How long will my laptop last? (i never turn it off)

[hmscott]

I wouldn't put any mesh or filter in front of your vents. That could restrict airflow and just cause overheating. Sorry about the misdirection with closing the lid. I hadn't though about the heat dissipation through the keyboard.

Nonpossible, the G750 keyboard is sealed, there is no airflow through it. The G750 is pretty much closed to airflow except through the bottom vents.

You should be able to leave the screen lid down and not affect the cooling or CPU/GPU temperature.

Filtering the intake air with a mesh can be done without causing air flow problems - as long as you keep the outside of the mesh clear from obstruction - wipe the dust off.

Use hwinfo64 to monitor the CPU/GPU temps while the keyboard is covered with the lid down, and when the lid is up, see if it makes a difference. Same for the mesh.

[GoAsusGo!]

Nonpossible, the G750 keyboard is sealed, there is no airflow through it. The G750 is pretty much closed to airflow except through the bottom vents.

You should be able to leave the screen lid down and not affect the cooling or CPU/GPU temperature.

Filtering the intake air with a mesh can be done without causing air flow problems - as long as you keep the outside of the mesh clear from obstruction - wipe the dust off.

Use hwinfo64 to monitor the CPU/GPU temps while the keyboard is covered with the lid down, and when the lid is up, see if it makes a difference. Same for the mesh.

thanks for the info :-)

[Amenophis]

My friends laptop caught fire after only 7 years of use. He refused to clean the exhaust pipes on it and never vacuumed the dust from the top of the unit. He kept it in the hottest place in the house where the humidity was very extensive. Suddenly, there was a burst of smoke that rolled out from the back and he poured a jug of water into it to stop the blaze. He says he'll never by another ASUS but will buy the infamous Toshiba. I'm sure if he treats his new laptop the way he did his ASUS, he'll have to have the fire department on speeddial. Let's hear it people! Let's have a round of applause for Jake for not taking care of his laptops. A silent applause will do now , won't it? Hmmmmmmm?

[Darnassus]

My friends laptop caught fire after only 7 years of use. He refused to clean the exhaust pipes on it and never vacuumed the dust from the top of the unit. He kept it in the hottest place in the house where the humidity was very extensive. Suddenly, there was a burst of smoke that rolled out from the back and he poured a jug of water into it to stop the blaze. He says he'll never by another ASUS but will buy the infamous Toshiba. I'm sure if he treats his new laptop the way he did his ASUS, he'll have to have the fire department on speeddial. Let's hear it people! Let's have a round of applause for Jake for not taking care of his laptops. A silent applause will do now , won't it? Hmmmmmmm?

... Blue Lady hit you now. ;x

[Korth]

Yes, fan bearings/motors do seize after years of use.
"Cheap" sleeve-bearing types are usually rated around 20,000 to 30,000 hours (27 to 41 months). They get louder and louder when they're dying and failure tends to be gradual, you'll probably get tired of hearing their angry noise long before they die.
"Better" ball-bearing types are usually rated around 50,000 hours (68 months). They develop annoying high-pitched rattling, ticking, or whining oscillations a little while before they die, again you'll tire of their noise before it happens.
"Superior" bearing types are usually rated much higher. Noctua's IndustrialPPC SSO2 bearings are rated >150,000 hours (over 17 years)! Never had one of these die on me before, lol, I don't know if there's any symptoms of failure.

A problem is that there's no standard for specifying fan longevity. Different fan manufacturers have different engineering (and marketing) methods. Unless specified, you don't know what parameters they use in their calculated MTBF/MTTF specs - fan duty-cycles, volts, rpms, measures for 50% or 75% or max 100% or what?

A bigger problem is that ASUS provides no specs for your G750 fans. The only way to know what their specs are is to examine their components and lookup their part numbers. At the very least, they should be marked with their Voltage (usually 12V or 5V) and their Amps or Watts, you can measure their frame dimensions, you can determine their motor type by the number of wires they feed into a mobo fan header - so replacement fans are always available, even if they aren't made by ASUS, even if you have to order generic Alibaba stuff.

If you have a mechanical HDD system drive then it might fail before the fans do. Especially if it stays busy all the time, or the machine gets moved around and bumped a lot.

If you have an SSD system drive then it also has a limited working lifespan, it should take years to fail completely but it suffers from gradually diminishing capacity and performance every time anything gets written to it, so you may find that while it still works it's also become a piece of junk after a few years.

The battery chemistry will last 2 or maybe 3 three years before it starts losing capacity and getting hot. It might keep working for many years. But most laptop batteries won't hold any charge within at most 5 years (depending on how they've been used or abused), they just become AC-powered laptop heaters.

The motherboard (or more likely, some electrolytic cap or inductor on it) will eventually fail, but good motherboards still work after a couple decades. CPU can fail. Thermal pastes tend to dry/burn and fail (and force the machine into thermal shutdown until repasted) after some years.

