Your fuel filter is so piss poor that it won't trap iron filings?

The DE3815TYKHE is also substantially bigger than the NUC. Roughly twice the size. Increased size confers better passive cooling.

The i3, i5, and i7 all require fans. Anything that size that takes more than a watt or two requires a fan.

I share those kind of concerns in general. The AOpen MP945 was an example of using an excellently engineered cooling system. There basically was nothing else in the box besides the CPU that made any appreciable heat. Mine was very quiet and never degraded. The NUC from what I've heard has similarly great thermal engineering. But when the cooling system on anything like this degrades or fails, you're going to have to try to find and pay for the expensive custom part. You can't just slap a new commodity fan in there.

Interestingly, my AOpens have held up better than anything else I've had. An endless train of motherboards have succumbed to capacitor failure, but none of my AOpens.

Without a dBa @ distance measurement, with comparisons to other equipment using the same measuring equipment, "quite loud" is not really a useful characterization. Even then the dBa level alone doesn't tell you all you need to know about the acousic objectionability factor. My good old AOpen MP945 with GMA950 graphics (exactly the same size as the good, original Mac Mini) idles and even does useful light work in silence in a quiet residential room with nobody else in the house to make any noise, and without any radio or TV or air conditioner running. Even all out, it is plenty tame, on the good side of laptop noise. The cooling system in a dynamite design.

But my AOpen GP7A with the power hogging NVidia graphics, even sitting idle, periodically roars like a bastard as some random daemon makes a quick tiny demand. When it is really working it sounds like a freight train or jet plane taking off, and oven-like air is rushing out of it.

The NUC takes even less power than the MP945. Certainly the 5i3 is damn quiet. I expect the 5i7 isn't all that noisy. I'm pretty certain it is a damn sight quieter than that GP7A.

And a million times the power drain and heat.

I don't disagree in the main, except that the NUC cooling system is not "whiney".

Why do I think if we actually had the exact specs of your Dell it would be severely inferior to the NUC in various ways.

It's not as high as that. Here is one with all the pieces (including 16GB and M.2 SSD) all assembled and tested for $755. Same thing with the 5i5 is $630; 5i3 is even cheaper.

A Mac Mini with an i7, 16 GB and the cheapest available SSD is $1400. I just went to the Apple store to check. And the Mac Mini is 19.7x19.7 cm. The NUC is 11.5x11.1 cm. A whole different class. Even the original Mac Mini before it got pointlessly squashed down vertically and bloated horizontally was 15x15 - 17x17 cm. If I could find my old shell I would tell you, but it was definitely in that range. The present Mac Mini doesn't even use an external power brick you can toss on the floor under the desk. The main case is bloated to hold the whole power supply.

Maybe you could tell us just what is out there that IS competetive with the NUC?

Mini-ITX is absolutely colossal compared to the NUC. Even the Mac Mini is gigantic in comparison. On the other end, ARM is not even in the ballpark in performance. ARM definitely has its place, but it is not in the same class as the NUC.

By the time you buy your mini-ITX motherboard, case, and power supply you are paying more than an equivalent NUC. The 3i7 is cheap. And, unlike mini-ITX jammed-together nightmares, the NUCs are beautifully engineered systems that go together neatly.

I have never seen a mini-ITX that had anything close to an acceptable cooling system. They were noisy and/or inadequate. I went through a phase where I built a number of mini-ITX systems, and none of them were ever anywhere near satisfactory.

The way the country goes, the powers that be would probably have Obama impeached if he tried to pardon Snowden.

Ha! There's an easy end-run around that scenario. He does what all presidents do. He issues a whole shitload of pardons during the last few days of his presidency. You can't even begin to get the slow wheels of impeachment working in hours, even if anyone is paying attention on his final day.

If a single commenter mentioned this, I didn't see it. The entity employing H1B workers is required by law to file a Labor Condition Application to ensure that they meet or exceed the prevailing wage, and an attestation designed to ensure that they are not used to break a strike nor to replace citizen workers - i.e., that the H1Bs are really needed because citizens cannot be found to do the jobs.

Obviously this does not work, or there would be little to no motivation to gratuitously replace citizens with H1B workers. What no one has satisfactorily explained to me (beyond waving the hands and mumbling "corruption") is, how is the law being flouted?

Yes, conceptually all flash data has to decay when powered off. But implementation tradeoffs vary widely. A dirt-cheap Microchip PIC18F2580 microcontroller has flash data retention without refresh "conservatively estimated" at 40 YEARS MINIMUM and 100 years typical (page 3, 10, 435). The number of previous erase/program cycles that retention is predicated on is not given, but is probably a single one, or a few, out of a typical endurance of 100,000 cycles and a minimum of 10,000 from -40 to +85 C. AFAIK there is no wear leveling or block sparing in microcontroller flash memories.

I have NEVER HEARD of an embedded guy ever running into a case of either cycle exhaustion or data decay in the program memory such microcontrollers (if data is written to flash during operation, specs do have to be considered).

SSDs have flash design tradeoffs remarkably different from this. For example, MLC individual cell endurance is around 1000-10,000, and retention is far less as seen in the article and comments here. In return, the access time is vastly faster and the density vastly higher.

Intel comes with botnet: Intel AMT/Vpro/VT.

VNC server built in to the the chipset, pulls from frame buffer of intel gfx card. Also can upload the contents of your ram over the network or 3g if you have it. Cannot be permanently disabled (can always be remotely re-enabled).

This is to find dissenters and pedophiles.

In the middle east they get bombed. In the west it's prison.

This is to keep men at the bottom and women managing them.

It's just sad to think our uninformed (and probably uninformable) coward probably believes the drivel he spouts. From end to end. And he's got a lot of company.

Credit card companies don't bear the losses in the sense you seem to think they do. The losses are simply treated as a cost of doing business, and passed on to their customers as a group - the credit card holders. To this is added the irrecoverable portion of the costs incurred by irresponsible credit card holders who don't/can't repay their debts, and the costs incurred by RESPONSIBLE credit card holders who fold due to catastrophic illness or injury, getting victimized by a bad economy, thrown into prison for selling someone the wrong kind of cigarettes, etc.

Why do you think credit card interest rates are so usurious in a time of supposedly spectacular low cost of money?