I am using a couple of these with success, but some motherboards simply refuse to see them on the PCIE buss.
A mobo bios update might help, or might not.

Any of Your kitties coming to Europe? I'd strap back a GTX780 and some kibble with instructions to the airlines to keep good care of.

I'm afraid my two remaining kitties are much too old to be gallivanting about anymore.
But, thanks for the offer.
Meow!

Meow - see my EDIT.

I had a friendship with a cat - that friendship lasted for 17 years. MeoW!

-- for the memory of my password recovery question. --

Well, mine are both just a couple of months away from 15yo....so, no traveling for them....LOL.

So, yeah

No travelling for them.

I'll have to wait some more agile kitties to come, hopefully with their families and ready to tell their stories and stay with us a little longer.

---

To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones

I use riser cables (not cards) on three of my boxes, two of them because there's only room for two large cards inside the box and one just because the friggin' power connector is too close to the PCIe slot for a double-wide to seat properly.

Only the one riser cable on my old T7400 is a powered riser, with the extra Molex connector. That cable connects a GTX 660 that resides outside the box, while a GTX 670 and GTX 780 are mounted in the full PCIe x16 slots. Because the power supply on that box is split into 4 rails, I had to do a lot of juggling of connectors to get an acceptable balance of power to all 3 of those cards, so the extra power through the Molex was definitely necessary for that riser. I would imagine that anyone running multiple cards on a system that's not single-rail could sometimes run into problems with powered risers if they're not careful about where the power draw for each slot and connector comes from.

I have had nothing but single rail PSUs for many years here. There are worth whatever small premium they may demand in price.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

I didn't even know usb cable risers existed!
I assumed risers were mostly for limited space U1 & U2 servers.

The only one I've found on eBay Canada (available in country) is this one:
[url]http://www.ebay.ca/itm/PCI-E-16X-to-1X-Adapter-USB-3-0-Cable-Enhanced-Extender-Riser-Adapter-Set-/141988490868?hash=item210f2c6e74:g:5tgAAOSwGYVW~xPO
[/url]

Is it possible to use a longer USB cable to connect a smaller computer (with a limited PSU) to an empty tower (with just a PSU) that would house and power the GPU(s)?

Also, any pics of your external riser setups would be great eye-candy!
Cheers,
Rob :-D

---

I didn't even know usb cable risers existed!
I assumed risers were mostly for limited space U1 & U2 servers.

Also, any pics of your external riser setups would be great eye-candy!

Rob, see my new post on mess fixing...if you need more pics, I'll be happy to share !

A.

Is there any specific brand of pci-e riser that is good to use, or can any unbranded model ok for use? On several forums I have read how the bad soldering on these risers have blown the motherboards or the GPUs. Even the Thermaltake ones have received bad reviews on Amazon.

Also, is there any limitation on how long these can be? I came across this site indicating various lengths.

Many sites gave good review for 3M Riser cables from Digikey, but the costs are exorbitant compared to the others for both 25cm and 50 cm length.
______________

Vipin, couple quick comments, I think that what is mostly being discussed in this thread is the USB3 style risers, and what you referred to is what I am using, 3M's PCI-E 3.0 physical riser cables. I had purchased a number of the no name brand riser cables a few years ago, which I had used for a very short time, and seemed to work, but I mothballed that project. Recently I got back into SETI crunching, and decided that I wanted to put together a multi GPU open air system. I did a bit of reading, because I heard somewhere that there were issues with cheap versions of those cables, and found many examples like you mentioned, smoked motherboards, etc.

At that point, I decided to toss my ~$10 cables, and purchased those 3M style. Heck yes, they are expensive, but this was my thought: The cable was I think around $60-70 or so, but my higher end video cards are between $300-600, and the motherboards I put them in are multi-hundred dollars, as well as the processors, memory, etc. So, in that light, the cost of those risers is minimal compared to the rest of the setup, so the peace of mind that is provided by purchasing a high quality part from a very trusted manufacturer is well worth the cost to me.

Because, gawd forbid if anything bad happens, because it is 3M (and the price), I believe that I have recourse if it turns out that the cables are the reason for the failure. If it was one of those no name cables, then good luck with that. So, to summarize, you get what you pay for, and what you are trying to protect with that purchase will determine how price conscious you are when deciding which you are going to buy. As an aside, my system Crunchmonster is running all 4 video cards exclusively on those risers and has been since I put it together (and as a test down the road plan on adding 3 more to the system for a total of 7). I have experienced no issues with it, so it appears that they are working well for me.

---

I probably should have clarified in my earlier post that the riser cables I'm using are all standard PCIe ribbon cables, not the USB risers. I think they range from about 7 to 10 inches. I try to use the minimum length necessary to get the GPU where it needs to go, with no excess cable.

Since virtually all my hardware is inexpensive, pre-owned stuff, scrounged off of eBay and elsewhere, I went with the cheaper cables, too ($12 being at the high end, I think). No problems found so far in nearly 3 years of operation.

