I really beg to differ, in fact I think the Beechcraft Starship is one of the best looking planes ever designed. I had the luck of seeing one in person and they just look like what an airoplane should.
That's why triple monitor is even better, documentation/debugger in the left window, IDE in the middle, and application in the right monitor. It's extremely useful for me and I'm not even primarily a coder (network admin). I use 3x17" monitors, some people might find a pair of 22+" monitors more useful depending on their workflow and work style, either way it's dirt cheap.
Sorry but HTML/CSS/XML is VERY powerful and offers flexibility in both the way that things get displayed and in the way we tie systems together. Two quick examples, the AJAX platform Google uses for Google Maps has allowed them to write native clients for a whole host of mobile platforms at a fraction of what a tradition client/server system would have, the second is a system we developed for adding accounts payable automation to our systems. We essentially use a local proxy server that intercepts the html from our ERP system and ties it into the AP system and our content management systems. All of this is done without touching the code in the ERP system which would have been massively expensive, required a HUGE testing effort, and would have made the ERP system much harder to upgrade. This is how you use open, flexible systems to avoid the hidden costs so common in legacy platforms.
I use a system with 3x120mm fans (front, back, PSU) and 1x80mm. The 120mm fans are normally running at 8V (they are 12V units) and are absolutely silent. The CPU fan is the 80mm and unless I have both cores on my 45W Athlon x2 pegged it is also silent. The only time both cores are pegged is during transcoding so I just have it do that in the middle of the night.
The reason they don't allow modding is that whole "hot coffee" thing you might have heard about, you know the lawsuit against the company based on a third party mod that almost shut them down.
All business class printers worth buying do automatic duplexing. Anything lacking duplexing is a consumer toy or someone saving capital costs but wasting consumables and/or employee time.
Although NASA split off from the Airforce a long time ago I think the Museum of the United States Air Force at Wright Patterson would be a good home. They have plenty of room, a fairly serious space collection (bigger than the Smithsonian's I believe), and are a mecca for airplane and space buffs.
That's to get rid of all the nasties onboard, primarily the hydrazine from the thrusters I would imagine though the asbestos in the caulk between the heat tiles might need to go too depending on its mobility (some aircraft have it pulled, some have it encased in place if it's not an exposure hazard).
Sorry but the fit is pretty good on the shuttle,it has to be or that plasma problem Columbia had would be a lot more common.... Not to mention that whole keeping the atmosphere in against the vacuum of space thing.
Nope, not nearly enough ram for even a moderately complex scene's geometry let alone the textures unless you want your output looking like a game (IE most graphic artists will want photorealistic output which is more than a game console is capable of).
The best MIPS/watt for CUDA is probably either the 9600 GSO or the GTX280 depending on whether you're memory or processor constrained. The 9600 can be had for about $75 for 768MB variety (forget the 512/1024 parts they much lower performing) and has 96 stream processors running at up to 650Mhz. The GTX280 costs about $400 and has 240 650Mhz stream processors (though I believe they might be slightly more advanced then the ones on the 9600 I'm not sure how much of that is exposed by CUDA). Power usage is 46W peak for the 9600 and 180W for the GTX280.
Old CPU's have a much lower MIPS/Watt and a lower MIPS/interconnect so they have a higher cost. Many organizations have found it's cheaper to retire an old supercomputer and add a few nodes to the new one even if it is more capital outlay to get the same performance. Basically a Cell does many times as much useful work than a P4 at a fraction of the power budget.
I've seen multiple presentations by naturalists specifically calling out the fact that North America is significantly more forested than it was before European settlers came. One big factor is the wholesale clearing of much of the great plains.
Just color copies/prints cost my midsized employer $200K last year and that didn't include the cost of paper. In this economy that's a real target for cost savings. If you can save 20% that's about enough to employ one low level person or enough to give an extra 1% cost of living adjustment to a department. As long as the results are legible on a marginal printer is there any reason NOT to do it?
We simply don't have anywhere near the technology to harness this sort of heat into energy.
Like hell we don't, molten salt solar plants use salts that boil at 1400C and magma only reaches about 1300C max, the solidified area that would form around the pipe would lower the delta T to well below what such a system could handle.
The server was built summer of 2006, there were no servers with a dedicated TOE card (not checksum offloading but full stack offloading). Granted today all of HP's servers ship with them, but you must be running Windows with either the Advanced Networking Pack (doesn't run so well on x86, much better on x64) or SP2.
