Meteorite data show that Mars had a much denser atmosphere billions of years ago, perhaps as dense as half that of the Earth (now it is only 0.06 times as dense as the Earth's). Back then it had a magnetic field to avoid atmosphere stripping by the solar wind.
In France, spent nuclear fuel is reprocessed at the La Hague site, where the remnants are temporarily stored into the permanent underground repository. La Hague reprocessed 1700 tonnes of fuel per year.
France has a project to permanently store long-life nuclear waste 500 metres below ground in impermeable clay in Bure, eastern France, but the plan has not yet received government approval and is strongly opposed by local groups and environmentalists.
The RF exposure is far higher with closer distances, so this will be different than cell towers, etc.
This is unlikely. 5G microcells will be much lower power than large 3G cells. Plus your phone in your pocket will be able to transmit at lower power levels.
The real question is how much did Google and Netflix have to invest additionally for their content distribution infrastructure since Comcast didn't have to pay for it themselves under "net neutrality"?
Somewhere in every media company is a person updating a spreadsheet that maximizes revenue from pricing & licensing deals for every mode of distribution in every geography. Broadcast, OTT SVOD, OTT ad-supported VOD, TV Everywhere, MVPD, MVPD VOD, Blu-Ray, DVD, etc.
Failures include retention errors caused due to leakage current, which worsens with time when not acted upon. Second, they also suffer from phenomenon such as read disturb and program disturb errors, where read or program of a row or block of cells affects the threshold voltage of untouched cells in its vicinity. data retention, program disturb, read disturb, endurance, and power faults.
Flash controllers have proactive and reactive mechanisms in place, to prevent the flash error propagation to higher levels in the system stack. Consequently, not all of the above-mentioned failures propagate to upper layers. But, ones that do propagate can result in fail-stop failures.
When my wife determines that it is no longer profitable for her to work because of the expansion of the Social Security payroll tax, she will drop out of her tech job and just watch the kids. Then SS gets no cash from her.
If a worker get injured or fired, it was nearly a death sentence, because it wasn't a case of getting an other job, but spending the rest of your life homeless, barring extensive effort in finding new work.
Your family would often take care of you. This happened to several members of my family in the 1800's. Instead of paying 20-30% in income tax to the government to take care of the invalid, they paid 20-30% of their income to take care of their invalid relatives.
People certainly are "retiring" in terms of collecting Social Security old age benefits.
There were about 2.8 workers for every Social Security Old Age old age, survivors, and disability insurance (OASDI) beneficiary in 2017. This is down from 3.7 in 1970, and 5.1 in 1960. Te ratio had been stable, remaining between 3.2 and 3.4 from 1974 through 2008, but now going down quickly.
The Social Security trustees forecast that the ratio of workers to beneficiaries reaches 2.2 by 2035 when the baby-boom generation will have largely retired.
"No one has "live" encoded video on anything other than dedicated encoding hardware in a really REALLY long time."
That is actually not true - the industry is moving away from specialized encoding ASICs to more flexible VM-based encoders (for example see the Cisco/Synamedia Virtual DCM.
But regardless, it takes years for the ASICs to be spun up, and complexity is complexity. ASICs are more efficient than CPU/GPU, but they still need to do all the computational work, so an HEVC encoder ASIC will take a lot less power than an AV1 encoder ASIC.
No wonder the public cloud people like it, you'll pay them a ton of money encoding files!
Live is going to be a real hassle, almost impossible. It is all we can do to get a good 4K live encode with HEVC at reasonable bit rates.
What is really exciting is the next MPEG JVET codec, VVC (likely H.266). Even better performance than AV1 or HEVC, but with a minor increase in complexity.
The only thing that hasn't been skyrocketing are wages.
According to the BLS Employment Cost Index, September 2018:
Compensation costs for civilian workers increased 0.8 percent, seasonally adjusted, for the 3-month period ending in September 2018...Wages and salaries (which make up about 70 percent of compensation costs) increased 0.9 percent and benefit costs (which make up the remaining 30 percent of compensation) increased 0.4 percent from June 2018.
nternational graduate students, out of state graduate students, other international and out of state students, and (finally) in state students. I will let you guess why those are the recruiting priorities.
Because non in-state students pay higher tuition rates?
You realize that our productivity rises are do to technological advances, not actually people working 50% harder.
Indeed, productivity can rise from capital investments (such as machinery & automation), labor human capital improvement (going to chef school to learn how to cook more efficiently), management human capital improvement (learn how to more effectively run a business), and government/regulatory improvements (more efficient permitting process, reducing time to clear customs, etc.).
Right wing think tanks will tell you that it's because greedy schools are following supply and demand and raising prices to soak up quick cash. I've got a kid in school and can tell you this is bullshit. There were half as many slots in my kid's 300 level classes as there were qualified applicants (GPA 3.8 or higher, yes, that's an eight, not a zero). If the schools were pricing based on demand they'd just raise the price until they got the right number of applicants.
