Also I am of the camp that says computers should be off when not in use. This statement obviously excludes servers. I just believe that turning it off increases the lifetime of the unit, and leaving it on is an unnecessary power consumption. The equivalent (albeit crude due to Energy Saver technology) is leaving your car running in the garage all the time. Being a tech, I see too many power adapters, power supplies, and hard drives fail.
Why would putting systems to sleep rather than shutting down cause hard drives to fail faster? Don't the drives spin down in either case? Are you sure you're a tech? As for power consumption, even a sleeping Mac Pro draws only about 5 watts. The various wall warts around your house almost certainly draw more collectively. Sleep it for a full day and you'll eat about the same power as a 100 watt light bulb in one hour. Macs don't draw 0 watts even when shut down, either. They're always drawing at least a watt or two since they don't have hard power switches, so the power supply is always active, unless you actually unplug it or switch it off with a power strip. For some of us, the increased productivity of being able to get back to work instantly with exactly the files and applications we've been working on still open is worth a few watts. And what about all the drive activity required to cold boot a Mac? Wouldn't that reduce HD lifetime?
Why would putting systems to sleep rather than shutting down cause hard drives to fail faster? Don't the drives spin down in either case? Are you sure you're a tech? As for power consumption, even a sleeping Mac Pro draws only about 5 watts. The various wall warts around your house almost certainly draw more collectively. Sleep it for a full day and you'll eat about the same power as a 100 watt light bulb in one hour. Macs don't draw 0 watts even when shut down, either. They're always drawing at least a watt or two since they don't have hard power switches, so the power supply is always active, unless you actually unplug it or switch it off with a power strip. For some of us, the increased productivity of being able to get back to work instantly with exactly the files and applications we've been working on still open is worth a few watts. And what about all the drive activity required to cold boot a Mac? Wouldn't that reduce HD lifetime?
Google has over 100,000 PATA and SATA drives in their storage cluster.
They recently released results on a study on hard drive reliability.
Their conclusion was that drives that spin up and spin down will fail faster than drives that don't.
They also found that heat had almost no effect on reliability.
They also found that drives are most likely to fail in the first 6 months.
Failure rates then drop down for a period of 18 months.
Once the drive is two years old the failure rate begins to increase at an appreciable rate.
And think of the heat and noise issues or lack thereof, especially when you have 4 drives in a tower like I do! This rocks, can't wait for larger sizes and price drops.
It's a few years too early for that. Flash storage is still about 10x more per GB than desktop hard drives. For me, most hard drives are quiet enough. If you have a drive that's making a little bit too much noise, usually replacing it with another hard drive will get you a quieter drive. Heat doesn't seem to be a problem.
Google has over 100,000 PATA and SATA drives in their storage cluster.
They recently released results on a study on hard drive reliability.
Their conclusion was that drives that spin up and spin down will fail faster than drives that don't.
They also found that heat had almost no effect on reliability.
They also found that drives are most likely to fail in the first 6 months.
Failure rates then drop down for a period of 18 months.
Once the drive is two years old the failure rate begins to increase at an appreciable rate.
Some of that is a bit of an oversimplification. Reading one of their charts (fig 2), it looks like the failure rates in the second year exceeds the failure rates of the entire first year, adding up the rates for the first three time periods (all first year) still doesn't exceed the 2yr point. Other charts in the same paper say things a bit differently, depending on use and, and less on temperature. I haven't read the paper in a bit and only reskimmed it today, so I may be a bit off-base on it.
I was just doing some looking, but it would be possible to purchase the 32GB flash drive, buy an external 2.5" sata house, and then plug it in via esata, firewire800. My question then is, can it still sustain its fast read speeds over a esata, firewire800? My initial thought would be yes, as both of those have enough bandwidth (firewire800 can support up to about 95MBps). So these could be used as very fast external drives (out the door for about $450 including housing).
I know personally, I don't shut down my laptop ever. So I could case less about the "increased startup time". But having a nice fast external drive would be nice.
That sandisk drive is cool, I'm interested in watching these things grow, with the rate of technology these days that capacity could triple in a year. When it triples is when I'll heavily consider it.
NAND flash has a finite number of read/write cycles, something in the order of 100,000. For storing digital photos, this is entirely adequate, as the memory will unlikely to be refreshed more than once or twice per day.
However, most modern operating systems create a paging file, which is frequently written to, erased, written, in a much more frequent cycle.
