Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×

Comment Re:Performance? (Score 1) 83

It depends on if they emulate it by translation and shadowing, or by interpretation. Software translation is rather fast, but not native-fast. To get native-fast, you have to go native.

I've been suggesting an accelerator chip (maybe even off-die) that decodes x86 instructions into the internal RISC instructions stored in the ICache, but people keep telling me it's impossible because... they're stupid. Modern x86, x86-64, and ARM chips all read instructions in their ISA and translate to an internal CPU RISC ISA, to the point that x86-64 chips actually translate x86 instructions to take advantage of around 60 hardware registers thanks to having not just twice as many GPRs (16!), but those GPRs being 64-bits wide (32 GPRs), plus the EBP, ESP, and EIC registers being 64 bits wide and only validly addressing a 32-bit address space (3 more registers). Multiple instructions hitting the same memory won't just work on cache (fast), but will actually load that cache line to register (extremely fast) and operate a series of instructions there--even out-of-order instructions.

No doubt the instruction decoder would be large-ish, and access to its own (I4) cache would prove a performance boon if that cache is kept consistent with actual RAM. Nevertheless, it should be roughly-trivial to produce a CPU chipset that can execute both x86-64 and ARM64 code, in the same way it's roughly-trivial to produce a CPU chipset that executes both x86 and x86-64 or ARM and THUMB. These aren't simple tasks by any means; but the fact is we routinely create chipsets which execute an ISA, and chipsets which execute multiple ISAs, and even chipsets which execute old ISAs while automatically leveraging internal facilities such as registers available to new ISAs (which isn't impressive when you think about OOE, parallel execution, and branch prediction). x86-64 is related to x86, but only at face value; they're different instruction sets, just like ARM and MIPS are different, and it should be fairly easy--not cheap, mind you--to wedge x86-64 in with ARM.

Comment Re: Americans? (Score 1) 320

That's true. That's a continuation of when Ikea adjusted the shape of their boxes to stack three times as much stuff onto one shipping pallet, reducing the number of shipping employees involved by 60%; and when the shipping pallet was invented; and when the automobile and train replaced the horse and buggy; and when overland shipping replaced a large population of sailors, which happened after the old cold blast process that made 400 pounds of iron was replaced with a hot-blast furnace that could make 86,400 pounds of iron with the same amount of labor time.

We've been cutting back at the amount of work to do anything since we sharpened spears. We changed from stone to bronze and then to iron and steel; we created better tools; we got rid of artisans who take forever to make anything and went to assembly lines that use 1/8 as many people to do the same shit, and then to cellular manufacture that achieves the same in 1/12 as much labor. Better logistics, better management, new tools, power tools, and the like have continued to cut this back. Fertilizer and GMOs have eliminated nearly 100% of the labor used to make food over the past 200 years.

It's a beautiful forward march of progress, isn't it? The standard-of-living of all classes goes up as the labor required to make any one thing goes down; and the laborers working their full time to make those things--fewer they are--still have the same money representing the same labor-hours, even though to buy anything they don't need to induce the labor of so many labor-hours. They can then buy more things, and so do so, which is why we tend to stabilize around 5% unemployment even as we eliminate nearly all jobs that have ever existed.

Imagine what it'll be like when only 1/4 as many people are required to make Tesla cars. That $85,000 top-tier Model S will be replaced by something ludicrous for the rich folks, and the common man will buy something roughly equivalent to the top-tier Model S performance box for $21k. That'll be the car we drive, the car the working-man owns.

Comment Re:Greed by any other name... (Score 1) 531

My point is that GDP-per-capita says what was produced; median income says how much people have to spend. If you produce $57,000 of stuff per capita and you have $52,000 to spend, guess what? That's more than $38,000 of stuff--the amount of stuff produced per-capita in 2000.

What I'm saying is $52,000 worth of stuff is more than $38,000 worth of stuff. Median income of $52,000 means you can buy $52,000 worth of stuff.

My point is that the median income remaining flat while that amount of income buys more and more stuff means you are getting richer; and we don't have a good indicator that shows what people can actually buy.

Median income in constant dollars is probably the best single number to indicate average income. GDP does not enter into that. Given constant income in constant dollars, it doesn't matter how the GDP changes.

