Actually, I think it's like this: Until a few months ago, the generaly available diesel fuel (called Low Sulphur Dielse) in the USA had up to 500 parts per million of sulphur. With this fuel and the technology used in the modern diesels (as sold in Europe), it's still not possible to make a competitive engine while keeping within USA's emission regulations. Thus, the few available diesel models are seriously limited in terms of engine choices and power.
A few months ago, Ultra Low Sulphur Diesel with less than 15 ppm was made mandatory in the USA. With this fuel and the existing technology, it's possible to produce competitive engines while still respecting the emission regulations. Thus, expect to see more diesel models next year.
The exception to this is California and a few more states. They have passed even stricter emission regulations which can't be met even with ULSD and current technology. IIRC, both the remaining USA states and the European Union are also scheduled to pass stricter emission regulations. It'll will take new technology to meet those emission levels, which should be comming to market in 2008-2009.
I'm afraid that the status quo in consumer electronics is much worse than that. Many, problably most, devices consume a few watts to a few tens of watts in standby mode.
1 watt or less in standby for most devices is what the International Energy Agency is pushing for in 2010.
I'm afraid it's still an issue. Despite the technology being available and cheap, mass consumer electronics does tend to cut corners in ways that are hard to understand.
Without copyright, I could take a computer program whose source code is available, introduce some improvements and then distribute only the binaries while I make a business of supporting my improved version of the program.
GPL *forces* me to give the source code to whomever I give the binaries.
No. Actually, it's very unlikely Microsoft, Sony or Nintendo depend on IBM for manufacturing. Most likely, all the three companies can take their manufacturing somewhere else if they choose to. That was the BIG mistake Microsoft did with Xbox: it depended on Intel and nVidia to manufacture the CPU and GPU. This kept the Xbox expensive to manufacture.
What IBM has going for them is that, like Intel and AMD, they have the know-how to design high-performance CPUs and they have their own fabs. And, unlike Intel and AMD, they're willing to design a CPU that can be manufactured by someone else (or manufacture a CPU designed by someone else for that matter).
Some years ago, a friend of mine had his PC pluged into a power socket without ground connection. Ocasionally, the PC would freeze. Connecting the ground pin on the power plug to his house's floor with a wire would unfreeze the computer.
"Coca-cola" is the trademark, not "coke". Both Kleenex and Xerox take measures similar to the ones Google has just engaged to ensure their trademarks remain as such.
Until now, Xeons supported 2 or 4 sockets in the same FSB, hence the lower speeds than the single socket versions. With the 5000 series, they droped that and went with one socket per FSB. Thus, these Xeons' FSB can match or surpass (since the boards can be a bit more expensive) the desktop version. 5100 series Xeons with 1333MHz FSB are avaible for buyers NOW.
Re:Sounds like a good idea.
on
EXT4 Is Coming
·
· Score: 3, Interesting
There are or were a few quirks.
First off the bat: you can't install the bootloader in a XFS partition since XFS uses the first 512 byte block on the partition. Of course, most people install the bootloader in the MBR but for some it's an issue.
GRUB had a bug with XFS. When you tried to use a XFS partition as/boot, you could corrupt XFS.
For a considerable period of time, ext3's code was more stable than XFS.
ext3 has an ordered data mode (which is the default). Other journaled file systems only support writeback mode. In general, ordered data mode doesn't provide any better warranty of consistency than writeback mode but does make an important difference for a few special cases but which can make a substancial difference to a desktop user.
Typical annoying case: - You're editing a file on your favorite text editor and you save it. - The editor opens the file in overwrite mode, meaning the file is actually deleted and a new one is created (under Linux's default settings, the OS will commit the changes to the metadata in 5 seconds or less and the changes to the data in 30 seconds or less). - The changes to the metadata are commited to disk. - The system crashes! When the system comes back up, the new file is there it's full of garbage.
With ext3's ordered data mode, the contents of the file would have been commited to disk before the associated changes to metadata. It's problable (but not assured!!) that after a crash you'll have either the old version or the new version of the file.
HD-DVD and BluRay drives are new devices, thus it's natural they're expensive. Think back at when DVD-ROM drives for PC came out. I wouldn't be surprised if the drive itself costs 100-200 USD.