Replacing a computer's boot drive with a flash unit, WITHOUT OS MODIFICATIONS TO TAKE ACCOUNT OF THE R/W LIMITATION, is likely to prematurely age the flash drive.
^^^ I am pretty sure Intel took care of that flash memory r/w thingy everyone wants to bring up.
Check out the article again. Intel has made a big stride here in progress.
Quote:
Originally Posted by AppleInsider
Intel says the Z-U130s will distinguish themselves from other solid state product offerings by their extensive validation, including more than 1,000 hours of accelerated reliability testing. The drives are expected to meet an average mean time between failure (MTBF) specification of five million hours.
Why would putting systems to sleep rather than shutting down cause hard drives to fail faster? Don't the drives spin down in either case? Are you sure you're a tech?
I think that last comment is a bit uncalled for.
However, I will address the previous sentences anyway.
It is ludicrous to assume that any piece of technology that remains on 100% the time is going to be just as reliable as the equivalent device that is powered up 25% of the time or less depending on usage, or have as long of a lifetime. Your laptop in precious "sleep mode" still draws electricity, still generates heat, and heat has effects over time. I will listen to your argument when electronic equipment generates zero heat, has zero chance of condensation damage, has zero change of being susceptible to other environmental factors like accidental dropping, water penetration, dust...the list goes on. And spin down is not the same as power down.
Only when sleeping laptops use no power, generate no heat, and have no moving parts that are susceptible to weardown will I tell people that it's ok to leave them on all the time. Until then, I will continue to recommend to shut down computers when not in use. Thankfully, I work with respectable people who ask for my humble recommendation and don't go around asking if I'm really a tech when I give it. They can then choose to follow it or do whatever they want.
What size is your average OS X install these days?
1.5GB+?
It's more than just the base OS. As shipped, before I did anything with it, my MacBook Pro had 30GB already taken. If you don't use iLife, you can probably clean out 10+GB there if you uninstall it, but that means gutting some of Apple's selling points of having that software included. The iWork apps take 2GB (included as demos), probably more with the sample themes that are in the library somewhere. 32GB really isn't much to work with when it comes to Macs.
^^^ I am pretty sure Intel took care of that flash memory r/w thingy everyone wants to bring up.
Check out the article again. Intel has made a big stride here in progress.
I really don't believe they've really addressed it. The press release and product brief don't discuss the write cycle limitation of flash memory. Intel did demo PRAM recently, which I think does take care of it, but that's not flash, but they have not mentioned it in the product information.
Only when sleeping laptops use no power, generate no heat, and have no moving parts that are susceptible to weardown will I tell people that it's ok to leave them on all the time.
I don't think it is necessarily clear cut. Even parts that don't look like they are moving do have to deal with thermal expansion on use, and contraction on cool down. This fatigues the metals, joints and contacts in a circuit, and something may eventually fail. Hard drive motors are rated for a certain number of start cycles because they have to expend a lot more energy per unit time spinning up than they do maintaining the spin through inertia. Notebook hard drives are rated for a lot more start/stop cycles because drives are supposed at every reasonable opportunity to save power.
I don't know how much of a problem start/stops are, it can happen, but I really don't think it's a major problem so long as one isn't stupid about shutting down too often. I figure once a day is probably fine. If you want to keep your workspace, hybernation will do the trick.
It is ludicrous to assume that any piece of technology that remains on 100% the time is going to be just as reliable as the equivalent device that is powered up 25% of the time or less depending on usage, or have as long of a lifetime. Your laptop in precious "sleep mode" still draws electricity, still generates heat, and heat has effects over time. I will listen to your argument when electronic equipment generates zero heat, has zero chance of condensation damage, has zero change of being susceptible to other environmental factors like accidental dropping, water penetration, dust...the list goes on. And spin down is not the same as power down.
Then what is power down? Do you really think motherboards keep power flowing to hard drives for no reason whatsoever? You must really think electronic engineers are fools. Sleep generates internal heat? How much heat? Less than 10ºF on any given day. Most of the heat will be in the transformer in the power adapter, and transformers don't wear out. You'll have more temperature variation between day and night or when taking the laptop outside. All your "concerns" about condensation, dust, water, dropping, etc. would seem to be better arguments against taking any laptop off its desk. Is that what you're advocating?