Except that the median income in constant dollars has fallen by like $5,000 in the past 15 years, yet the amount of stuff you can buy with the current median income far-exceeds what you could buy with the median income 15 years ago.

The same percentage of the median income buys a car that's got more-complex technology like complex suspensions, electronic stability control, fuel-injected engines, power locks, bluetooth radio; that's a lot of complex, expensive shit that you could get in a high-dollar luxury vehicle a decade or so ago, if you could afford to spend 2-5 times as much for a car.

Each family spends a smaller percentage of their income on food now; and they eat out about twice as much as they did a decade ago, meaning they buy food and pay servants to cook and wait on them while they eat with less money than they previously paid to cook their own food.

Services like internet have exploded. In 1998, you would pay $35/month for 128K ISDN service, demarcated by a $250 ISDN modem; today you pay $83/month for 200,000K cable internet demarcated by an $80 DOCSIS 3.0 modem. That's 1,562 $35 ISDN lines. What percentage of the median $52,000 income is $54,687?

Smart phones. We have high-speed computers in our pocket that can get e-mail, stream music, play games, and do voice chat. Do you remember paying $600 for a Compaq iPaq with 32MB of RAM that used RAM as storage (yes, if you removed the battery, it wiped the phone!) in 2001? Do you remember it not having a cell phone radio? Instead you got that Motorola V3 RAZR for $350; and today you can throw $350 and get a cell phone with 64GB of storage, 2,048MB of RAM, and a 1080p AMOLED screen--and it's got four friggin' radios so you can do Wifi, Bluetooth, GSM voice, and LTE data simultaneously.

What the hell happened that our real income went down yet we became fucking rich as the Sultan of some backwater oil capital? How does that work? How do you sit on coal, turn it into diamonds, and get poorer?

Comment Re: Americans? (Score 1) 320

Actually, outsourcing and importing for cheaper is part of what gives the poor a better quality-of-life. The unemployment argument is a red herring: bringing any trade jobs back incurs an added cost, which raises prices and thus eliminates other jobs. If those prices raise enough to diminish the newly-created jobs beyond the number of jobs lost, then you have a net-loss of employment; this means American workers producing trade-import goods would need to be paid little to net-create jobs.

Even then, the change is disturbing.

Right now, Men and Boys's Cotton Trousers and Shorts retail for an average of $14.97 per pair (this is a rough Google number, and is probably inaccurate; it's also the only factor that can be variable and still correctly-demonstrate the principle). The Chinese import cost is 6 cents per pair (40,000 pairs shipped in a 40-foot shipping container, at an import cost less than $1,300 from China to US), with the Chinese labor cost at $6.14 per pair (via the published total number of imports of MBCT from China PRC and the total cost of those imports at import time). The difference in import and price includes the domestic shipping (truck drivers), retailing (inventory associates, cashiers, managers), logistics, and infrastructure (power, maintenance, rent) involved in local sale, as well as the profits.

If we paid American factory workers above $18/hr to make MBCT, with a retail average of $14.97, we would lose total American jobs; if we paid under $18/hr, we would gain jobs. This is because the cost of MBCT would increase, and the total purchaseable goods would thus decrease, impacting the entire logistics chain of shipping and selling them, as well as reducing the number of factory jobs to make them; and the factory jobs recovered from China are added to the job market, offsetting this. If more jobs are lost than gained, you lose jobs in total.

That's not the issue.

Say you pay your factory workers $21/hr, the same salary as a GM line worker. The price of MBCT goes up from $14.97 to $50.57 (remember: $8.83 of that goes to American wages for cashiers, truck drivers, shelf stockers, and the like, with some carved out for taxes and profits; I'm assuming profit margins and taxes fall instead of increasing as well, instead of adjusting that $8.83 larger). Today, a $21/hr income lets you buy MBCT at 0.71 hours's work per pair. With $21/hr factory workers, they'd work for 2.4 hours to afford a pair.

If you pay them an $8.25 minimum wage, the price rises to only $25. Today, an $8.25/hr wage lets you work for 1.8 hours and buy pants; at $25, an $8.25/hr wage requires you to work for 3.0 hours to afford the same pants you're making.

