Original SSDs Beginning 7th Year of Running In Trading Rigs

Scataphagos

Scataphagos

In case anyone is wondering about "reliability" or "longevity" of SSDs in real-time use...

My original SSDs... Intel X25-Ms, are beginning their 7th year of operation in trading rigs which run about 80 hrs/week.

FWIW...
 
Last edited:
xandman said:
Do you think it's necessary to turn off the computer for longevity?
More...

Don't know. On the one hand, constant current running through the hardware has a slowly degrading effect on the longevity of certain components. But "current surge" from repeated startups has one as well. Which effect is greater ??

Some people leave their computer on 24/7. I've always rebooted mine daily, as none of them need to run 24/7.. Seems intuitive the cost of electricity to run constantly should be considered.

As a side note... Mythbusters tested the myth that leaving the lights on all the time saved energy (to avoid "current surge") over turning them off when nobody was in the room. Data showed conclusively that the lights should be turned off. Likely same applies to computers if considering energy cost savings.
 
Last edited:
One thing to note. The first generation product from a name brand company, which you probably have, is usually built bullet proof or simply overbuilt.

One can only imagine what happens when they start outsourcing production. Nonetheless, I'm sold on the advantages. HDD is a big bottleneck.
 
xandman said:
One thing to note. The first generation product from a name brand company, which you probably have, is usually built bullet proof or simply overbuilt.
More...

Likely true. The original SSDs were talking, "should be good for 5,000-10,000 write cycles". There was an Intel X25-V which had done 24/7 reads/writes for 3 years... 40,000 write cycles and didn't fail. Now days, the "warranty" on some SSDs is for as few as 150 write cycles... likely they will run many multiples of that, but still.

There was another endurance test which ended last year(?) where 250GB SSDs were tortured. The first to fail, as expected, was the Samsung 840 Evo (with newer/cheaper TLC NAND)... and it lasted for 600TB of writes.... that's 2,400 cycles. All of the rest of the test drives made it past 1PB.
 
Last edited:
stevegee58 said:
I've run several computers 24/7 for years. The only component I've had to replace was a power supply.
More...

My experience is that the components most likely to fail are PSU, optic drive, and RAM... then the system drive. I've had virtually every component fail at one time or another (even video cables), except for the CPU.
 
So is the HD still the biggest bottleneck? Or is there something else we can set our sights on.

Seems with storing and manipulating financial data, it still is. Perhaps, there will be ways to hardcode the OS/Apps in a ROM chip.
 
xandman said:
So is the HD still the biggest bottleneck? Or is there something else we can set our sights on.

Seems with storing and manipulating financial data, it still is. Perhaps, there will be ways to hardcode the OS/Apps in a ROM chip.
More...

I guess it depends upon use. In data intensive enterprise utilization with constant writes, the drive still likely the greatest bottleneck.

In high video output applications like gaming... CPU, GPU, and RAM work together. The drive doesn't do much except load everything into RAM.
 
Last edited:
You must log in or register to reply here.
Back
Top