Quote:
Only when sleeping laptops use no power, generate no heat, and have no moving parts that are susceptible to weardown will I tell people that it's ok to leave them on all the time. Until then, I will continue to recommend to shut down computers when not in use. Thankfully, I work with respectable people who ask for my humble recommendation and don't go around asking if I'm really a tech when I give it. They can then choose to follow it or do whatever they want.
Ah, so only "respectable" people agree with you. Maybe we're just not as susceptible to being impressed by anyone who comes along using self-invented "tech" terms like "weardown." If you can cite any REAL information other than your "learned personal opinion," I'd be happy to accept it. Otherwise, I'll treat it no differently than any other old wives' tales and urban legends, all of which have staunch supporters who swear they're indisputable.
Besides, I don't hold "techs" in very high regard, having dealt with way too many ignorant ones in the past. In fact, one of my closest friends is not only a technician, he's the system administrator for a medium-size newspaper. And even he admits he would never have made it past the first week in an engineering college.
I'm willing to bet that CompactFlash is much slower than the standard of the new flash memory being introduced.
Also I am of the camp that says computers should be off when not in use. This statement obviously excludes servers. I just believe that turning it off increases the lifetime of the unit, and leaving it on is an unnecessary power consumption. The equivalent (albeit crude due to Energy Saver technology) is leaving your car running in the garage all the time. Being a tech, I see too many power adapters, power supplies, and hard drives fail.
I agree! That's why I disconnect the battery from my car when I park, lest the motor in the clock wear out.
Seriously, even a small flash drive complements a HDD and I want one. For a sub-notebook, a 32 GB flask drive and a WiFi connection to a 'virtual' remote drive on my desktop for rarely-accessed files. Keep Apple TV, give me a Mac I can comfortably use on my lap.
[QUOTE=Brendon;1054973]64 GB is more than enough, no one will ever need more than 64 GB, in fact Apple should just hard wire this into their OS, nothing above 64 GB for the flash drives. /QUOTE]
InfoWorld:
"Flash memory-based solid-state disks look poised to quickly replace hard-disk drives in laptops"
For example, SanDisk debuted its first SSD, a 32GB model, at January's CES but a mere six months on at Computex it's showing a 64GB model. The company says much higher-capacity drives are possible today but will be too expensive for most enterprise users, so it's increasing the capacity of its drives while keeping them at what it considers the sweet-spot of price and storage space.
Apacer Technology is demonstrating a 128GB industrial SSD that can replace a 2.5-inch hard drive and operate at temperatures between -40 degrees Celsius and 85 degrees Celsius. It will be available in the fourth quarter, and a second version with double the data read speed of 200Mbps will be available in early 2008.
Comments
Also I am of the camp that says computers should be off when not in use. This statement obviously excludes servers. I just believe that turning it off increases the lifetime of the unit, and leaving it on is an unnecessary power consumption. The equivalent (albeit crude due to Energy Saver technology) is leaving your car running in the garage all the time. Being a tech, I see too many power adapters, power supplies, and hard drives fail.
Why would putting systems to sleep rather than shutting down cause hard drives to fail faster? Don't the drives spin down in either case? Are you sure you're a tech? As for power consumption, even a sleeping Mac Pro draws only about 5 watts. The various wall warts around your house almost certainly draw more collectively. Sleep it for a full day and you'll eat about the same power as a 100 watt light bulb in one hour. Macs don't draw 0 watts even when shut down, either. They're always drawing at least a watt or two since they don't have hard power switches, so the power supply is always active, unless you actually unplug it or switch it off with a power strip. For some of us, the increased productivity of being able to get back to work instantly with exactly the files and applications we've been working on still open is worth a few watts. And what about all the drive activity required to cold boot a Mac? Wouldn't that reduce HD lifetime?
Why would putting systems to sleep rather than shutting down cause hard drives to fail faster? Don't the drives spin down in either case? Are you sure you're a tech? As for power consumption, even a sleeping Mac Pro draws only about 5 watts. The various wall warts around your house almost certainly draw more collectively. Sleep it for a full day and you'll eat about the same power as a 100 watt light bulb in one hour. Macs don't draw 0 watts even when shut down, either. They're always drawing at least a watt or two since they don't have hard power switches, so the power supply is always active, unless you actually unplug it or switch it off with a power strip. For some of us, the increased productivity of being able to get back to work instantly with exactly the files and applications we've been working on still open is worth a few watts. And what about all the drive activity required to cold boot a Mac? Wouldn't that reduce HD lifetime?