Heat accelerates electrical failures. Although an always-on machine is actually held in a steady thermal state (until something changes or fails) which is gentler on components than the thermal surges which flow through a machine whenever it's powered on. Overclocking anything will burn stuff out faster.
Dust accelerates mechanical failures. It even gets into HDDs (which are barometrically sealed, not hermetically sealed as many people believe). And it's a wonderfully efficient thermal insulator so it also increases component temps and accelerates thermal failures.

Lacking any specific component data, the warranty period of the laptop is a fair indicator of useful lifespan.

I wonder why you'd dedicate a somewhat costly G750 laptop to this always-on media server role when an inexpensive immobile desktop could do it just as well.

[Amenophis]

... Blue Lady hit you now. ;x

Yeah, oooook ... not sure what you're driving at but apparently you do have something on your mind???

[Korth]

Server's CPU lifespan is 10 years for being constant on. Will they last longer? likely... but they are made for 10 years. Since these laptops do not have server quality CPU, I assumed ~50% of server life. I got the info from a friend that works at intel.

Most servers can run much longer than 10 years, lol, or at least servers built over 10 years ago could.

Very few actually maintain this sort of epic uptime with extremely high availability. Mostly because of downtime caused by software faults, not caused by hardware faults.

I don't know where this "10 years" figure comes from. Xeon CPUs have 3 year Intel warranty. Xeon Phi cards have 1-2 year Intel warranty. Workstation (Xeon-chipset) mobos have (usually 2-3 year, sometimes 5 year) manufacturer warranty. GPU cards (workstation and consumer) have 3 year NVIDIA/AMD warranty. Server-grade PSUs have (at least 5 year, sometimes 7 year or 10 year) manufacturer warranty.

In fact, most enterprise hardware isn't even "owned" by the end-users, components and platforms are leased over a contract period (typically up to 3 years) from an OEM like ASUS, Supermicro, or Dell. The apparently high hardware purchase cost is actually divided up into a per-year hardware cost, it includes full service and support from the OEM, it ends up costing almost exactly the same in the accounting as purchasing the hardware asset up front then depreciating the value of the asset (as the hardware becomes "obsolete") over time. Hardware problems are fixed "for free". Hardware upgrades are frequent (every new lease). Depreciated hardware doesn't need to be stored (ongoing expense) or liquidated (below actual value). It's a multi-billion-dollar industry, they OEMs have exhaustively calculated every penny and every detail (even not-yet-invented computing technologies) with exacting precision, and they make their money by making this sort of leasing model the most "affordable" way for corporations to access enterprise machinery.

The stuff is built to last, sure, unexciting but proven reliability and redundancy and scalability. A typical rack holds a bunch of blade units which are each packed with multiple (modestly or massively multi-core) Xeon CPUs, multiple (modestly or massively GPGPU) workstation GPUs, and/or multiple (very large or very fast) storage devices, along with error-correction and mirroring and power backups built into every major block. They emphasize stability and reliability over security or performance or longevity (although these are all parameters linked to reliability), and could hardly care less about cosmetics because an industrial heatsink or generic green PCB works better or costs less than stylized consumer parts anyhow. A properly-configured stable server can last 10 years, even 20, maybe more - but hardly any of them are kept running that long or even kept out of the landfill that long.

The last few generations of high-end consumer mobos have been using server-grade components along with very robust VRM designs, socket/slot enhancements, DIMM placements, thermal management, etc. They might last just as long as enterprise gear, at least if they're run within rated spec (not overclocking anything). The "~50% of server life" assumption seems reasonable in the absence of real data. Then again, I've got an ancient computer (875P mobo, Pentium 4 CPU, 4GB DDR, HD4670/1GB AGP, a mighty gaming beast back then) which has been running continuously (with many reboots) since 2008, and I had another ancient computer (780G mobo, Athlon II X3 CPU, 8GB DDR2, strange HD7870/2GB GPU) which ran nonstop for over 4 years before a power outage then would never restart again (dead mobo, not worth fixing lol). Gaming mobo companies like ASUS are now building gaming mobos with extreme performance and excellent reliability, while enterprise companies like Supermicro are building gaming mobos with extreme reliability and excellent performance - the lines were never clearly defined before, but now they're just downright blurry.

[Clintlgm]

I am shocked that it only lasted 7 years in those conditions!!!

There are 3 components of fire
Fuel Dust Flammable and explosive if saturate airborne
Heat Dust matting make a great heat retaining insulator
Oxygen the fans are bring it in all the time or with out fans the heat it self rising draws in fresh air carrying O2 to the heat and fuel.

[JustinThyme]

Personally I dont understand leaving a portable machine on 24x7. Much cheaper alternatives that will last longer.

As for how long it will last?
There is no magic number. Could be a month could be ten years.
Just dont expect the battery to hold up.

[Korth]

Personally I dont understand leaving a portable machine on 24x7. Much cheaper alternatives that will last longer.

As for how long it will last?
There is no magic number. Could be a month could be ten years.
Just dont expect the battery to hold up.

That's what I said!

Although you were far more concise, lol.