credible citations required. As in, direct links to pages in motherboard manuals that do NOT specify ECC memory modules.

uh, the fucking slashdot thread from yesterday?

The fix here would be to issue a firmware update that disables that slow 500MB*, and update the labelling on the card to reflect the fact that it's a 3.5GB card not a 4GB.

*Ever run RAM of two different speeds in a desktop? Wonder where those crashes are coming from? It's not a case of the faster RAM waiting for the slower RAM, the slower stuff is tripping over trying to keep up with the faster stuff. It doesn't work in the same way as a PATA channel where the bus runs at the speed of the slowest device.

no, I did not say that. The claim is that these cards, in the first instance, are being sold as 4GB cards. That's as may be, but the top 500MB is deliberately crippled (to turn a 980 into a 970? I don't get the logic) to the point where it can and does cause repeated and repeatable crashes when the 3.5GB ceiling is met under certain conditions, which is ENTIRELY doable when you have a multiple screen setup. That is the issue.

depends, what's your data worth to you sitting on a drive that's not spinning while you wait for your shiny new WD Red?

(I use "obsolete" commodity components in building my NAS gear. My current one uses a 2-port SATA riser on a Via Eden board, mounted in a Shuttle XPC case (equipped with a 100W PSU) and running LAMP docuwiki headless. Total hardware worth: £12 for the SATA riser, total cost: + about £200 for the mainboard and case back in 2006. Just because it's obsolete doesn't mean I should simply bin it).

What about those users (more than one, anecdotes are data not anomalies!) whose use causes the GPUs to attempt to address more than 3.5GB VRAM causing them to crash out? If what NVidia are claiming here according to TFS is accurate, then this should not be happening. It is happening, the 3.5GB roof is being hit hard and people are feeling it. What say you, NVidia?

heat? We're talking about a technology reuse in places such as rural India and middle Africa where daytime temperatures often exceed 45C, I don't think they're going to be overly concerned about something they're a: not pushing that hard anyway - use cases for these things are going to be about as mundane as GDOs and routers - and b: costs them next to nothing to obtain and deploy. If your use case requires investment in fluid pumped cooling I would say that your use case also calls for something with a bit more pep than an ARM SOC.

I get the difference bwtween a 40W brick for a laptop and half a rack of isolators and switchgear delivering 17kW for a five Petabyte cluster, I've dealt with both. I don't imagine for one minute that TFA is talking about competing with Big Iron either in terms of power efficiency or in terms of raw computing power. It's talking about using existing hardware that would otherwise find its way into landfill simply because Johnny Facebook has no further use for it after buying his iPhone 20z, for processes where that hardware can be proven in its environment rather than $Third_World_Telecoms_Startup having to deal with seven-digit sums in maintenance contracts and leasing and IBM.

if compute efficiency were really an issue, we'd all be using RasPis and running RISC OS. As it is, we are all, each and every one of us, using what is available to us "until something we can afford which is in our own mind better comes along". Right now, personally speaking, the most efficient thing for me to do is keep my AMD APU until it burns out and THEN worrying about specifying my next hardware purchase. I'm not about to go buy the latest greatest >1.0-efficient process platform just because it's there because I simply CANNOT AFFORD IT. I'll use what is available to me NOW.

unless you're talking about a shop in middle Africa or even Outback, China that proposes to utilise such a system in a mesh network to bring remote communities one step closer to being Facebook zombies.

Hell, for that matter - how much processing power do you need to run a DHCP router?
Or a DVR?
Or a home automation system? Something as simple as an automatic garage door opener?
An RFID reader?

There's a BUNCH of uses for low power/small iron that Big Iron would be utterly WASTED on. The aforementioned is not, by any means, exhaustive.

citations needed.

yeah but the half life of 40K is something like 1.27 billion years. That's not enough of a curve to estimate the age of a nonfossilised biological sample. The half life of 14C is 5280 years, which is good up to around 60,000 years. We know to a fair degree of precision how much 14C is in the environment and the rate at which it is produced and absorbed, and we also know precisely how it decays and into what: alpha decay to 14N at the rate of 14 events per gram of pure carbon per second (Libby). The mechanism of production is so well known in fact that an accurate estimation can be made of the date of birth of any living individual simply by looking at the relative amount of 14C in their teeth*.

(*according to Choppin Liljenzin and Rydberg, there is more 14C in the biosphere due to the decay products resulting from nuclear testing (ie from thorium and radium fallout following atmospheric detonations) than from any other source - including n + 14N -> 14C + p in the upper atmosphere which is the main natural source).

actually, thanks to NCLB they have been afforded the same opportunities in education as everybody else. The fault of their being dumbasses lies solely and entirely on them. Ergo, the funpokery is entirely justified, in my opinion.

what should terrify you is that these fuckers breed.

banana bread. Contains bananas which are yellow when very ripe (for some definition of "yellow", they're actually about as banana-like as gourds - the modern yellow bent banana didn't actually exist before 1840, it's an entirely artificial cultivar). Bananas of course, being high in potassium which is slightly radioactive.