Google has over 100,000 PATA and SATA drives in their storage cluster.
They recently released results on a study on hard drive reliability.
Their conclusion was that drives that spin up and spin down will fail faster than drives that don't.
They also found that heat had almost no effect on reliability.
They also found that drives are most likely to fail in the first 6 months.
Failure rates then drop down for a period of 18 months.
Once the drive is two years old the failure rate begins to increase at an appreciable rate.
And think of the heat and noise issues or lack thereof, especially when you have 4 drives in a tower like I do! This rocks, can't wait for larger sizes and price drops.
It's a few years too early for that. Flash storage is still about 10x more per GB than desktop hard drives. For me, most hard drives are quiet enough. If you have a drive that's making a little bit too much noise, usually replacing it with another hard drive will get you a quieter drive. Heat doesn't seem to be a problem.
Google has over 100,000 PATA and SATA drives in their storage cluster.
They recently released results on a study on hard drive reliability.
Their conclusion was that drives that spin up and spin down will fail faster than drives that don't.
They also found that heat had almost no effect on reliability.
They also found that drives are most likely to fail in the first 6 months.
Failure rates then drop down for a period of 18 months.
Once the drive is two years old the failure rate begins to increase at an appreciable rate.
Some of that is a bit of an oversimplification. Reading one of their charts (fig 2), it looks like the failure rates in the second year exceeds the failure rates of the entire first year, adding up the rates for the first three time periods (all first year) still doesn't exceed the 2yr point. Other charts in the same paper say things a bit differently, depending on use and, and less on temperature. I haven't read the paper in a bit and only reskimmed it today, so I may be a bit off-base on it.
1.5GB+?
I know personally, I don't shut down my laptop ever. So I could case less about the "increased startup time". But having a nice fast external drive would be nice.
NAND flash has a finite number of read/write cycles, something in the order of 100,000. For storing digital photos, this is entirely adequate, as the memory will unlikely to be refreshed more than once or twice per day.
However, most modern operating systems create a paging file, which is frequently written to, erased, written, in a much more frequent cycle.
Replacing a computer's boot drive with a flash unit, WITHOUT OS MODIFICATIONS TO TAKE ACCOUNT OF THE R/W LIMITATION, is likely to prematurely age the flash drive.
Check out the article again. Intel has made a big stride here in progress.
Intel says the Z-U130s will distinguish themselves from other solid state product offerings by their extensive validation, including more than 1,000 hours of accelerated reliability testing. The drives are expected to meet an average mean time between failure (MTBF) specification of five million hours.
Why would putting systems to sleep rather than shutting down cause hard drives to fail faster? Don't the drives spin down in either case? Are you sure you're a tech?
I think that last comment is a bit uncalled for.
However, I will address the previous sentences anyway.
It is ludicrous to assume that any piece of technology that remains on 100% the time is going to be just as reliable as the equivalent device that is powered up 25% of the time or less depending on usage, or have as long of a lifetime. Your laptop in precious "sleep mode" still draws electricity, still generates heat, and heat has effects over time. I will listen to your argument when electronic equipment generates zero heat, has zero chance of condensation damage, has zero change of being susceptible to other environmental factors like accidental dropping, water penetration, dust...the list goes on. And spin down is not the same as power down.
Only when sleeping laptops use no power, generate no heat, and have no moving parts that are susceptible to weardown will I tell people that it's ok to leave them on all the time. Until then, I will continue to recommend to shut down computers when not in use. Thankfully, I work with respectable people who ask for my humble recommendation and don't go around asking if I'm really a tech when I give it. They can then choose to follow it or do whatever they want.
What size is your average OS X install these days?
1.5GB+?
It's more than just the base OS. As shipped, before I did anything with it, my MacBook Pro had 30GB already taken. If you don't use iLife, you can probably clean out 10+GB there if you uninstall it, but that means gutting some of Apple's selling points of having that software included. The iWork apps take 2GB (included as demos), probably more with the sample themes that are in the library somewhere. 32GB really isn't much to work with when it comes to Macs.
^^^ I am pretty sure Intel took care of that flash memory r/w thingy everyone wants to bring up.
Check out the article again. Intel has made a big stride here in progress.
I really don't believe they've really addressed it. The press release and product brief don't discuss the write cycle limitation of flash memory. Intel did demo PRAM recently, which I think does take care of it, but that's not flash, but they have not mentioned it in the product information.