Actually the network CAN be the bottleneck even at low throughput. We had an Oracle server that saw an ~20% decrease in CPU load and significant reduction in latency by going with a TOE card, a $50K server (running several times that in software) sped up significantly with the addition of an $800 card, the best bang for the buck I have seen in over a decade in IT.
Modern HDD's can easily do 50MBps sustained for bulk transfer so even a 3 drive RAID5 with a CPU and network card that can keep up should be able to saturate a Gig-e network without a problem.
If you want decent throughput build it yourself. Seriously. I have a coworker that bought 5 different NAS devices to do a bakeoff for a small skunkworks office and they all sucked for throughput. We ended up buying a $1K NAS that still wasn't great but sure beat all the SOHO ones. Numbers were ~8MB/s max on the fastest SOHO unit vs 25MB/s on the midrange one.
Theoretically it's about risk assessment and market segmentation, allow the people making the loans set their acceptable level of risk and required rate of return. Obviously the system broke down, but that's because they came up with instruments so complex that they eliminated information from the system and thus hid the risk.
Some people do it to compete with others, some do it to compete with themselves and against the mountain and mother nature. If you know you wouldn't be able to do it without oxygen but you think you have a shot with it, it might be the toughest thing you ever do and so it might be worth doing. Besides that Everest isn't anywhere the most difficult so by your logic it's not worth doing, K2 is MUCH more difficult for instance. Personally I can't do 5km peaks very well (I'm a moderately out of shape flatlander) so I know I will never see the summit of an 8km peak.
Considering TIV-I made it through a EF2 tornado without any problems I would say he understood the forces fine, he never though the TIV could survive an EF5 which is the whole reason for teaming with the DOW.
Lol, the web interface could never keep up with all the stuff that the JAVA interface does. There is telnet/SSH as well as SNMP and web services interfaces to perform management tasks, but for most configuration as well as non-detailed monitoring the JAVA GUI is by far the best tool. I wouldn't use it to diagnose serious problems, but if I just want to update a zone or monitor the storage and E ports it's just plain easier to do through the GUI. Not only that but the SAN maps that the GUI provides are the fastest way to show your configuration to someone not familiar with your environment as I found out while working with two outside support organizations last month.
The 256GB monster is actually non-production, due to the quirks of the management in our software side they wanted to have "identical" testing capabilities vs prod. Since we are CPU licensed the cheapest way I could get them two non-prod environments to match prod was to match the prod CPU count and double the ram. This way we have room to go prod if we need to. Our DR scenario is currently to either move the prod HBA's into the non-prod server or zone the prod disks to the non-prod HBA's depending on the failure mode, the app is currently not covered by our DR SLA, but should that change then we will begin either log shipping to a like DR box or buy a second SAN for DR and use SAN replication. Oh, and to answer your questions directly:
1)Cost, Oracle RAC is expensive per transaction, it's more of an availability tool then a performance one.
2)Data transform tool and the fact that the best way we have found to maintain decent I/O performance without turning down Oracle's data integrity options is to throw more log writers at the problem, one I/O writer per core.
3)Like I said prod is only 128GB and since our OLTP DB is currently only about 60GB uncompressed I don't forsee us outgrowing a maxed box before the 3 year hardware cycle is out.
4)Currently our primary table is growing about 1.2M rows a month, but we are adding addition capabilities about twice a year so data growth is hard to quantify over a long period of time.
5)Our SLA is something like 95% during SLA hours, hardly hard to achieve with decent equipment. We recently experienced some of the worst downtime in my career due to prematurely outgrowing our old Cisco 9140's (they fell over at ~1.7Gb max traffic, very pathetic), but it was a total of about 4 hours of user visible downtime and even less for the financial systems.
6)DR is talked about above.
Other)Storage, we use a Xiotech SAN, we have 36TB of raw space over 224 spindles which is utilized for file storage, SQL Server, Lotus Notes, and multiple Oracle installations as well as for some boot from SAN application servers. Our next move will probably be to their Emprise 7000 line which will probably suck in all of the data in our current until as well as host document archiving for ediscovery. The Emprise is a beast of a system, scalable from 1TB to 1EB, the bigger limitations are the connected servers (248) and LUN's (1,024).
Dear god, the reason you didn't want to go over ~1.25GB of heap in Java x86 isn't that it can't go higher, it's that garbage collection times become significant enough that they cause user noticeable delay and possibly timeouts.
