Memory transistors are about a thousand times larger than CPU transistors. Do try to keep up.

I've a very good idea that RAM prices are artificially inflated, that the fab plants are poorly managed, that the overheads are unnecessarily high because of laziness and the mentality in the regions producing RAM.

I'm absolutely certain that 15nm-scale RAM on sticks the same size as sticks used today would cost not one penny more but would have a capacity greater than I've outlined.

It could be done tomorrow. The tools all exist since the scale is already used. The silicon wafers are good enough, if they can manage chips 4x and 9x the size of a current memory chip with next to zero discards, then creating the far smaller dies (so you can discard more chips and still get the same absolute yield) is not an issue. It would reduce idle time for fabs, as fabs are currently run semi-idled to avoid the feast/famine cycle of prior years but 15nm would let them produce other chips in high demand, soaking up all the extra capacity.

What you end up with is less waste, therefore lower overheads, therefore higher profit. The chip companies like profit. They're not going to pass on discounts, you getting a thousand times the RAM for the same price is discount enough!

Not really. RAM is only expensive because of the transistor size used. Fab plants are expensive. Packaging is expensive. Shipping is expensive. Silicon is expensive. If you add all that up, you end up with expensive products.

Because fab plants are running very large transistor sizes, you get low yields and high overheads.

Let's see what happens when you cut the transistor size by three orders of magnitude...

For the same size of packaging, you get three orders of magnitude more RAM. So, per megabyte, packaging drops in cost also by three orders of magnitude.

Now, that means your average block of RAM is now around 8 Tb, which is not a perfect fit but it's good enough. The same amount of silicon is used, so there's no extra cost there. The shipping cost doesn't change. As mentioned, the packaging doesn't change. So all your major costs don't change at all.

Yield? The yield for microprocessors is just fine and they're on about the scale discussed here. In fact, you get better. A processor has to work completely. A memory chip also has to work completely, but it's much smaller. If the three round it fail testing, it doesn't affect that one. So you end up with around a quarter of the rejection rate per unit area of silicon to a full microprocessor.

So you've got great yield, same overheads, but... yes... you can use the fab plant to produce ASICs and microprocessors when demand for memory is low, so you've not got idle plant. Ever.

The cost of this memory is therefore exactly the same as the cost of a stick of conventional RAM of 1/1000th the capacity.

Size - Exactly the same as the stick of RAM.

Power budget - of no consequence. When the machine is running, you're drawing from mains power. When the machine is not running, you are refreshing the dirty bits of memory only, nothing else. And 99.9% of the time, there won't be any because sensible OS' like Linux sync before a shutdown. The 0.1% of the time, the time when your server has been hit by a power cut, the hard drive is spun down to save UPS and the main box is in the lowest possible energy mode, that's when this sort of system matters. Even on low energy mode, buffers will need flushing, housekeeping will need to be done, transactions will need to be completed. This system would give you all that.

And the time when the machine is fully powered, fully up? Your hard drive spends most of its time still spun down. Not for power, although it'll chew through a fair bit - mechanical devices always do and the high-speed drives being proposed will chew through far, far more. They'll be spun down because a running hard drive suffers rapid deterioration. Can you believe hard drives only last 5 years??! Keep the damn thing switched off until last minute, then do continuous write. Minimizes read head movement (there's practically none), minimizes bearing wear-and-tear, eliminates read head misalignment (a lot of times, you can write the entire disk in one go, so what the hell do you care if the tracks are not perfectly in line with the ones they're replacing?) and (by minimizing read head time over the drive) minimizes the risk of a head crash.

I reckon this strategy should double the expected lifetime of drives, so take the cost of one 10 Tb drive and calculate how much power you'd need to consume extra for the memory in order for the memory's power budget to exceed the value of what you're doing.

Oh, and another thing. Because I'm talking memory sticks, you only need to buy one, subsequent drives of the same or lower capacity would not need to have memory there. You could simply migrate it. RAM seems to hold up ok on old computers, so you can probably say that the stick is good for the original drive and the replacement. That halves the cost of the memory per drive.

So, no, I don't see anything unduly optimistic. I think your view of what the companies could be doing is unduly pessimistic and more in line with what the chip companies tell you that you should think than what the chip companies can actually do.

Agreed, which is why it should be there.

Nonetheless, there needs to be a backup plan in case it does turn out that the NSA or GCHQ have a backdoor to it. If it's been deliberately compromised (and I'm not keen on changes made AFTER it had been approved as SHA3 for that very reason), then the more paranoid amongst us need to have a backup plan. I certainly wouldn't suggest HTTPS over TOR use algorithms that are considered three-letter-agency-unsafe for any part of the security protocol, for example, since they're the ones doing most of the attacking.

There's no easy answer to this, but I think that having SHA3 and NESSIE as the two standard choices and limited support for some third algorithm for when approval simply isn't good enough is the only real solution. The first two can be standard on all browsers and by all certificate authorities, the third only needs support on special-purpose browsers and OpenCA/OpenSSL/LibreSSL (since most uber-secure sites will roll their own certs).

