that are planned to be built over the next decade or so, thus ensuring a nightmare 4°C global average temperature rise (if we believe the predictions).

How about we invent something better FIRST?

From Zero to hero: Why mini 'puter Oberon should grab Pi's crown

JeffyPoooh
Pint

Maybe somebody could write an Oberton emulator for the $5 Pi

$140. Not exactly spur-of-the-moment mad-money pricing.

Like the $5 Pi, the $140 Oberton is also not actually available (Sold Out). So they're both hypothetical for the time being.

Android on Windows is disruptive because neither Microsoft nor Google can stop it

JeffyPoooh
Pint

Re: An OS is 'just' SW...

DaLo, "Not exactly a stellar prediction... ...quite some time..."

So why do so many fail to see the obvious long-term future?

One can imagine some future OS accepting Apps written for other OSes. Its embedded 'intelligence' would identify the App's native OS and intent, and then execute it within an emulation. Transparently.

It seem obvious.

So why they downvotes?

JeffyPoooh
Pint

An OS is 'just' SW...

I got heavily down-voted when I suggested that the future includes OSes running other OSes, perhaps all within another OS.

The end of eco-systems on phones. Ghetto walls crashing down.

It's starting.

Who owns space? Looking at the US asteroid-mining act

JeffyPoooh
Pint

Re: Home Delivery

Space elevator.

Put a pulley at the top and loop the cable around an electric motor on the ground. Makes the whole concept much easier to implement.

JeffyPoooh
Pint

"...without regulatory oversight during an eight-year period..."

Eight years?

Might as well be eight minutes.

Eight years is the blink of an eye, unless someone has got 24 billion 1969-dollars to spend.

California cops pull over Google car for driving too SLOWLY

JeffyPoooh

Re: @JeffyPoooh

@AC

Putting it simply, you do NOT sum the speeds because of the speed squared function to energy. You'd be creating energy if you can sum speeds. So don't.

You can sum the masses. Two cars head-on, twice the energy, but distributed between two cars. Indistinguishable to each car hitting a brick wall.

This isn't quite as simple, but has the advantage of being correct.

JeffyPoooh
Pint

Re: I you want a guarantee, get a toaster

So, how do you certify a Machine Learning system to Level 'C'?

Answer: You can't. Not practically.

In reality, the system might not recognize pedestrians wearing kilts. You just don't know.

Testing (vice Analysis) to the required level of safety will take a Very Long Time.

JeffyPoooh
Pint

Re: @JeffyPoooh Silver badge

@Tom 13

You're confused. The original issue was with the phrase "sum of speeds".

With two cars in a head-on collision, you may SUM THE MASS, not sum the speeds.

So with two cars, each brings their own energy. But it's distributed between the TWO cars. Indistinguishable from each hitting a brick wall.

This is high school Physics.

If you know the Physics, then perhaps it's a reading comprehension problem on your part.

JeffyPoooh
Pint

Re: @Charlie 9

@DaLo

Your discussion includes "an immovable Car2" which is a VERY poor choice of an example.

Do you mean that Car 2 is parked but ready to move with impact, or it's fixed in place like a 'brick wall' ? Does it have crumple zones? Is it in Neutral, or Park? How good are the tires on Car 2?

Makes your entire argument unclear.

JeffyPoooh
Pint

Machine Learning meet DO-178...

Some of you might know about the DO-178B/C process. Design Assurance Level 'A' and all that analysis and documentation. Often required for safety-critical software.

But AI often relies on 'machine learning'.

How do you certify a system that's been trained, not explicitly designed?

As far as I know, the tools you'd need to do 'code review' of the resultant training data do not even exist. Maybe they do, but I suspect there's a gap exactly there.

Some may try to claim that testing can support certification of such systems. They'd just need BILLIONS of hours of testing. This might take a while...

'Just around the corner.' Like flying cars. Like fusion power. Like Middle East peace.

JeffyPoooh
Pint

Re: Here in California ...

jake "...you can be cited for this ... "

Citation are issued at the whim of the officer. I doubt that such a ticket would survive review by the prosecution, let alone a judge. A phone call to the prosecution would make it go away.

Many people are confused on this point. Similar to 'You can be sued...'.

JeffyPoooh
Pint

Re: What's funny...

AC: "Strong AI does exist today..."

BS.

AI exists, arguably.

Strong AI is still a dream.

JeffyPoooh
Pint

@Charlie 9

"Now you have two cars crashing at the sum of their respective speeds..."

That's not how Physics works. Summing non-relativistic speeds, which are squared in the energy formula, would lead to creation of energy from nothing.

If two cars (same mass) doing 100 kmh crash head-on, that's basically the same as each hitting a brick wall at the very same, non-summed, 100 kmh.

Using bold print and being wrong must be embarrassing.

D-Wave heads for New Mexico

JeffyPoooh
Pint

Re: "thousand-qubit"

"Quantum annealing superimposes the equally-weighted candidate states, then gradually evolves them in parallel toward the global minimum, with quantum tunneling between states."

Factor 1 * Factor 2 = a given 1000-bit number

Find the two factors. This is a trivial problem, but difficult.

If a quantum computer can't solve it, then what's the point?

Seriously! If it can't find a solution to this, then what's it for?

The D-Wave TreeStump 4000?

JeffyPoooh
Pint

Re: "thousand-qubit"

Quantum annealing is suppose to instantaneously consider all possible solutions at once, and then settle on the solution directly.

Factoring a big number becomes simple multiplication, f1 x f2 = given big integer

Assuming you'd have to use a legacy algorithm is obviously nonsense. Why on Earth would you make that assumption?

If a D-wave can't perform this simple canonical problem on demand, then it smells like BS-ware.

PS, your stackexchange link has a highly rated answer that supports my position.

How TV ads silently ping commands to phones: Sneaky SilverPush code reverse-engineered

JeffyPoooh
Pint

Re: Those frequencies are too high

@Tom "...khz....db....Khz..."

kHz (lowercase k uppercase H lowercase z)

dB (lowercase d uppercase B)

JeffyPoooh
Pint

I doubt that the speakers in my TV are capable of achieving such lofty freqs

Home theater, sure. Don't use that every day.

TV speakers, seems unlikely that'd reach even 15kHz.