Slashdot Mirror
I really beg to differ, in fact I think the Beechcraft Starship is one of the best looking planes ever designed. I had the luck of seeing one in person and they just look like what an airoplane should.
That's why triple monitor is even better, documentation/debugger in the left window, IDE in the middle, and application in the right monitor. It's extremely useful for me and I'm not even primarily a coder (network admin). I use 3x17" monitors, some people might find a pair of 22+" monitors more useful depending on their workflow and work style, either way it's dirt cheap.
Sorry but HTML/CSS/XML is VERY powerful and offers flexibility in both the way that things get displayed and in the way we tie systems together. Two quick examples, the AJAX platform Google uses for Google Maps has allowed them to write native clients for a whole host of mobile platforms at a fraction of what a tradition client/server system would have, the second is a system we developed for adding accounts payable automation to our systems. We essentially use a local proxy server that intercepts the html from our ERP system and ties it into the AP system and our content management systems. All of this is done without touching the code in the ERP system which would have been massively expensive, required a HUGE testing effort, and would have made the ERP system much harder to upgrade. This is how you use open, flexible systems to avoid the hidden costs so common in legacy platforms.
I use a system with 3x120mm fans (front, back, PSU) and 1x80mm. The 120mm fans are normally running at 8V (they are 12V units) and are absolutely silent. The CPU fan is the 80mm and unless I have both cores on my 45W Athlon x2 pegged it is also silent. The only time both cores are pegged is during transcoding so I just have it do that in the middle of the night.
The reason they don't allow modding is that whole "hot coffee" thing you might have heard about, you know the lawsuit against the company based on a third party mod that almost shut them down.
All business class printers worth buying do automatic duplexing. Anything lacking duplexing is a consumer toy or someone saving capital costs but wasting consumables and/or employee time.
Although NASA split off from the Airforce a long time ago I think the Museum of the United States Air Force at Wright Patterson would be a good home. They have plenty of room, a fairly serious space collection (bigger than the Smithsonian's I believe), and are a mecca for airplane and space buffs.
That's to get rid of all the nasties onboard, primarily the hydrazine from the thrusters I would imagine though the asbestos in the caulk between the heat tiles might need to go too depending on its mobility (some aircraft have it pulled, some have it encased in place if it's not an exposure hazard).
Sorry but the fit is pretty good on the shuttle,it has to be or that plasma problem Columbia had would be a lot more common.... Not to mention that whole keeping the atmosphere in against the vacuum of space thing.
Nope, not nearly enough ram for even a moderately complex scene's geometry let alone the textures unless you want your output looking like a game (IE most graphic artists will want photorealistic output which is more than a game console is capable of).
The best MIPS/watt for CUDA is probably either the 9600 GSO or the GTX280 depending on whether you're memory or processor constrained. The 9600 can be had for about $75 for 768MB variety (forget the 512/1024 parts they much lower performing) and has 96 stream processors running at up to 650Mhz. The GTX280 costs about $400 and has 240 650Mhz stream processors (though I believe they might be slightly more advanced then the ones on the 9600 I'm not sure how much of that is exposed by CUDA). Power usage is 46W peak for the 9600 and 180W for the GTX280.
Old CPU's have a much lower MIPS/Watt and a lower MIPS/interconnect so they have a higher cost. Many organizations have found it's cheaper to retire an old supercomputer and add a few nodes to the new one even if it is more capital outlay to get the same performance. Basically a Cell does many times as much useful work than a P4 at a fraction of the power budget.
I've seen multiple presentations by naturalists specifically calling out the fact that North America is significantly more forested than it was before European settlers came. One big factor is the wholesale clearing of much of the great plains.
Just color copies/prints cost my midsized employer $200K last year and that didn't include the cost of paper. In this economy that's a real target for cost savings. If you can save 20% that's about enough to employ one low level person or enough to give an extra 1% cost of living adjustment to a department. As long as the results are legible on a marginal printer is there any reason NOT to do it?
We simply don't have anywhere near the technology to harness this sort of heat into energy.
Like hell we don't, molten salt solar plants use salts that boil at 1400C and magma only reaches about 1300C max, the solidified area that would form around the pipe would lower the delta T to well below what such a system could handle.
The server was built summer of 2006, there were no servers with a dedicated TOE card (not checksum offloading but full stack offloading). Granted today all of HP's servers ship with them, but you must be running Windows with either the Advanced Networking Pack (doesn't run so well on x86, much better on x64) or SP2.