Whoever was in charge of the live stream are a bunch of amateurs, incompetent idiots and should be fired, publicity shamed and never hired again.

According to this article there were a number of issues which were all caused by Apple rather than Akamai:

The bottom line with this event is that the encoding, translation, JavaScript code, the video player, the call to S3 single storage location and the millisecond refreshes all didn't work properly together and was the root cause of Apple's failed attempt to make the live stream work without any problems. So while it would be easy to say it was a CDN capacity issue, which was my initial thought considering how many events are taking place today and this week, it does not appear that a lack of capacity played any part in the event not working properly. Apple simply didn't provision and plan for the event properly.

I don't know enough about streaming to comment on the validity of the assertions made.

High capacity I can understand, but high speed is senseless. At current transistor sizes, you could easily have 10Tb of battery-backed RAM on a hard drive. You can then peel the data off the hard drive into RAM and write changes when there are enough or when a sync command is sent. RAM doesn't eat battery significantly, it only needs to maintain state and then only on dirty portions. That'll easily buy enough time to survive power outages and Windows crashes.

If everything is in RAM, access times are insignificant for always-on machines (the ones likely to need 10Tb of disk space). Since writes can be postponed until critical, the disk can spend most of the time totally powered down.

Now, if you're REALLY clever, you have twice that RAM. One lot for working space (which doesn't need battery backing) and one lot for writing to disk. This second set can be permanently defragmented, with writes designed to be compact on space and the hard drive spun to specifically provide for that.

Microsoft will probably implement SHA0. There's no value in SHA2 (and variants) now that SHA3 has been ratified, since SHA2 is just SHA1 with some lengthening. If SHA1 is brutally compromised, SHA2 will fall shortly after. Best to switch to NESSIE (Whirlpool) and SHA3 (something that sounds vulgar).

Having said that, SHA3 involved dubious mid-contest rule changes and spurrious rejection criteria that might well have been NSA-inspired. I'd take a very close look at the Hashing Lounge for any second or third round reject that shows greater resilience across the board (pre-image vulnerabilities, etc) as a backup in case NESSIE and SHA3 are seriously compromised.

That's why you should never hire people, just small furry creatures from Alpha Centauri.

I deny all knowledge about the epson fx spontaneously catching fire.

The short circuit that blew up two power transformers and an embedded computer had nothing to do with me. And you didn't see me. And I was in disguise anyway.

Nobody saw me insert the radio direction finder valves into the R1155, switch it on and jam all televisions in the neighbourhood.

So, no, I've no knowledge of using technology to get into trouble. None whatsoever.

I had forgotten about the gridded fins...you're right, that should provide substantial cross-range capability.

A big challenge for water landing will be wind during the descent of the rocket. If the wind is blowing 100 miles an hour for a minute as the rocket is falling, then it's going to be dragged a mile from the ballistic landing point. (When things move quickly through the air, the lift generated by wind is extremely high; bullets move with the wind.) I don't believe that the booster will have the capacity to fly horizontally too far, and it won't be firing at all for the bulk of the descent.

If the wind could be predicted accurately, it would be easy enough to steer the rocket to the right place -- or move the landing platform to the right place.

If you're landing back at the launch pad; there will have been a rocket that could have sampled the wind speed just a few minutes previously, so you could have very precise wind speed vs. altitude data.

There was a very nice system, including PIN numbers to manage the POS terminals. Way back when stock trade was 49$, it was 25 cent per transaction irrespective of the size of transaction. This should have become zero. But that is not what happened.

The 25 cent transaction fee is charged by the acquiring bank, not Visa and Mastercard - whose fees for debit are typically 1 cent per transaction as they are a volume based business.

The reason that acquirers charge is because they incur costs associated with that transaction (including, but not limited to, interchange fees). If they didn't charge, it would fail as a viable business model.

Pre-paid cards still have to use Visa or MC to get the request for the money from the acquirer (who has the relationship with the merchant and typically provides the terminal) to the card issuer (the bank that supplied the pre-paid card).

Regarding AppStore vs MC+Visa, in order for Apple to be able to accept payments directly they would have to get an e-money licence so they could issue virtual debit or credit cards for use on their phones. By doing so, they'd still need the rails that Visa and MC provide - unless they really want to get into the business of connecting themselves to all the banks worldwide (aka becoming a payment processor).

It's harder than you think, unfortunately. Nuclear weapons have a few kilograms of radioactive material, reactors have more than a few tons. The Yucca Mountain repository, the best that nuclear engineers could come up with, had to be certified to be safe for 10,000 years...but literally after 10,000 years things could have gotten out of control. It's a tough problem.

That said, it means that we have to try harder. The problem is not going to go away; we have to pursue better approaches.

Those who are dead via saline solution can be told once they're... ummm... undead.

Yeah, I can see you do great on statistics, too.

Death stopped being binary some years back (suggest you read medical news) but this isn't about that. This is simple numbers. If device X kills N times out of 100 and device Y kills M times out of 100, where N != M, the lethality of the devices is not the same.