Around here, the power company has to pay you the amount that they charge you. If they charge 6c/KWH at night and charge 9c/KWH during the day, they must pay you that same difference.
A heavy read database, like a news site, will have nearly everything cached in memory.
I've done ad-hoc aggregate functions against non-indexed tables with over 100mil rows, and they return in sub 10ms times. Cached table data can be fast.
1) Energy locality. Cars charge at night when the load is low and discharge during the day when load is high, and the cars are quite *near* the places that consume power. Less load on the transmission lines
2) Well... Extra energy supply while load is high. Less reason for rolling blackouts.
3) With more and more green energy, power supply can increase and decrease quite rapidly. This is very hard on power plants.
In the end, it's not a question of efficiency, it's a question of cost. It will cost less to give away some free electricity and soak the lost energy from power conversions, than to keep adjusting the power plants.
There's also a good chance that efficiency is the same or better because of lost energy of the system unable to absorb sudden excess power generated from green sources. You also gain value from less peak load on the system.
There are a lot of variables in a "smart grid", and they all need to be included. My "guess" would be that it's an over all better design.
There was talk that the benefit of Electronic Vehicles as a distributed storage medium could be so great, that one might be able to power a car like a Nissan Leaf for free for some people. The idea is that one gets "credits" from the power company when allowed to use your car to help smooth out the power. This "leveling" is so useful, that they would be willing to give you some free power. The amount of free power would be a sizable amount compared to the "average" distance traveled to work.
To go along with that, Toyota traded someone for their old Prius because it was about 10 years old and had something like 250k+ miles on it. The batteries could still hold 80% of their charge, and it was a car with heavy city mileage on it, so lots of charging/discharging. This was with an older car with older batteries.
The new battery charging systems are VERY intelligent.
Like you said, "most modern batteries die of age rather than cycles"
Great for servers, bad for clusters. Very power efficient, but very picky on the type of work they do. Very expensive, but worth it if you need high densities. Semi-niche market.
The T3 has a shared FPU among all those cores. The FP performance is horrible, but it can crank out the int performance required for an efficient web/db server.
When reading about the "smart grid" many years ago, they talked about programmable settings. You say by which time you want the battery topped off, the lowest amount of discharge, stuff like that.
The general idea is most cars are parked during peak usage. So, in the middle of the hot day when you're eating your lunch, your car is helping to supply power to the grid. By the time you get out of work, your car is filled back up.
The idea is you also get "credit" for helping. So you actually get some amount of free electricity for helping smooth the load.
In the middle of the night, power usage is the lowest, so that's the best time to charge the car. Again, all customizable. If you don't like it, don't enable it.
Fake News: In other news, OEM X hast lost millions of dollars because Linux runs over 50% of the internet and all those data centers can't use their servers. OEM Y saw this as an opportunity and now offers manageable certs.
What makes you think OEMs won't allow manageable certs when it's easy for them to do so? Actually, it's beyond "easy" as the default is to allow it from UEFI motherboard makers. OEMs would have to specifically request to have that feature disabled. Also sounds like a great way for a class action law suit or even anti-trust.
Where are you getting this info that IT can't sign their own boot loaders?
"allow customers to import and manage those certificates, and manage secured boot"
If the OEM doesn't allow it, don't buy it. In a market where competition is fierce and Linux is quite "main stream" for enthusiasts/servers, there will be someone else willing to allow certs to be managed.
"What are the socioeconomic ramifications of the disease and how do they reflect upon any surviving children? How will social stigmas affect copulation opportunities? These and other feedback mechanisms cannot be marginalized by hyperbole."
When you're talking about natural-selection, quantity > quality, so long as the quality of the quantity is just enough to keep it running.
When you got countries with 30% and higher HIV rates, it's not an issue of having it or not, it's how often you have sex that determines if you have kids to keep your DNA alive.
There are two types of "deleted" data. There is "remapped" deleted data caused by wear leveling, and "File System" deleted data.
In the case of a deleted file, the TRIM command would let the SSD know that it can clear that block, In the case of no TRIM, the SSD will still think data is there. If the block is cleared and the SSD writes to it, nothing has to be done. If TRIM has not happened and the block is not cleared, the SSD will read up that block, remap it, then write that data to another block.
In this case, TRIM saves a read. There are worse cases.
Assume you deleted a file and the filesystem deleted all 256k from a block. To the FS, that 256k is clear, but to the SSD, it isn't. Now assume you write 128k of data to the block. With TRIM, that 128k get written, no extra work to be done and the other free 128k can be written to with no extra work either. Without TRIM, the SSD copies up all 256k of that block, then modifies 128k of it, then writes all 256k to another block. If the other 128k get modifed, the SSD must again read up all 256k of data, modify the other 128k, then write to another block.
In this case, you save two reads and one write. You gained write performance while saving an extra write cycle.
The real question is how much overhead and complexity TRIM adds.
"Abstinence" is good to teach because it is sound advice, but *most* of the time, it doesn't work or just postpones it a bit.
