My very layman understanding of the brain is because of the way the synapses interact, they also act as part of the processing. You may think of the brain as an ASIC special purpose CPU, but which ASIC that is used is determined by the input. As input triggers a neuron to do something, that neuron may trigger several other neurons in different orders, timings, and strengths. Each of these many combinations causes a cascade of triggering other such neurons in the same way. After only "50 cycles", a cat is identified. Millions of neurons and billions of synapses may have been used during this time.
Using the same neurons in different timings and strengths results in different answers. It's like a hybrid of CPUs, memory, and routing. I was also under the impression that neurons use a hybrid of analog and digital processing, and even a possibility of quantum.
Because which input synapses and the signal strength of those synapses affect how the neuron sends data to the output synapses, you could say the neurons also have routing inside of themselves.
Pretend you have a neuron with 2500 input and 2500 output synapses. Lets assume each synapse has some 8 states it can be in. Not sure if it's even remotely close, but lets go with it. The neuron can have 8^2500 different inputs resulting in a probably constant-time O(1) result in output. That one neuron. If a single neutron can translate an 8^2500 number of inputs into a useful output of a partial amount of data processed, it doesn't take too many iterations to process data.
Depends on your work load. If you spend a lot of time idle, a slower simpler CPU can save more power, but you still have the fixed waste electricity per power supply. Some work loads are more efficient if you have a single large server with strong CPUs and use virtualization to increase density.
Even in the situation of many-core CPUs with lots of simple cores, many cores do no scale well for most work loads. If your workload has lots of random memory access, a multi-core large CPU can save more power.
Or you're driving many miles. The rate at which they get read-ended is directly attributed to the number of miles driven. They drive more miles, they get hit more often.
I guess you've never seen a lay person with 12 months of updates that they never bothered to install after their tech friend disabled Windows Updates to automatically install. The average end user is too stupid to make educated decisions. They want their cake and eat it to. They don't want to ever be inconvenienced, they don't every want to reboot, they don't ever want to make decisions, they don't want their system to have malware, they don't want to be told not to run naked.britney.exe from their email, they don't want to be told not to install yet another search bar, they don't want all of those search bars, etc.
All of that being said, PC-BSD uses ZFS to snapshot the entire OS and allows you to roll-back. And because it's ZFS, rolling back is an O(1) perceptibly instant operation that only requires a reboot. Windows need a modern FS and modularize their OS to synergize with snapshots.
*Purchase 100Gb/100Gb at your local Internet Exchange for $5k/month. Assuming the national **average of 10Mb/s, a $30/month bill, and a fairly standard 20:1 oversubscription ratio, that 100Gb will bring in $6mil/m of revenue. That's a 0.083% overhead cost, leaving you with 99.916% for whatever else.
**Some ISPs are much worse, like Comcast, Verizon, etal of the past, going past 20:1 and charging closer to $70/month and still unable to stream 3Mb/s videos. If you adjust for these target values, 100Gb will bring in $70mil/m of revenue, bringing that $5k/month down to 0.007% overhead, leaving them with 99.993% margin for other costs and profits.
Free peering typically means that each carrier sends and receives about the same amount of data
That only applies to transit providers where the data needs to get hauled. Residential ISPs are the endpoint, and as long as Netflix is dropping off the data near the customer, it's ALWAYS in the ISP's best interest to get free bandwidth than to pay their transit provider for access. Your logic falls flat on its face with the simple example that my ISPs sends me way more data than I send it, so it should pay me.
However, if that wasn't clear, and this setting involves uploading data to Google silently and automatically, then the current behaviour is shady as hell. A device that is recording and/or uploading anything without its user's knowledge, or worse when its user explicitly thinks they have turned that behaviour off, is always a usability and privacy issue, and it is always the software developers' responsibility to fix it.
I don't know about Google Photo, but when I just recently got my first smart phone, it prompted me during the setup if I want to backup my phone's data using Google Sync. At least during this configuration of the sync functionality, it was made clear that pretty much all of my data would be getting sent to Google.
We have to pay interns to attract them. We also need to have enough work to keep them busy. If we can't guarantee them enough money to have fun during the summer, we won't win them over.
My federal taxes are much lower because my gross income is lower. They can deduct more than me, but it doesn't make up enough of a difference. Not to mention being able to live 5-10 minutes from work.