If you think that sounds ludicrous, consider: before globalization, the median American household spent 12% of its income on clothing; once we started outsourcing to China, this rapidly fell to 4%. It's now under 3.5%--it's only slowly continued to fall since the great globalization revolution. That means globalization in fact decreased costs to 1/3 what they were.

Outsourcing your jobs to another country that does the work a hell of a lot more cheaply creates an enormous capacity to buy, but somebody has to transport all that shit you're buying once it comes off the docks. You can't sail a ginormous shipping friggate up the mid-western basin to Colorado. That creates local jobs. Even then, unemployment dips nice and friendly-like, but it gets buffed out as population expands to fill the abundance of jobs, until some factor of scarcity (job scarcity, food scarcity, etc.) creates an expanding population in poverty and slows growth. Likewise, a small loss of jobs slows population growth until it adjusts to fit its economy's capacity, and can have a profound effect on the size of the labor force.

Comment Re:Americans? (Score 1) 320

Actually, there's something to that, but you're missing a factor.

Remember that part where I said the break-over point wasn't zero? For men and boys's cotton trousers and shorts eliminating all import from China, if all American jobs created to make those average above $18/hr wage, you have a net-loss of American jobs; if they average below $18/hr, you have a net gain. In either case, the cost of MBCT increases in terms of labor-hours, especially to the line workers--where a minimum-wage line worker model would raise the price from $14.97 average per pair to $25 and 1.8 hours's wage to 3.0, and a $21/hr worker model (GM factory line worker wage) would raise the price to $50.57/pair and from 0.71 hours's wage to 2.4.

In both cases, the worker making the given salary (and every other American) spends more time working to buy the same good. That's called "being poorer", and the net result is fewer things shipped, fewer things sold, fewer retail workers, and so forth. The break-over point is non-zero because while you reduce the number of goods shipped and sold, the reduced number of formerly-Chinese jobs are added onto American jobs when the dust settles: if you lose 40,000 Chinese manufacture jobs, 57,000 American retail jobs, and you transfer the remaining 57,000 manufacture jobs form China to America, you get a net-zero change in American employment. Reduce the purchaseable goods further (by making them more-expensive) and you lose jobs; reduce it less (by paying lower wages and thus lowering the price) and you gain jobs.

As for eliminating all American jobs, that's actually a viable goal. The main factor is time: if you eliminate some American jobs today, we will experience some unemployment, and also an increase in wealth (purchaseable goods per person). Give it time and our spending and labor force adjust to fit, buffing out the unemployment. Obviously, this means eliminating 30% of jobs today would wreck the economy; while eliminating 30% of today's jobs over the next 30 years would end in no increase in unemployment while every American ended up that much richer and capable of buying that much more stuff.

Projecting this back, far more than 100% of all jobs which have ever existed have been eliminated (by technical progress), and we've outsourced a hell of a lot of jobs. There are things Americans simply don't do anymore, and the outsourced workforce is enormous. Furthermore, combining trade with technical progress, the sheer amount of stuff we import greatly exceeds what we were able to make before it was outsourced--which is only an amusing interpretation, because it's also true that technical progress has replaced pretty much all of our jobs several times over, and now we make here in America several times more than we could make a hundred years ago.

The technical progress thing should be obvious, too. Imagine the farm workers it took to irrigate a field before we proposed a better way. It takes far fewer people to build, maintain, and operate the factories, pumps, and irrigation equipment to pivot-irrigate a farm than it does to have folks carry buckets of water back and forth all damned day.

Comment Re:Americans? (Score 3, Insightful) 320

Actually, it's not.

50,000 American jobs created by factory work? Okay. Now, those iPhones have to generate revenue to pay for those jobs. Apple has huge profit margins, so this isn't much fair (the US business average is under 10% profit margin); but let's say Apple isn't altruistic and is trying to keep those huge profit margins, or just pretend Apple is a normal US business with normal profit margins that fit its business growth and risk control needs (because the only violators of this are what, Apple, Google, and Microsoft?).

To keep the same margins at higher American worker costs, you charge more for the phone.

If you ship fewer phones, you'll have fewer jobs. That includes fewer exports, too, so less international revenue coming to the U.S.; but let's assume that doesn't happen. Everyone buys iPhones at $1,400 instead of $700.