Actually the network CAN be the bottleneck even at low throughput. We had an Oracle server that saw an ~20% decrease in CPU load and significant reduction in latency by going with a TOE card, a $50K server (running several times that in software) sped up significantly with the addition of an $800 card, the best bang for the buck I have seen in over a decade in IT.
Modern HDD's can easily do 50MBps sustained for bulk transfer so even a 3 drive RAID5 with a CPU and network card that can keep up should be able to saturate a Gig-e network without a problem.
If you want decent throughput build it yourself. Seriously. I have a coworker that bought 5 different NAS devices to do a bakeoff for a small skunkworks office and they all sucked for throughput. We ended up buying a $1K NAS that still wasn't great but sure beat all the SOHO ones. Numbers were ~8MB/s max on the fastest SOHO unit vs 25MB/s on the midrange one.
Theoretically it's about risk assessment and market segmentation, allow the people making the loans set their acceptable level of risk and required rate of return. Obviously the system broke down, but that's because they came up with instruments so complex that they eliminated information from the system and thus hid the risk.
Some people do it to compete with others, some do it to compete with themselves and against the mountain and mother nature. If you know you wouldn't be able to do it without oxygen but you think you have a shot with it, it might be the toughest thing you ever do and so it might be worth doing. Besides that Everest isn't anywhere the most difficult so by your logic it's not worth doing, K2 is MUCH more difficult for instance. Personally I can't do 5km peaks very well (I'm a moderately out of shape flatlander) so I know I will never see the summit of an 8km peak.
Considering TIV-I made it through a EF2 tornado without any problems I would say he understood the forces fine, he never though the TIV could survive an EF5 which is the whole reason for teaming with the DOW.
Lol, the web interface could never keep up with all the stuff that the JAVA interface does. There is telnet/SSH as well as SNMP and web services interfaces to perform management tasks, but for most configuration as well as non-detailed monitoring the JAVA GUI is by far the best tool. I wouldn't use it to diagnose serious problems, but if I just want to update a zone or monitor the storage and E ports it's just plain easier to do through the GUI. Not only that but the SAN maps that the GUI provides are the fastest way to show your configuration to someone not familiar with your environment as I found out while working with two outside support organizations last month.
The 256GB monster is actually non-production, due to the quirks of the management in our software side they wanted to have "identical" testing capabilities vs prod. Since we are CPU licensed the cheapest way I could get them two non-prod environments to match prod was to match the prod CPU count and double the ram. This way we have room to go prod if we need to. Our DR scenario is currently to either move the prod HBA's into the non-prod server or zone the prod disks to the non-prod HBA's depending on the failure mode, the app is currently not covered by our DR SLA, but should that change then we will begin either log shipping to a like DR box or buy a second SAN for DR and use SAN replication. Oh, and to answer your questions directly:
1)Cost, Oracle RAC is expensive per transaction, it's more of an availability tool then a performance one.
2)Data transform tool and the fact that the best way we have found to maintain decent I/O performance without turning down Oracle's data integrity options is to throw more log writers at the problem, one I/O writer per core.
3)Like I said prod is only 128GB and since our OLTP DB is currently only about 60GB uncompressed I don't forsee us outgrowing a maxed box before the 3 year hardware cycle is out.
4)Currently our primary table is growing about 1.2M rows a month, but we are adding addition capabilities about twice a year so data growth is hard to quantify over a long period of time.
5)Our SLA is something like 95% during SLA hours, hardly hard to achieve with decent equipment. We recently experienced some of the worst downtime in my career due to prematurely outgrowing our old Cisco 9140's (they fell over at ~1.7Gb max traffic, very pathetic), but it was a total of about 4 hours of user visible downtime and even less for the financial systems.
6)DR is talked about above.
Other)Storage, we use a Xiotech SAN, we have 36TB of raw space over 224 spindles which is utilized for file storage, SQL Server, Lotus Notes, and multiple Oracle installations as well as for some boot from SAN application servers. Our next move will probably be to their Emprise 7000 line which will probably suck in all of the data in our current until as well as host document archiving for ediscovery. The Emprise is a beast of a system, scalable from 1TB to 1EB, the bigger limitations are the connected servers (248) and LUN's (1,024).
Dear god, the reason you didn't want to go over ~1.25GB of heap in Java x86 isn't that it can't go higher, it's that garbage collection times become significant enough that they cause user noticeable delay and possibly timeouts.