If this was a perfect world, then no one would have to worry about wars/poverty/etc. Alas, people are human, and society has to pay for the mistakes of others. Anyone who thinks the general population will remain abstinate, is naive(willful ignorance). Any policies based around these as core ideas will just cause more harm than good. "Abstinence" can augment other ideas, but it should never be a "core" idea, except for religious reasons.
"People having sex" is as much the reason for people having AIDs as people sports getting broken bones. It may contribute a lot, but isn't the only reason.
I'm also assuming, from your judgmental tone, that you think people "sleeping around" is the issue. Well, some times some people cheat on their spouses and the other spouse doesn't know. The person who is "faithful" still gets AIDs.
If they didn't get over 800bil in subsidies over the past 15 years and didn't run their fiber over public land and other's private property, then you may have a case. But as it stands, they've gotten over 1 trillion(after inflation) of free money and they run their cables along high-ways and cross state lines. There is NOTHING private about that.
When you pay for "the internet", you're paying for full access to the "public network". If the company interferes with that, then they should not be able to advertise it as an "internet connection" in any form.
Also, your connection may flow "over" their private network, but you do not actually have access to that private network. Not to mention, their "private network" runs over public lands and crosses state lines.
If they don't like it, then they should have to fully purchase an own the land where their fiber/cables run.
A read-only bootloader is a horrible idea for a common desktop computer. IT would hate you.
"That this will block certain classes of rootkits is entirely incidental"
If everything from the boot process to the software is signed and requires signature, malware's only hope to getting installed is bugs in the OS, and even then, it would go away on reboot. Even if the malware could insert itself into some start-up location, it would never start because it wasn't signed. The uEFI won't load it, the bootloader won't load it, the OS won't load it.
Slashdot Mirror
Which is why you tell the system to only discharge your battery to a certain point
"Keep my battery at: 80%" There you go, never a dead car.
"Keep my battery at least 70% and have it at 80% by 4:30p" there you go
We're not talking about dumb systems here.
Around here, the power company has to pay you the amount that they charge you. If they charge 6c/KWH at night and charge 9c/KWH during the day, they must pay you that same difference.
A heavy read database, like a news site, will have nearly everything cached in memory.
I've done ad-hoc aggregate functions against non-indexed tables with over 100mil rows, and they return in sub 10ms times. Cached table data can be fast.
1) Energy locality. Cars charge at night when the load is low and discharge during the day when load is high, and the cars are quite *near* the places that consume power. Less load on the transmission lines
2) Well... Extra energy supply while load is high. Less reason for rolling blackouts.
3) With more and more green energy, power supply can increase and decrease quite rapidly. This is very hard on power plants.
In the end, it's not a question of efficiency, it's a question of cost. It will cost less to give away some free electricity and soak the lost energy from power conversions, than to keep adjusting the power plants.
There's also a good chance that efficiency is the same or better because of lost energy of the system unable to absorb sudden excess power generated from green sources. You also gain value from less peak load on the system.
There are a lot of variables in a "smart grid", and they all need to be included. My "guess" would be that it's an over all better design.
There was talk that the benefit of Electronic Vehicles as a distributed storage medium could be so great, that one might be able to power a car like a Nissan Leaf for free for some people. The idea is that one gets "credits" from the power company when allowed to use your car to help smooth out the power. This "leveling" is so useful, that they would be willing to give you some free power. The amount of free power would be a sizable amount compared to the "average" distance traveled to work.
To go along with that, Toyota traded someone for their old Prius because it was about 10 years old and had something like 250k+ miles on it. The batteries could still hold 80% of their charge, and it was a car with heavy city mileage on it, so lots of charging/discharging. This was with an older car with older batteries.
The new battery charging systems are VERY intelligent.
Like you said, "most modern batteries die of age rather than cycles"
Great for servers, bad for clusters. Very power efficient, but very picky on the type of work they do. Very expensive, but worth it if you need high densities.
Semi-niche market.
The T3 has a shared FPU among all those cores. The FP performance is horrible, but it can crank out the int performance required for an efficient web/db server.
Don't expect Ray Tracing with an INT heavy CPU.
When reading about the "smart grid" many years ago, they talked about programmable settings. You say by which time you want the battery topped off, the lowest amount of discharge, stuff like that.
The general idea is most cars are parked during peak usage. So, in the middle of the hot day when you're eating your lunch, your car is helping to supply power to the grid. By the time you get out of work, your car is filled back up.
The idea is you also get "credit" for helping. So you actually get some amount of free electricity for helping smooth the load.
In the middle of the night, power usage is the lowest, so that's the best time to charge the car. Again, all customizable. If you don't like it, don't enable it.
Or for half the price, build you own computer that supports secure booting Linux.
"by making it a requirement that the OEM provide the key to the buyer"
So a malware creator can use that key to create signed malware?
There is a reason we call it a "private" key. If you don't like it, make your own key and import it.