Someone concluded 36 million iPhones sold in the US. If we're imagining a doubling in price above (for illustration; order of magnitude is controlled by depth of price difference, and the difference isn't at 0--we'll get to that), that's $25.2 billion. That's equivalent to 1.52 million minimum-wage incomes.

So for 50,000 jobs shipping 36 million iPhones to US customers (which I doubt actually happens) at $700 additional cost (doubling the price), you lose a maximum of 1.52 million jobs. It's only 50,000 jobs lost if the wages are on average $262/hr for those lost jobs ($524,000/year).

As I said: the price increase controls magnitude. If you increase it by $100 ($700 becomes $800), you're looking at $3.6 billion. That's 218,000 minimum-wage jobs, or 50,000 $72k jobs ($36/hr average wage). That's your exchange.

All of that is based on the sales of US phones to US people. That doesn't count international sales. The biggest take-aways here are that job creation or loss in practice depends on how much you pay the workers--pay them less and you create more jobs, as you noticed--and that everyone who isn't a factory worker and who buys the factory worker's product has less money to spend.

At best, this is a way to enrich factory workers at the expense of all other Americans, reducing the number of available products and services (e.g. we could have expensive iPhones and no Spotify) by drawing both domestic and international money to a subset of peoples's hands, with the international money being spendable back into the US economy. At worst, this is a way to create poor US factory workers, a poor US middle class, and less-competitive United States business, causing a rapid fall in sales as people in Europe roll their eyes at higher-priced iPhones and just go to buy the Chinese-made competitor's product--or maybe Apple will sell Chinese-made phones outside the US and stay competitive, but the US factory workers won't get that international money (14,000 employees at Apple HQ are still getting that cash and propping up Cupertino's economy with $2 billion of wages from across the world).

Again, as you observe: the net job change will be positive (an increase) if we pay the factory workers little and abuse them with minimal benefits and other cost-cutting measures, making them poor even as the products they produce become more-expensive than the import product. Even then, the US consumer still has less money to spend on everything else, and is thus poorer: he can buy fewer things with the same income.

Comment Re:Greed by any other name... (Score 1) 531

It's difficult. If the numbers don't make sense at a glance, look at them harder. The fault is more likely to be with you than with the numbers.

The problem is it is difficult, and we've created indicators that tell a funny story. GDP-per-capita tells what we produce; real median income tells what our income looks like per household in the center of the population (not the mean). People look at this and say, "Look! We produce twice as much, but our income adjusted to inflation is no higher!" The problem is that your income buys twice as much--or it roughly buys that proportion of the GDP-per-capita. That means when your real income doesn't go up at all, but your GDP-per-capita doubles, your standard-of-living has doubled.

Median income is what the average household makes, but per capita income is not by itself indicative of what individuals can buy.

The problem is: what is? Individuals are capable of buying a hell of a lot more today with their dollars than they could in 2000 or 1995.

If the upper decile suddenly gets a 10)% raise, the per capita GDP has gone up 10%, and the median income doesn't show any change.

GNP is nominal; GDP is inflation-adjusted. For GDP to go up, more has to actually be produced. Median, as you observe, is in the middle of the population, so changes at the ends that don't affect the middle don't change the median.

Per-capita GDP can only increase if more productivity occurs. If it's just more money, it's inflation.

Internet access has become, to many people, a necessity like telephone service. It's no longer just a luxury. The cost of necessities has not been going down like that of certain luxuries.

Hey, I buy 1,562 $35/month ISDN lines today, but I only pay $83/month.

Certain luxuries have been going down quickly; necessities are a different story. Fuel has been fluctuating, with natural gas getting cheap, and oil following. Shelter per square foot has decreased rapidly over the years, but has increased slightly during the mid-2000s housing market bubble (thanks to mortgage prime rates falling sharply and a bunch of marketing bullshit). Food productivity increased rapidly up to the 80s; and since then, food has only fallen in cost slowly.

The rapid march of technology has brought ridiculous, unsustainable, expensive things into the hands of everyone by making them common, sustainable, and cheap. Cell phones are a primary example, both in becoming viable ($4,000 for the phone and $250/month for 2hrs/week voice service in 1983 when they went commercial) and in becoming technically-advanced ($350 for a OnePlus phone, $30/month for service with high speed data that can support audio streaming, $40/month for service with high-speed data that can support voice calls; streaming 1080p video gets you into the high usage tiers).