The decoding hardware is also expensive. Both HD-DVD and BluRay mandates H.264 support. At 1024p, it takes a medium range PC to handle it. Toshiba's HD-DVD player actually uses PC hardware (1.8GHz P-M, 512MB RAM IIRC), running Linux.
Or the couple of added execution units. Or the 1 MB L2 instruction cache. Or just saving sillicon area.:) They didn't remove it. They simply didn't implement it in the new chip. The IA32 compatibility required logic that not only took area, also sat in the critical paths. So, not implementing it simplified things a bit. Maybe a big bit.
It's not like Intel just popped out the IA32 circuit and droped some new stuff in that spot. Although it's microarchitecture is an evolution of the current Itanium2, the upcoming is a brand new chip, where Intel chose not to implemente the IA32 logic.
I don't know where the quote stops and your words start, so take as needed.
I agree that parallel hybrids (like Honda Insight) are not the way to go, while serial hybrids are much more interesting.
However, Toyota Prius is neither a parallel nor a serial hybrid. Toyota managed to get the best of both worlds in one design. Like your text suggests, it does keep the gasoline engine running in a narrow range of rpm for maximum efficiency. But unlike a pure serial hybrid, it outputs directly into the transmission, therefore skipping the power loss in mechanical -> electrical -> mechanical conversion in pure serial hybrids.
You can also use CVTs to keep the gasoline engine running at a constant speed. Honda Insight for example is available with CVT (Insigh always uses the gasoline engine, while the electric motor is only used for extra power).
The limitations for CVT are that you'd still need a clutch to start movement, with the associated power dissipation. Also, I don't nobody ever build a car with CVT and more than 200HP.
Now, back to the Prius, it uses a smart solution. Problably the reason Toyota is being sued over. They have a planetary system that connects the electric motor, the gasoline engine, the generator and the transmission all together. The system allows them to keep the gasoline engine runnig at a whatever speed they want, weather it's proving power to the transmission and/or to the generator. Unlike Insight, in Prius it's the electric motor that provides power at zero-30MPH, so they don't need a clutch either to start movement.
I just remembered something. What I've described above is described in RFC 3056 and is supported, for example, by Linux's tun6to4 network device. However, you may be configuring a "6to4" tunnel and always route your non-local packets to a 3rd party 6 to 4 router. For example, I've seen howto's that use the sit0 device. The sit0 device doesn't know about RFC 3056 and will always send packets to the 3rd party router.
I'm aware of mapped and 6to4 tunneling, but those don't do exactly what I have specified, and that's why they haven't seen massive use. 6to4 tunneling requires a 3rd party and doesn't necessarily take care of the routing for you. Consider, at home I have a network NATed behind a single static IP on DSL. At work, I have a workstation with a static v4 address. I wish to talk to my home machine at (for the sake of argument) 192.168.1.200.
Certainly, I can play various games with ipip tunnels and creative routing. OR I could aggregate my home static IP, and the non-routed IP of the target machine into a v6 address and just use it. The traffic will not need to go through 192.88.99.1 at all. It won't matter if my work machine is also behind NAT. The networks on either end can be either v4 or v6 native. If v6, it's easy to make machines selectively available by having (or not having) the appropriate alias. I can also filter the traffic at the firewall as usual (using choke routing rather than NAT).
6to4 automatic tunneling should do just that.
At work, configure 6to4 on your workstation (I assume it has a public IP address). At home, configure 6to4 on your gateway, based on it's public IP address. Assign your other home computers IPv6 addresses with the same 2002:some:thing: prefix the gateway got. radvd supports this.
There, your other home computers now have a public IPv6 address and should be reachable from the IPv6 Internet.
The packets won't be going through a 3rd IPv6 IPv4 router either. 3rd party is only needed to connect hosts on 6to4 tunneling to native IPv6 hosts or IPv6 hosts on other kind of tunnel.
Your workstation at work and gateway at home will route exchange their IPv6 in IPv4 packets directly between themselves.
I'm aware of mapped and 6to4 tunneling, but those don't do exactly what I have specified, and that's why they haven't seen massive use. 6to4 tunneling requires a 3rd party and doesn't necessarily take care of the routing for you. Consider, at home I have a network NATed behind a single static IP on DSL. At work, I have a workstation with a static v4 address. I wish to talk to my home machine at (for the sake of argument) 192.168.1.200.