I wish we had better historical data on actual college spending per student, but I really have to think it has risen. The quality of the physical plant of most public universities I know of have risen dramatically (dorms used to be cinder-block "jails without bars").
Also the number of executive, administrative and managerial employees on US university campuses are up by 15% between 2007 and 2014.
I think it is a realistic market model to expect that some universities wish to realize the high student payments that can be extracted by the richest families, and thus many universities are competing to be "highly selective". If you "just let anyone attend", your selectivity will become low, and thus you will attract students of less well-off families who can pay less. "Less selective" universities may also face reduction in ability to enter into non-educational research transactions, and will attract less alumni funding.
Today we're at about 70% of recent high school grads enrolling in college, which is up from 60% in 1990. so clearly colleges are enrolling more students, again an indication that the market model is correct. It might be that "more selective" colleges are happy not to admit more, but "less selective" colleges are expanding enrollment.
It should also be noted that 13.7% (44.5 million people) in the US are foreign born, meaning they chose to come to the US for a better life. They are likely doing much better than if they remained where they were, even if they bring down average US compensation data.
The question, though, is what is the distribution of those two items. Has the rise in average productivity been driven by the productivity of a few, high-earning non-average workers, or has productivity risen for all workers? Also, what portion of productivity increase has come from capital investment in equipment as opposed to improved labor skills?
These are US numbers as well. It is possible that US productivity increases may be due to improvements in labor and capital in countries like China (where labor has seen tremendous wage increases) which provide inputs into US companies.
Slashdot Mirror
Meteorite data show that Mars had a much denser atmosphere billions of years ago, perhaps as dense as half that of the Earth (now it is only 0.06 times as dense as the Earth's). Back then it had a magnetic field to avoid atmosphere stripping by the solar wind.
In France, spent nuclear fuel is reprocessed at the La Hague site, where the remnants are temporarily stored into the permanent underground repository. La Hague reprocessed 1700 tonnes of fuel per year.
France has a project to permanently store long-life nuclear waste 500 metres below ground in impermeable clay in Bure, eastern France, but the plan has not yet received government approval and is strongly opposed by local groups and environmentalists.
The RF exposure is far higher with closer distances, so this will be different than cell towers, etc.
This is unlikely. 5G microcells will be much lower power than large 3G cells. Plus your phone in your pocket will be able to transmit at lower power levels.
The real question is how much did Google and Netflix have to invest additionally for their content distribution infrastructure since Comcast didn't have to pay for it themselves under "net neutrality"?
Somewhere in every media company is a person updating a spreadsheet that maximizes revenue from pricing & licensing deals for every mode of distribution in every geography. Broadcast, OTT SVOD, OTT ad-supported VOD, TV Everywhere, MVPD, MVPD VOD, Blu-Ray, DVD, etc.
The correct philosophy is to assume YOU ARE ALREADY COMPROMISED.
I don't think this was SJW, this was Verizon buying Yahoo! for $4.8 billion and finding out you can't sell much advertising on "porn".
See SSD Failures in Datacenters: What? When? and Why?.
Failures include retention errors caused due to leakage current, which worsens with time when not acted upon. Second, they also suffer from phenomenon such as read disturb and program disturb errors, where read or program of a row or block of cells affects the threshold voltage of untouched cells in its vicinity. data retention, program disturb, read disturb, endurance, and power faults.
Flash controllers have proactive and reactive mechanisms in place, to prevent the flash error propagation to higher levels in the system stack. Consequently, not all of the above-mentioned failures propagate to upper layers. But, ones that do propagate can result in fail-stop failures.
Beverly Hills is STILL fighting the purple line subway.
When my wife determines that it is no longer profitable for her to work because of the expansion of the Social Security payroll tax, she will drop out of her tech job and just watch the kids. Then SS gets no cash from her.
If a worker get injured or fired, it was nearly a death sentence, because it wasn't a case of getting an other job, but spending the rest of your life homeless, barring extensive effort in finding new work.
Your family would often take care of you. This happened to several members of my family in the 1800's. Instead of paying 20-30% in income tax to the government to take care of the invalid, they paid 20-30% of their income to take care of their invalid relatives.
People certainly are "retiring" in terms of collecting Social Security old age benefits.
There were about 2.8 workers for every Social Security Old Age old age, survivors, and disability insurance (OASDI) beneficiary in 2017. This is down from 3.7 in 1970, and 5.1 in 1960. Te ratio had been stable, remaining between 3.2 and 3.4 from 1974 through 2008, but now going down quickly.
The Social Security trustees forecast that the ratio of workers to beneficiaries reaches 2.2 by 2035 when the baby-boom generation will have largely retired.
"No one has "live" encoded video on anything other than dedicated encoding hardware in a really REALLY long time."
That is actually not true - the industry is moving away from specialized encoding ASICs to more flexible VM-based encoders (for example see the Cisco/Synamedia Virtual DCM.