Only when sleeping laptops use no power, generate no heat, and have no moving parts that are susceptible to weardown will I tell people that it's ok to leave them on all the time.
I don't think it is necessarily clear cut. Even parts that don't look like they are moving do have to deal with thermal expansion on use, and contraction on cool down. This fatigues the metals, joints and contacts in a circuit, and something may eventually fail. Hard drive motors are rated for a certain number of start cycles because they have to expend a lot more energy per unit time spinning up than they do maintaining the spin through inertia. Notebook hard drives are rated for a lot more start/stop cycles because drives are supposed at every reasonable opportunity to save power.
I don't know how much of a problem start/stops are, it can happen, but I really don't think it's a major problem so long as one isn't stupid about shutting down too often. I figure once a day is probably fine. If you want to keep your workspace, hybernation will do the trick.
It is ludicrous to assume that any piece of technology that remains on 100% the time is going to be just as reliable as the equivalent device that is powered up 25% of the time or less depending on usage, or have as long of a lifetime. Your laptop in precious "sleep mode" still draws electricity, still generates heat, and heat has effects over time. I will listen to your argument when electronic equipment generates zero heat, has zero chance of condensation damage, has zero change of being susceptible to other environmental factors like accidental dropping, water penetration, dust...the list goes on. And spin down is not the same as power down.
Then what is power down? Do you really think motherboards keep power flowing to hard drives for no reason whatsoever? You must really think electronic engineers are fools. Sleep generates internal heat? How much heat? Less than 10ºF on any given day. Most of the heat will be in the transformer in the power adapter, and transformers don't wear out. You'll have more temperature variation between day and night or when taking the laptop outside. All your "concerns" about condensation, dust, water, dropping, etc. would seem to be better arguments against taking any laptop off its desk. Is that what you're advocating?
Only when sleeping laptops use no power, generate no heat, and have no moving parts that are susceptible to weardown will I tell people that it's ok to leave them on all the time. Until then, I will continue to recommend to shut down computers when not in use. Thankfully, I work with respectable people who ask for my humble recommendation and don't go around asking if I'm really a tech when I give it. They can then choose to follow it or do whatever they want.
Ah, so only "respectable" people agree with you. Maybe we're just not as susceptible to being impressed by anyone who comes along using self-invented "tech" terms like "weardown." If you can cite any REAL information other than your "learned personal opinion," I'd be happy to accept it. Otherwise, I'll treat it no differently than any other old wives' tales and urban legends, all of which have staunch supporters who swear they're indisputable.
Besides, I don't hold "techs" in very high regard, having dealt with way too many ignorant ones in the past. In fact, one of my closest friends is not only a technician, he's the system administrator for a medium-size newspaper. And even he admits he would never have made it past the first week in an engineering college.
This would be ideal, IMO, if it were combined with a dock having a larger HDD and optical drive. That would be sweet.
I'm willing to bet that CompactFlash is much slower than the standard of the new flash memory being introduced.
Also I am of the camp that says computers should be off when not in use. This statement obviously excludes servers. I just believe that turning it off increases the lifetime of the unit, and leaving it on is an unnecessary power consumption. The equivalent (albeit crude due to Energy Saver technology) is leaving your car running in the garage all the time. Being a tech, I see too many power adapters, power supplies, and hard drives fail.
I agree! That's why I disconnect the battery from my car when I park, lest the motor in the clock wear out.
Seriously, even a small flash drive complements a HDD and I want one. For a sub-notebook, a 32 GB flask drive and a WiFi connection to a 'virtual' remote drive on my desktop for rarely-accessed files. Keep Apple TV, give me a Mac I can comfortably use on my lap.
InfoWorld:
"Flash memory-based solid-state disks look poised to quickly replace hard-disk drives in laptops"
http://www.infoworld.com/archives/em...ming-on_1.html
For example, SanDisk debuted its first SSD, a 32GB model, at January's CES but a mere six months on at Computex it's showing a 64GB model. The company says much higher-capacity drives are possible today but will be too expensive for most enterprise users, so it's increasing the capacity of its drives while keeping them at what it considers the sweet-spot of price and storage space.
Apacer Technology is demonstrating a 128GB industrial SSD that can replace a 2.5-inch hard drive and operate at temperatures between -40 degrees Celsius and 85 degrees Celsius. It will be available in the fourth quarter, and a second version with double the data read speed of 200Mbps will be available in early 2008.