Certainly, I can play various games with ipip tunnels and creative routing. OR I could aggregate my home static IP, and the non-routed IP of the target machine into a v6 address and just use it. The traffic will not need to go through 192.88.99.1 at all. It won't matter if my work machine is also behind NAT. The networks on either end can be either v4 or v6 native. If v6, it's easy to make machines selectively available by having (or not having) the appropriate alias. I can also filter the traffic at the firewall as usual (using choke routing rather than NAT).
6to4 automatic tunneling should do just that.
At work, configure 6to4 on your workstation (I assume it has a public IP address).
At home, configure 6to4 on your gateway, based on it's public IP address. Assign your other home computers IPv6 addresses with the same 2002:some:thing: prefix the gateway got. radvd supports this.
There, your other home computers now have a public IPv6 address and should be reachable from the IPv6 Internet.
The packets won't be going through a 3rd IPv6 IPv4 router either. 3rd party is only needed to connect hosts on 6to4 tunneling to native IPv6 hosts or IPv6 hosts on other kind of tunnel.
Your workstation at work and gateway at home will route exchange their IPv6 in IPv4 packets directly between themselves.
Simple: they're not shared memory systems. XD1 is built on 2 socket smp blocks, with Opteron 2XX series. XT3 (RedStorm) is built on single socket blocks, with Opteron 1XX series.
Linpack is not very interconnect sensitive. Despite all the bragging companies make on being number 1, the top500.org are the first to recognize that Linpack is not a proper benchmark, just a quick way to compile a list of the biggest machines on the planet.
But FLOPS/CPU numbers come primarly from cpu design, clock speed and memory bandwidth. For example, among the scalar CPUs, some can issue 2 double precision floating point ops per clock cycle, some can issue 4. And they run at different frequencies, from 700 MHz to 2.2GHz+. The vector based machines can perform a lot of operations per cycle, but they tend to run at low clock frequencies. For example, the Earth Simulator's processor (based on NEC SX-6) can perform 16 but it runs at 500 MHz.
The ideia behind isn't new. IBM's very successfull AS/400 line used the same concept. However, I think that now is a good time to take another look at the concept, using the improvements in JIT and applying OOP concepts to the OS.
You are wrong.:) PostgreSQL has a propper query planner (MS SQL's may be better or worse, I have no idea), support for prepared statements and will "cache" the execution plans for statements in Pl/SQL stored procedures. I don't know about Firebird, but I'd expect it to have about the same features. These are fairly common features in RDBMS. Not having them is the exception.
Replication is avaliable for MySQL, PostgreSQL (and Firebird too I think), although PostgreSQL's replication are still third party tools and I don't know which replication modes are avaliable to each RDBMS.
Regarding data integrity, what flaws would you point to PostgreSQL or Firebird?
I have no idea about what you mean by "remote queries".
Slashdot Mirror
Actually, I think it's like this:
Until a few months ago, the generaly available diesel fuel (called Low Sulphur Dielse) in the USA had up to 500 parts per million of sulphur. With this fuel and the technology used in the modern diesels (as sold in Europe), it's still not possible to make a competitive engine while keeping within USA's emission regulations. Thus, the few available diesel models are seriously limited in terms of engine choices and power.
A few months ago, Ultra Low Sulphur Diesel with less than 15 ppm was made mandatory in the USA. With this fuel and the existing technology, it's possible to produce competitive engines while still respecting the emission regulations. Thus, expect to see more diesel models next year.
The exception to this is California and a few more states. They have passed even stricter emission regulations which can't be met even with ULSD and current technology.
IIRC, both the remaining USA states and the European Union are also scheduled to pass stricter emission regulations.
It'll will take new technology to meet those emission levels, which should be comming to market in 2008-2009.
Is this uncommon in the USA?
Ok, you've just shed a new light on the energy saving situation in the USA.
I'm afraid that the status quo in consumer electronics is much worse than that.
Many, problably most, devices consume a few watts to a few tens of watts in standby mode.
1 watt or less in standby for most devices is what the International Energy Agency is pushing for in 2010.