Comment Re:Hey Slashdot: (Score 2) 131

When you trade money for news, you tend to get the news that makes the most money. It's human nature, unless controlled by regulation. Just as corporations, utilities, colleges, all mostly get financially out of hand unless regulated, because people are mostly naturally greedy. There's scant sense of fairness, and gross excess of "take the market for all it can bear."

Look, news is all mostly biased anyway. Biased by what they cover and what they choose not to cover; by the editor's influence; by the publisher's influence; by the advertiser's influence; by the stockholder's influence; by ridiculous "equal time for superstitious nonsense" policies (because the news consumers are bewildered, so in order to get their money, they are pandered to), etc. I'm just not going to actually pay for more bias.

It's a complete waste of time to put a paywalled link in front of me. Not going to click it if I know what it is; not going to stay if I am snookered into clicking.

For news, here's what I want: facts and relevance to actual news. Not the Kardumbians, not some actor's opinion, not breathless reporting of some lab result as if it was tech coming down next Friday, Politics, cover the candidates and what they say. Even handedly. Don't leave some out (Sanders, cough) don't over-cover some (Trump, cough), don't report bland, content free remarks as if they were incoming legal doom (Clinton, cough)... you get the idea.

Simple enough, you'd think. Just do a good job. But they don't. Okay then, fine. But expecting me to pay for that crap? Not happening. They oughta pay me for having to fact check every goddam thing they write and speak about.

Comment Re:That can't be right (Score 1) 533

Food happens to be a relatable tangible good. I've had trouble with people claiming things like Netflix or cellular communication aren't "making things" and that the US doesn't "make things" because anything that's not concrete isn't real. If people are going to point at an increase in medical care, high-speed internet availability, and personal entertainment services and call that "not really making anything" to argue that the economy is failing and the US has collapsed, I'm going to have to start pointing at things that people can actually relate to.

Things like food.

By the by, I like my phone, and I like my 200Mbit/s Internet; but poor people need food and shelter first and foremost, and their ability to survive (and the general ability of a population to expand) kind of relies on their ability to eat. I encourage you to try not eating for one month and tell me how it goes.

Comment Re:That can't be right (Score 1) 533

Uh, that rebuttle isn't so obvious. Food prices continue to be extremely low because population doesn't expand to drive prices up. The moment scarcity starts to set in, unemployment starts going up, poverty starts increasing, and we get more poor people and fewer middle- and upper-class. Do you see that happening around you?

My argument is that population expands to fit abundance. Do you see population rapidly expanding to consume all of our employment opportunities? What if I told you that the labor force would slow its expansion during high unemployment? What if that actually happened from 2008 to 2012? What if the population somewhat dipped during that time?

You haven't provided any argument that says that expanding beyond our means would not cause population to slow its growth, while I have shown good reasoning that it does and demonstrated the effect actually occurring during times when poverty (and thus individual access to means) has increased.

Comment Re:That can't be right (Score 1) 533

You don't know that. There are many companies that operate on a tight profit margin

UH, I said that maximized competition can't push prices down below actual costs. If your business spends $1,000 and charges the customer $500, it goes out of business. That is a 100% guarantee. That's not a "tight profit margin"; that's a loss.

Maximized competition and whatever else can't push prices down below actual costs. If you charge customers less than you spend to produce goods, you go out of business. If a hospital is spending $63,000 to provide care and charging $58,000 for that care, it goes out of business. Period.

Comment Re:That can't be right (Score 1) 533

Again, you're focused on insurance market competition and not on healthcare competition.

The insurance market is selling insurance to individuals.

The healthcare market is selling healthcare to patients.

The interaction between healthcare and insurance is that patients are less-accessible to healthcare providers if those providers don't provide rates to those patients which are competitive with the rates other providers provide; and that insurers engage in group negotiation on behalf of the insured.

That means healthcare providers are competing for patients when they negotiate with the insurance providers. The patients are paying healthcare providers to act as effective negotiators for them, and backing them with their patronage. If the healthcare providers don't negotiate a good deal with the consumer via their elected arbiter (insurer), then the consumer goes to someone who did negotiate a good deal with them; and we call those people who negotiate deals "in-network providers".