Fake News: In other news, OEM X hast lost millions of dollars because Linux runs over 50% of the internet and all those data centers can't use their servers. OEM Y saw this as an opportunity and now offers manageable certs.
What makes you think OEMs won't allow manageable certs when it's easy for them to do so? Actually, it's beyond "easy" as the default is to allow it from UEFI motherboard makers. OEMs would have to specifically request to have that feature disabled. Also sounds like a great way for a class action law suit or even anti-trust.
tin foil hat much?
Where are you getting this info that IT can't sign their own boot loaders?
"allow customers to import and manage those certificates, and manage secured boot"
If the OEM doesn't allow it, don't buy it. In a market where competition is fierce and Linux is quite "main stream" for enthusiasts/servers, there will be someone else willing to allow certs to be managed.
"What are the socioeconomic ramifications of the disease and how do they reflect upon any surviving children? How will social stigmas affect copulation opportunities? These and other feedback mechanisms cannot be marginalized by hyperbole."
When you're talking about natural-selection, quantity > quality, so long as the quality of the quantity is just enough to keep it running.
When you got countries with 30% and higher HIV rates, it's not an issue of having it or not, it's how often you have sex that determines if you have kids to keep your DNA alive.
Curious which SSD you had.
There are two types of "deleted" data. There is "remapped" deleted data caused by wear leveling, and "File System" deleted data.
In the case of a deleted file, the TRIM command would let the SSD know that it can clear that block, In the case of no TRIM, the SSD will still think data is there. If the block is cleared and the SSD writes to it, nothing has to be done. If TRIM has not happened and the block is not cleared, the SSD will read up that block, remap it, then write that data to another block.
In this case, TRIM saves a read. There are worse cases.
Assume you deleted a file and the filesystem deleted all 256k from a block. To the FS, that 256k is clear, but to the SSD, it isn't. Now assume you write 128k of data to the block. With TRIM, that 128k get written, no extra work to be done and the other free 128k can be written to with no extra work either. Without TRIM, the SSD copies up all 256k of that block, then modifies 128k of it, then writes all 256k to another block. If the other 128k get modifed, the SSD must again read up all 256k of data, modify the other 128k, then write to another block.
In this case, you save two reads and one write. You gained write performance while saving an extra write cycle.
The real question is how much overhead and complexity TRIM adds.
If it's a write cycle death, it should die gracefully in a read-only state. If it dies for defect reasons, then it's a up to luck.
"Byte-by-byte they're about the same price as the 15k SAS"
If that's the case, what about power costs and gained IO performance?
I'm curious from someone who works with them in servers.
"Does anybody have a backup plan for when their SSDs die? After all, unlike magnetic media, SSDs have a limited number of writes"
Most 128GB drives are good for 10TB+ per day for 5 years. Even when you hit the write cap, you can still read from the drive, it just turns read-only.
"Abstinence" is good to teach because it is sound advice, but *most* of the time, it doesn't work or just postpones it a bit.
If this was a perfect world, then no one would have to worry about wars/poverty/etc. Alas, people are human, and society has to pay for the mistakes of others.
Anyone who thinks the general population will remain abstinate, is naive(willful ignorance). Any policies based around these as core ideas will just cause more harm than good. "Abstinence" can augment other ideas, but it should never be a "core" idea, except for religious reasons.
"People having sex" is as much the reason for people having AIDs as people sports getting broken bones. It may contribute a lot, but isn't the only reason.
I'm also assuming, from your judgmental tone, that you think people "sleeping around" is the issue. Well, some times some people cheat on their spouses and the other spouse doesn't know. The person who is "faithful" still gets AIDs.
The difference is now there are a bunch of cancer cures on human testing.
If they didn't get over 800bil in subsidies over the past 15 years and didn't run their fiber over public land and other's private property, then you may have a case. But as it stands, they've gotten over 1 trillion(after inflation) of free money and they run their cables along high-ways and cross state lines. There is NOTHING private about that.
When you pay for "the internet", you're paying for full access to the "public network". If the company interferes with that, then they should not be able to advertise it as an "internet connection" in any form.
Also, your connection may flow "over" their private network, but you do not actually have access to that private network. Not to mention, their "private network" runs over public lands and crosses state lines.
If they don't like it, then they should have to fully purchase an own the land where their fiber/cables run.
A read-only bootloader is a horrible idea for a common desktop computer. IT would hate you.
"That this will block certain classes of rootkits is entirely incidental"
If everything from the boot process to the software is signed and requires signature, malware's only hope to getting installed is bugs in the OS, and even then, it would go away on reboot. Even if the malware could insert itself into some start-up location, it would never start because it wasn't signed. The uEFI won't load it, the bootloader won't load it, the OS won't load it.
I fail to see how this is bad.
Each department gets funding from the school, but also from private donations and other grants.
Yes, a Women's Outreach dept may have full staffing and proper equipment, but that's because other people think it it's important and donate.
If you think non-special group people need more help with money, then you should start donating.