I'm afraid it's still an issue. Despite the technology being available and cheap, mass consumer electronics does tend to cut corners in ways that are hard to understand.
e rtest.htm
This is old, but it's a good example how bad it can be:
http://home.comcast.net/~machrone/playertest/play
Wrong.
Without copyright, I could take a computer program whose source code is available, introduce some improvements and then distribute only the binaries while I make a business of supporting my improved version of the program.
GPL *forces* me to give the source code to whomever I give the binaries.
These drives will be available with PATA and SATA interfaces.T 13_Presntation.pdf
A new revision of the standards is being worked upon to support hybrid drives.
Take a look at http://www.t13.org/docs2005/e05113r0-Hybrid_Disk_
No, we drink alchool. ;-)
No.
Actually, it's very unlikely Microsoft, Sony or Nintendo depend on IBM for manufacturing. Most likely, all the three companies can take their manufacturing somewhere else if they choose to.
That was the BIG mistake Microsoft did with Xbox: it depended on Intel and nVidia to manufacture the CPU and GPU. This kept the Xbox expensive to manufacture.
What IBM has going for them is that, like Intel and AMD, they have the know-how to design high-performance CPUs and they have their own fabs.
And, unlike Intel and AMD, they're willing to design a CPU that can be manufactured by someone else (or manufacture a CPU designed by someone else for that matter).
Some years ago, a friend of mine had his PC pluged into a power socket without ground connection. Ocasionally, the PC would freeze.
Connecting the ground pin on the power plug to his house's floor with a wire would unfreeze the computer.
"Coca-cola" is the trademark, not "coke".
Both Kleenex and Xerox take measures similar to the ones Google has just engaged to ensure their trademarks remain as such.
Until now, Xeons supported 2 or 4 sockets in the same FSB, hence the lower speeds than the single socket versions.
With the 5000 series, they droped that and went with one socket per FSB. Thus, these Xeons' FSB can match or surpass (since the boards can be a bit more expensive) the desktop version.
5100 series Xeons with 1333MHz FSB are avaible for buyers NOW.
There are or were a few quirks.
/boot, you could corrupt XFS.
First off the bat: you can't install the bootloader in a XFS partition since XFS uses the first 512 byte block on the partition. Of course, most people install the bootloader in the MBR but for some it's an issue.
GRUB had a bug with XFS. When you tried to use a XFS partition as
For a considerable period of time, ext3's code was more stable than XFS.
ext3 has an ordered data mode (which is the default). Other journaled file systems only support writeback mode. In general, ordered data mode doesn't provide any better warranty of consistency than writeback mode but does make an important difference for a few special cases but which can make a substancial difference to a desktop user.
Typical annoying case:
- You're editing a file on your favorite text editor and you save it.
- The editor opens the file in overwrite mode, meaning the file is actually deleted and a new one is created (under Linux's default settings, the OS will commit the changes to the metadata in 5 seconds or less and the changes to the data in 30 seconds or less).
- The changes to the metadata are commited to disk.
- The system crashes!
When the system comes back up, the new file is there it's full of garbage.
With ext3's ordered data mode, the contents of the file would have been commited to disk before the associated changes to metadata. It's problable (but not assured!!) that after a crash you'll have either the old version or the new version of the file.
All of the above.
:)
HD-DVD and BluRay drives are new devices, thus it's natural they're expensive. Think back at when DVD-ROM drives for PC came out. I wouldn't be surprised if the drive itself costs 100-200 USD.
The decoding hardware is also expensive. Both HD-DVD and BluRay mandates H.264 support. At 1024p, it takes a medium range PC to handle it.
Toshiba's HD-DVD player actually uses PC hardware (1.8GHz P-M, 512MB RAM IIRC), running Linux.
And then, there's profit.
Or the couple of added execution units. :)
Or the 1 MB L2 instruction cache.
Or just saving sillicon area.
They didn't remove it. They simply didn't implement it in the new chip.
The IA32 compatibility required logic that not only took area, also sat in the critical paths. So, not implementing it simplified things a bit. Maybe a big bit.
It's not like Intel just popped out the IA32 circuit and droped some new stuff in that spot.
Although it's microarchitecture is an evolution of the current Itanium2, the upcoming is a brand new chip, where Intel chose not to implemente the IA32 logic.