They also have very little reason to worry about competition since health markets tend to be very regional

So what? Nearly half of all providers are out of my care network. They're excluded from my healthcare options because they cost me three to ten times as much. If the hospitals, doctors, psychiatrists, and pharmacies wanted my business, they should have signed on with CareFirst's BlueChoice PPO network and CVS Caremark's Pharmacy Benefits Manager. I've got out-of-network providers charging me $1,079 for things my insurance won't cover, at all; my in-network providers have negotiated a rate of $101, refuse to cover it, and so I have to pay the whole $101 out-of-pocket. Guess who I go to for doctor's services, prescriptions, and drugs?

ACA has guaranteed their profit to be capped to a percentage of premiums

That's actually a constant, and it doesn't much matter if you don't have customers. Let's take a look before the ACA.

A great many insurance companies have operated as NPOs for decades. Healthcare Services Corporation is a customer-owned for-profit insurance company--the same way a Credit Union is a customer-owned bank--with 15 million customers. CareFirst Blue Cross formed as non-profit in the 1930s, and considered (but didn't go through with) converting to a for-profit entity in 2001. Most Blue Cross and Blue Shield association members are NPOs or customer cooperatives.

Because of the NPO status of most insurers, they operate largely on carry-over and can't distribute profits in giant bursts to board members. Essentially, NPOs pay no taxes on profits; and the IRS may revoke your 501(c)(3) status if you generate too much profit and don't spend it in pursuit of your NPO-eligible business activity. For an insurer, that means you have to spend your profits maximizing the effectiveness of your insurance.

For example: CareFirst's CEO, Chett Burrell, got $2.5 million in salary and incentive pay in 2013; in 2014, CareFirst had an $865 million surplus, which prompted the city of DC to investigate whether CareFirst DC had a compelling need to float that much cash. In 2015, CareFirst was ordered to spend $268 million of this surplus to pay for public health services in DC, Maryland, and Virginia, including wellness programs, free clinics, public vaccination programs, and so forth. If CareFirst doesn't reduce its excess holdings or simply starts liquidating them via $13 million dollar executive bonuses each year, the IRS can revoke their 501(c)(3) status and make them pay 40% of their profits (that means $350 million of that $865 million) in taxes.

How does that tie profits to premiums?

In this situation, as you observe, CareFirst's profits are reliant on their premiums: if they have higher costs, they can charge higher premiums; and if they have to charge higher premiums, they have to carry over a larger surplus to cover for risk. Adverse risk events would easily take down an insurer who didn't carry over enough money to pay for sudden, unplanned costs; the size of those events scales with the size of costs, meaning rendering the same services at half the cost allows you to secure yourself with half as big a bank account in reserve.

Thus, as you point out, bigger premiums means you can carry over more money, taking more profits, and paying bigger executive bonuses without drawing IRS scrutiny. You still can't pay dividends--unless you're a customer-owned cooperative, in which case you can pay dividends to your customers (including yourself).

The other side of this is you need customers in the first place.

ACA or not, "a percentage of your premiums" only matters if you actually manage to charge premiums. If your customers go on the exchange and see your shit-level plan for $800/month and a silver-level plan for $235/month, they're going to buy the silver-level plan. Insurers have an incentive to draw customers so they can keep a percentage of those premiums as profit.

The only way to get those premium prices down is to get costs down. To get costs down, you must get the healthcare providers to lower their prices. Healthcare providers can lower prices until margins make a sustainable profit, but no lower; to get lower, they must lower costs, which means medical devices, drug costs, and the efficiency of treatments (labor time spent) must go down. Turning a 27-hour surgery with 6 weeks of in-patient care and 4 months of follow-up into a 35-minute outpatient procedure really cuts down the amount of wages paid to doctors, nurses, and staff. We pretty much spend all of our time researching medical technology to cut those costs back in that way; less-traumatic surgery tends to go hand-in-hand with fewer complications and greater patient outcomes.

You can only disconnect competition from the healthcare insurance market by ignoring half of the stuff actually going on. You have to ignore insurance customers, or you have to ignore healthcare providers. If you include customers, insurers, and providers, you can only identify a highly-competitive market.

Slashdot Top Deals

Beware of Programmers who carry screwdrivers. -- Leonard Brandwein

Working...