"Average" is kind of useless when you have a large variety that are directly dependent on the local market. I think a better metric would be what percentile the wage is relative to the local economy of where the job is located. Fresh out of college, I had a lot of jobs to choose from, all at or above the median household wage for the city it was located.
HTTPS proxies are also massive security holes, it break authentication. Trading one evil for another. The most recent OpenSSL bug that was declared a huge security issue is pretty much what HTTPS proxies do. Blindly forwarding traffic from the Internet and signing it.
A person is not fractional, so a single person doing a job faster may not benefit, but people can be fractional, and reducing the number of toilet cleaners by 50% can save a lot of money.
Adblock plus has a coloured history of cherrypicking advertisers to quietly ignore
They openly advertise this, pun intended. They do claim to have a minimum quality requirement for whitelisted companies, and companies need to pay to get on that white list. Breaking the terms by having questionable ads can remove the companies from the whitelist forcing them to pay all over again. I seems to be done in a way that would cost a company more to break the rules.
I know a lot of dim-wits that can do math in their head but have no ability to apply math. I can't do math in my head but I can apply it. Two separate issues at hand.
If I get fewer than 7 hours of sleep, yes even 6 hours, for more than 7 days in a row, I start to get dizzy spells, numbness of limbs, muscle twitching, migraines, strange distortions to my vision, hearing strange sounds, confusion of where I am, I can feel my pulse throughout my body, random weakness in that may cause my legs to give out, my head to suddenly tip, or my hands to lose grip.
I assume this is bad, so I make sure I get at least 7+ hours. Of course I can handle 1-2 days of almost no sleep with almost no issue if I'm "caught up" on sleep.
I read something at some point that mentioned that while the eye can only see only a few million colors, it can see many many shades of each color. Which is why we easily see color banding.
Slashdot Mirror
My very layman understanding of the brain is because of the way the synapses interact, they also act as part of the processing. You may think of the brain as an ASIC special purpose CPU, but which ASIC that is used is determined by the input. As input triggers a neuron to do something, that neuron may trigger several other neurons in different orders, timings, and strengths. Each of these many combinations causes a cascade of triggering other such neurons in the same way. After only "50 cycles", a cat is identified. Millions of neurons and billions of synapses may have been used during this time.
Using the same neurons in different timings and strengths results in different answers. It's like a hybrid of CPUs, memory, and routing. I was also under the impression that neurons use a hybrid of analog and digital processing, and even a possibility of quantum.
Because which input synapses and the signal strength of those synapses affect how the neuron sends data to the output synapses, you could say the neurons also have routing inside of themselves.
Pretend you have a neuron with 2500 input and 2500 output synapses. Lets assume each synapse has some 8 states it can be in. Not sure if it's even remotely close, but lets go with it. The neuron can have 8^2500 different inputs resulting in a probably constant-time O(1) result in output. That one neuron. If a single neutron can translate an 8^2500 number of inputs into a useful output of a partial amount of data processed, it doesn't take too many iterations to process data.
CPUs are two dimensional, 28nm is 4x as large as 14nm.
Depends on your work load. If you spend a lot of time idle, a slower simpler CPU can save more power, but you still have the fixed waste electricity per power supply. Some work loads are more efficient if you have a single large server with strong CPUs and use virtualization to increase density.
Even in the situation of many-core CPUs with lots of simple cores, many cores do no scale well for most work loads. If your workload has lots of random memory access, a multi-core large CPU can save more power.
Or you're driving many miles. The rate at which they get read-ended is directly attributed to the number of miles driven. They drive more miles, they get hit more often.
ignorant Windows users can be about security
It's not "windows users", it's "users". Just feel lucky that Windows users haven't moved to Linux yet. Once they do, the ecosystem will get polluted.
I guess you've never seen a lay person with 12 months of updates that they never bothered to install after their tech friend disabled Windows Updates to automatically install. The average end user is too stupid to make educated decisions. They want their cake and eat it to. They don't want to ever be inconvenienced, they don't every want to reboot, they don't ever want to make decisions, they don't want their system to have malware, they don't want to be told not to run naked.britney.exe from their email, they don't want to be told not to install yet another search bar, they don't want all of those search bars, etc.
All of that being said, PC-BSD uses ZFS to snapshot the entire OS and allows you to roll-back. And because it's ZFS, rolling back is an O(1) perceptibly instant operation that only requires a reboot. Windows need a modern FS and modularize their OS to synergize with snapshots.
