Why give up on risers? The one you sent me works great.
So does the 12 GPU mining board I bought.
So much easier to cool if the cards are spaced out next to the computer.
Peter,
I have spent years and probably thousands of USD trying to get a reliable riser-based system to work.
That is why I was perfectly willing to send you that card.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
I used to have a lot of problems, but everything is very stable now, by:
1) Very stable and strong power supply - in my case 0 guage wire with three 1kW and one 500W supplies on it, set to 12.6V (the highest allowed in the PCI-E specs). All GPUs connected straight onto it with PCI-E power connectors on short leads.
2) Make sure the PCI-E power connector pins on the card and the plugs are shiny to make good contact. If you can't polish the ones on the card (which I can't work out how to), solder the power wires straight onto the card.
3) Make sure you have the latest driver for the cards.
4) Try not to mix different GPUs on the same computer.
5) Any misbehaving cards, replace the heatsink gunk and pads on the GPU and VRAM. If you can't make one work reliably, lower the clock speed.
6) Don't let stuff go over 80C. Increase fan power if required.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
PCIe card edge connectors can be "polished" by just scrubbing them with a pencil eraser.
If you are talking about the PCIE 6 or 8-pin power connector pins on the gpus, you can't really "polish" them physically.
BUT, as a long-time electronics tech, I learned about contact cleaners and the best one I have ever come across and used for a long-time is a contact oil called Cramolin. Comes in blue or red flavors in little glass bottles. Blue for heavily oxidized connectors and red for normal connectors. You just dipped a Q-tip into the bottle and scrubbed the connection or deposited the oil on male and female mating positions and plugged and unplugged the connection to exercise the connection and break up the oxidation on both connectors.
The cleaners are able to dissolve copper and lead or tin oxidations and get down to bright metal that produces the lowest milli-ohm resistances and then protects the metal layer from further oxidation by keeping oxygen away from the base metal. Also has a high dielectric strength that prevents arcing and strong EM fields under high-voltages.
I first learned of this product when I was in the Navy at a base repair depot and learned that it was available from well stocked electronics shops outside of military and federal contracts. Been using the product for over 30 years now. Best in class in my opinion. Always solved a myriad of issues with contact integrity and stability.
The company that produced Cramolin Red and Blue went away for a while and came back as Caig DeoxIT and then made a resurrection. You can read the history here. Anyone remember Cramolin Red?
(Although the saving seems to be because the Italian selling the Cramolin is using a very expensive postal service.)
I've tried them all. Nothing beats Caig DeoxIT or what used to be Cramolin Red.
Spray can is extremely wasteful. You don't need to bathe in the stuff. Just a tiny little micro-drop or a wet Q-tip, squeeged out, is all the amount that is ever needed.
Why give up on risers?
)
Why give up on risers? The one you sent me works great.
So does the 12 GPU mining board I bought.
So much easier to cool if the cards are spaced out next to the computer.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Peter Hucker wrote: Why give
)
Peter,
I have spent years and probably thousands of USD trying to get a reliable riser-based system to work.
That is why I was perfectly willing to send you that card.
Tom M
A Proud member of the O.F.A. (Old Farts Association). Be well, do good work, and keep in touch.® (Garrison Keillor) I want some more patience. RIGHT NOW!
I used to have a lot of
)
I used to have a lot of problems, but everything is very stable now, by:
1) Very stable and strong power supply - in my case 0 guage wire with three 1kW and one 500W supplies on it, set to 12.6V (the highest allowed in the PCI-E specs). All GPUs connected straight onto it with PCI-E power connectors on short leads.
2) Make sure the PCI-E power connector pins on the card and the plugs are shiny to make good contact. If you can't polish the ones on the card (which I can't work out how to), solder the power wires straight onto the card.
3) Make sure you have the latest driver for the cards.
4) Try not to mix different GPUs on the same computer.
5) Any misbehaving cards, replace the heatsink gunk and pads on the GPU and VRAM. If you can't make one work reliably, lower the clock speed.
6) Don't let stuff go over 80C. Increase fan power if required.
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
PCIe card edge connectors can
)
PCIe card edge connectors can be "polished" by just scrubbing them with a pencil eraser.
If you are talking about the PCIE 6 or 8-pin power connector pins on the gpus, you can't really "polish" them physically.
BUT, as a long-time electronics tech, I learned about contact cleaners and the best one I have ever come across and used for a long-time is a contact oil called Cramolin. Comes in blue or red flavors in little glass bottles. Blue for heavily oxidized connectors and red for normal connectors. You just dipped a Q-tip into the bottle and scrubbed the connection or deposited the oil on male and female mating positions and plugged and unplugged the connection to exercise the connection and break up the oxidation on both connectors.
The cleaners are able to dissolve copper and lead or tin oxidations and get down to bright metal that produces the lowest milli-ohm resistances and then protects the metal layer from further oxidation by keeping oxygen away from the base metal. Also has a high dielectric strength that prevents arcing and strong EM fields under high-voltages.
I first learned of this product when I was in the Navy at a base repair depot and learned that it was available from well stocked electronics shops outside of military and federal contracts. Been using the product for over 30 years now. Best in class in my opinion. Always solved a myriad of issues with contact integrity and stability.
The company that produced Cramolin Red and Blue went away for a while and came back as Caig DeoxIT and then made a resurrection. You can read the history here. Anyone remember Cramolin Red?
This stuff? Bit
)
This stuff? Bit expensive.
https://www.ebay.co.uk/itm/264980294458
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
I'll try this at half the
)
I'll try this at half the price:
https://www.ebay.co.uk/itm/203975751982
(Although the saving seems to be because the Italian selling the Cramolin is using a very expensive postal service.)
If this page takes an hour to load, reduce posts per page to 20 in your settings, then the tinpot 486 Einstein uses can handle it.
Peter Hucker wrote: I'll try
)
Amazon sells it too $18US for 7.4ml bottle.
[url]CAIG LABORATORIES DeoxIT D100L-2DB Brush Applicator, More Than A Contact Cleaner, 7.4mL, Pack of 1 (amazon.com)[/url]
Peter Hucker wrote: This
)
No, this stuff. Caig - DeoxIT
Not that expensive when that
)
Not that expensive when that little bottle will last you 20 years.
Peter Hucker wrote: I'll try
)
I've tried them all. Nothing beats Caig DeoxIT or what used to be Cramolin Red.
Spray can is extremely wasteful. You don't need to bathe in the stuff. Just a tiny little micro-drop or a wet Q-tip, squeeged out, is all the amount that is ever needed.