But regardless, it takes years for the ASICs to be spun up, and complexity is complexity. ASICs are more efficient than CPU/GPU, but they still need to do all the computational work, so an HEVC encoder ASIC will take a lot less power than an AV1 encoder ASIC.
An AV1 encoder is about 130x times more complex to encode than HEVC (H.265) when comparing reference models.
See UNDERSTANDING THE VIDEO CODEC JUNGLE: A COMPARISON OF TCO AND COMPRESSION EFFICIENCY.
No wonder the public cloud people like it, you'll pay them a ton of money encoding files!
Live is going to be a real hassle, almost impossible. It is all we can do to get a good 4K live encode with HEVC at reasonable bit rates.
What is really exciting is the next MPEG JVET codec, VVC (likely H.266). Even better performance than AV1 or HEVC, but with a minor increase in complexity.
The only thing that hasn't been skyrocketing are wages.
According to the BLS Employment Cost Index, September 2018:
Compensation costs for civilian workers increased 0.8 percent, seasonally adjusted, for the 3-month period ending in September 2018...Wages and salaries (which make up about 70 percent of compensation costs) increased 0.9 percent and benefit costs (which make up the remaining 30 percent of compensation) increased 0.4 percent from June 2018.
Don't worry - Amazon, Google, Facebook, and Twitter have competition: Alibaba, Baidu, Tencent and Xiaomi
nternational graduate students, out of state graduate students, other international and out of state students, and (finally) in state students. I will let you guess why those are the recruiting priorities.
Because non in-state students pay higher tuition rates?
You realize that our productivity rises are do to technological advances, not actually people working 50% harder.
Indeed, productivity can rise from capital investments (such as machinery & automation), labor human capital improvement (going to chef school to learn how to cook more efficiently), management human capital improvement (learn how to more effectively run a business), and government/regulatory improvements (more efficient permitting process, reducing time to clear customs, etc.).
Right wing think tanks will tell you that it's because greedy schools are following supply and demand and raising prices to soak up quick cash. I've got a kid in school and can tell you this is bullshit. There were half as many slots in my kid's 300 level classes as there were qualified applicants (GPA 3.8 or higher, yes, that's an eight, not a zero). If the schools were pricing based on demand they'd just raise the price until they got the right number of applicants.
Current US average state/local spending on higher education is $7,120 per student. And that is before any student tuition payments.
I wish we had better historical data on actual college spending per student, but I really have to think it has risen. The quality of the physical plant of most public universities I know of have risen dramatically (dorms used to be cinder-block "jails without bars").
Also the number of executive, administrative and managerial employees on US university campuses are up by 15% between 2007 and 2014.
I think it is a realistic market model to expect that some universities wish to realize the high student payments that can be extracted by the richest families, and thus many universities are competing to be "highly selective". If you "just let anyone attend", your selectivity will become low, and thus you will attract students of less well-off families who can pay less. "Less selective" universities may also face reduction in ability to enter into non-educational research transactions, and will attract less alumni funding.
Today we're at about 70% of recent high school grads enrolling in college, which is up from 60% in 1990. so clearly colleges are enrolling more students, again an indication that the market model is correct. It might be that "more selective" colleges are happy not to admit more, but "less selective" colleges are expanding enrollment.
"If you believe real inflation has been low, I'm got a great deal - just for you! - on some delightful oceanfront property in Phoenix, Arizona."
Since 1991, the US inflation rate has been under 5%.
Compare this with 1969-1982 when inflation was over 5% for 12/13 years and over 10% for 4/13 years.
"Anything more complex than oatmeal or beans and rice has skyrocketed in price over the last decade. "
Consumer Price Index for All Urban Consumers: Food at home is only up 15% over the last 10 years. (That is a real price, of course, which tracks very closely with the 17% inflation rate).
"Real wages in the west have stagnated or gone down for two decades now."
Real total compensation per hour in the US is up 20% over the last 20 years.
It should also be noted that 13.7% (44.5 million people) in the US are foreign born, meaning they chose to come to the US for a better life. They are likely doing much better than if they remained where they were, even if they bring down average US compensation data.
Don't live in SF where government artificially restricts the housing market. Move to Memphis, TN, where the median home price is $82,700.
"We've doubled productivity in the last 20 years while wages remained the same or went down. "
Average non-farm real output of workers per hour has risen by 50% over the least 20 years.
Average non-farm real total compensation of workers per hour has risen by 20% over the last 20 years.
The question, though, is what is the distribution of those two items. Has the rise in average productivity been driven by the productivity of a few, high-earning non-average workers, or has productivity risen for all workers? Also, what portion of productivity increase has come from capital investment in equipment as opposed to improved labor skills?
These are US numbers as well. It is possible that US productivity increases may be due to improvements in labor and capital in countries like China (where labor has seen tremendous wage increases) which provide inputs into US companies.
Who do you *think* that benefits? The consumer? You think your service bill is going to go down?
I think you will find that the US mobile users experienced a 400-fold price-per-Internet-bps decline over the last 12 years.