I don't know where the quote stops and your words start, so take as needed.
I agree that parallel hybrids (like Honda Insight) are not the way to go, while serial hybrids are much more interesting.
However, Toyota Prius is neither a parallel nor a serial hybrid. Toyota managed to get the best of both worlds in one design. Like your text suggests, it does keep the gasoline engine running in a narrow range of rpm for maximum efficiency. But unlike a pure serial hybrid, it outputs directly into the transmission, therefore skipping the power loss in mechanical -> electrical -> mechanical conversion in pure serial hybrids.
You can also use CVTs to keep the gasoline engine running at a constant speed. Honda Insight for example is available with CVT (Insigh always uses the gasoline engine, while the electric motor is only used for extra power).
The limitations for CVT are that you'd still need a clutch to start movement, with the associated power dissipation. Also, I don't nobody ever build a car with CVT and more than 200HP.
Now, back to the Prius, it uses a smart solution. Problably the reason Toyota is being sued over.
They have a planetary system that connects the electric motor, the gasoline engine, the generator and the transmission all together.
The system allows them to keep the gasoline engine runnig at a whatever speed they want, weather it's proving power to the transmission and/or to the generator.
Unlike Insight, in Prius it's the electric motor that provides power at zero-30MPH, so they don't need a clutch either to start movement.
High mechanical geek factor, I'd say.
I just remembered something.
What I've described above is described in RFC 3056 and is supported, for example, by Linux's tun6to4 network device.
However, you may be configuring a "6to4" tunnel and always route your non-local packets to a 3rd party 6 to 4 router. For example, I've seen howto's that use the sit0 device. The sit0 device doesn't know about RFC 3056 and will always send packets to the 3rd party router.
6to4 automatic tunneling should do just that.
At work, configure 6to4 on your workstation (I assume it has a public IP address).
At home, configure 6to4 on your gateway, based on it's public IP address. Assign your other home computers IPv6 addresses with the same 2002:some:thing: prefix the gateway got. radvd supports this.
There, your other home computers now have a public IPv6 address and should be reachable from the IPv6 Internet.
The packets won't be going through a 3rd IPv6 IPv4 router either. 3rd party is only needed to connect hosts on 6to4 tunneling to native IPv6 hosts or IPv6 hosts on other kind of tunnel.
Your workstation at work and gateway at home will route exchange their IPv6 in IPv4 packets directly between themselves.
Simple: they're not shared memory systems.
XD1 is built on 2 socket smp blocks, with Opteron 2XX series.
XT3 (RedStorm) is built on single socket blocks, with Opteron 1XX series.
Linpack is not very interconnect sensitive. Despite all the bragging companies make on being number 1, the top500.org are the first to recognize that Linpack is not a proper benchmark, just a quick way to compile a list of the biggest machines on the planet.
But FLOPS/CPU numbers come primarly from cpu design, clock speed and memory bandwidth.
For example, among the scalar CPUs, some can issue 2 double precision floating point ops per clock cycle, some can issue 4. And they run at different frequencies, from 700 MHz to 2.2GHz+.
The vector based machines can perform a lot of operations per cycle, but they tend to run at low clock frequencies.
For example, the Earth Simulator's processor (based on NEC SX-6) can perform 16 but it runs at 500 MHz.
The ideia behind isn't new. IBM's very successfull AS/400 line used the same concept.
However, I think that now is a good time to take another look at the concept, using the improvements in JIT and applying OOP concepts to the OS.
SGI for example has done just that in some configurations.p uter/2009-1039_3-5428431-6.html
Check http://news.com.com/Photos+SGIs+Columbia+supercom
You are wrong. :)
PostgreSQL has a propper query planner (MS SQL's may be better or worse, I have no idea), support for prepared statements and will "cache" the execution plans for statements in Pl/SQL stored procedures.
I don't know about Firebird, but I'd expect it to have about the same features. These are fairly common features in RDBMS. Not having them is the exception.
Replication is avaliable for MySQL, PostgreSQL (and Firebird too I think), although PostgreSQL's replication are still third party tools and I don't know which replication modes are avaliable to each RDBMS.
Regarding data integrity, what flaws would you point to PostgreSQL or Firebird?
I have no idea about what you mean by "remote queries".