Bandwidth is *free, transit is not.
*Purchase 100Gb/100Gb at your local Internet Exchange for $5k/month. Assuming the national **average of 10Mb/s, a $30/month bill, and a fairly standard 20:1 oversubscription ratio, that 100Gb will bring in $6mil/m of revenue. That's a 0.083% overhead cost, leaving you with 99.916% for whatever else.
**Some ISPs are much worse, like Comcast, Verizon, etal of the past, going past 20:1 and charging closer to $70/month and still unable to stream 3Mb/s videos. If you adjust for these target values, 100Gb will bring in $70mil/m of revenue, bringing that $5k/month down to 0.007% overhead, leaving them with 99.993% margin for other costs and profits.
Bandwidth is cheap, transit is not.
Free peering typically means that each carrier sends and receives about the same amount of data
That only applies to transit providers where the data needs to get hauled. Residential ISPs are the endpoint, and as long as Netflix is dropping off the data near the customer, it's ALWAYS in the ISP's best interest to get free bandwidth than to pay their transit provider for access. Your logic falls flat on its face with the simple example that my ISPs sends me way more data than I send it, so it should pay me.
If it wasn't, I fear you're overly serious and part of that group of people that makes this world a horrible place.
>
However, if that wasn't clear, and this setting involves uploading data to Google silently and automatically, then the current behaviour is shady as hell. A device that is recording and/or uploading anything without its user's knowledge, or worse when its user explicitly thinks they have turned that behaviour off, is always a usability and privacy issue, and it is always the software developers' responsibility to fix it.
I don't know about Google Photo, but when I just recently got my first smart phone, it prompted me during the setup if I want to backup my phone's data using Google Sync. At least during this configuration of the sync functionality, it was made clear that pretty much all of my data would be getting sent to Google.
We have to pay interns to attract them. We also need to have enough work to keep them busy. If we can't guarantee them enough money to have fun during the summer, we won't win them over.
My federal taxes are much lower because my gross income is lower. They can deduct more than me, but it doesn't make up enough of a difference. Not to mention being able to live 5-10 minutes from work.
"Average" is kind of useless when you have a large variety that are directly dependent on the local market. I think a better metric would be what percentile the wage is relative to the local economy of where the job is located. Fresh out of college, I had a lot of jobs to choose from, all at or above the median household wage for the city it was located.
HTTPS proxies are also massive security holes, it break authentication. Trading one evil for another. The most recent OpenSSL bug that was declared a huge security issue is pretty much what HTTPS proxies do. Blindly forwarding traffic from the Internet and signing it.
A person is not fractional, so a single person doing a job faster may not benefit, but people can be fractional, and reducing the number of toilet cleaners by 50% can save a lot of money.
Probably is 5x for simple sites and much less for YouTube like sites where the content is large.
Adblock plus has a coloured history of cherrypicking advertisers to quietly ignore
They openly advertise this, pun intended. They do claim to have a minimum quality requirement for whitelisted companies, and companies need to pay to get on that white list. Breaking the terms by having questionable ads can remove the companies from the whitelist forcing them to pay all over again. I seems to be done in a way that would cost a company more to break the rules.
How's that going to work once most everything is HTTPS?
No, they charge you for the channels, which are the content.
Better outlaw bit rot, it changes binaries all the time. Non-checksumming filesystems are now illegal.
100% faster just means 2x.
I know a lot of dim-wits that can do math in their head but have no ability to apply math. I can't do math in my head but I can apply it. Two separate issues at hand.
If I get fewer than 7 hours of sleep, yes even 6 hours, for more than 7 days in a row, I start to get dizzy spells, numbness of limbs, muscle twitching, migraines, strange distortions to my vision, hearing strange sounds, confusion of where I am, I can feel my pulse throughout my body, random weakness in that may cause my legs to give out, my head to suddenly tip, or my hands to lose grip.
I assume this is bad, so I make sure I get at least 7+ hours. Of course I can handle 1-2 days of almost no sleep with almost no issue if I'm "caught up" on sleep.
I need at least 7-8, but 9 is optimal.
I read something at some point that mentioned that while the eye can only see only a few million colors, it can see many many shades of each color. Which is why we easily see color banding.