As to the Frankenbox mentioned in my earlier post, I don't know that the "eye candy" label would be particularly fitting. ;^) Inside are the GTX 780 and GTX 670, while the GTX 660 hangs out the back.

Could be that Your cable is missing a Connection to tell the computer that a card is present. On the card there are some "fingers" on the pci slot linked together to tell the computer that there are a card in the slot, and also the numbers of PCI lanes the card are using.

Some of the ribbon cables are missing this link, so the computer does not see the card as inserted.

Google pcie present and you'll find how to check for and make the correct Connection.

Rune

Thank you for the info Al and Jeff. It will surely help in making my decisions.

I have an update to my last post regarding bus load and risers.

It seems like one of the PCIE ports on my motherboard are the culprit.

Are now crunching away again With risers, but without the BSOD.

Is it a new motherboard, under warranty, so you can get it replaced? Hope so, bummer not having everything working properly on a new build.

---

After reading this thread it has gotten the old gears in the head going.

So opinions appreciated if my logic doesn't follow correctly.

So if I build a custom perspex case with its own powersupply and place three or so of the following http://www.gumptrade.com/pci-express-1x-to-16x-adapter-riser-card-extension-powered-usb-3-0-cable-for-bitcoin-mining.html and run several GTX660TI's seeing they are rather cheap are there any issues that I havent forseen such as power feedback through the USB cable due to the other powersupply.

Any thoughts, ideas or comments are welcome.

---

"I know I am insignificant...Just look how many stars their are...!"

I would really like to know the thoughts on these as well, looking forward to hearing from people who are knowledgeable about this technology. My only thought after taking a quick look at them is how inexpensive they are. Not that cheap always equals bad, but I have a little bit of a bad taste in my mouth from the cheap 16x PCI-E riser cables and their reputed/reported issues with occasionally shorting out components, so I am wary of things that appear to be lower cost than I would expect, if that makes sense.

*edit* Was just thinking about this a little after posting it, I have been following the USB/Thunderbolt improvements over the years, and was wondering if instead of using USB 3.0 that Thunderbolt 3.0 might be a better interconnect interface? I just did a quick search on it, and found this current article about it.

Here is an excerpt that caught my eye:

USB Type C to the rescue

With Thunderbolt seemingly on the ropes, Intel had one last move—one that likely put the technology on a winning path at last. At last year’s Computex, to the surprise of many, the company announced a faster version of the spec called Thunderbolt 3, with speeds up to 40Gbps—and it could do it over the new USB Type-C connector, instead of the funky MDP cable.

Intel essentially uses the same alternate mode that DisplayPort does to pass Thunderbolt signalling over PCIe. And by integrating a USB 3.1 10Gbps controller into the Thunderbolt 3 controller, it could fully support USB 3.1 too.

What Thunderbolt 1 and 2 couldn’t do, Thunderbolt 3 has finally achieved in its vision of “one cable to rule them all.”

A single USB Type C connector could support: DisplayPort, PCIe, high-wattage charging, and USB’s fastest spec.

Even pricing, which was always a controversial topic with Thunderbolt 1 and Thunderbolt 2, seems to have been nullified. Intel’s public price for its Thunderbolt 3 chips is about $8, with volume pricing closer to $5, according to customers I’ve interviewed. That low pricing has driven down the cost of the primary competitor, the Asmedia USB 3.1 controller, with one report putting it well under $3.

Numerous OEMs, though, told me it’s not just the price that’s changed their mind on Thunderbolt 3; it’s the move to USB-C and giving consumers a port that can do it all.

Would using this technology instead of USB 3 and a power cable be a better solution to the PCI-E 3.0 16X riser? From what it stated, the price of the components is supposedly way down, but even at the under $5 for the chip (in quantity) and from what I surmise will be a significantly less expensive cable than those earlier Thunderbolt versions, it will be a bit more expensive than the part that you linked to. I haven't seen such an animal produced yet, but to me, that would seem to be a no-brainer?

After another quick search, one cable I found doesn't appear to have done all that much in terms of improving the pricing issues they have compared to USB 3.0 cables. Ugg, that will be a tough sell for implementing it into less expensive devices, if the cheapest cable is $25. Hopefully mass production will help drive down the prices significantly.

---

Just curious, how much power does a GTX 750Ti SC pull when crunching full tilt, has anyone figured that out recently? I think 40-50 watts comes to mind, but I'm not sure. Also, are these Molex connectors on the USB-PCI-E adapters able to (safely) supply that amount of power over a longer term?

---

Just curious, how much power does a GTX 750Ti SC pull when crunching full tilt, has anyone figured that out recently? I think 40-50 watts comes to mind, but I'm not sure.

The GTX 750Ti is rated at 60W.
The highest peak load I've seen on mine has been around 75% of the maximum, so around 45W.

Also, are these Molex connectors on the USB-PCI-E adapters able to (safely) supply that amount of power over a longer term?

Indefinitely, as long as they're reasonably well made.
If they're really cheap connectors, cable, soldering/crimping then all bets are off.

GPU & CPU power connections.

---

Grant
Darwin NT

Thanks for the heads up, Grant, much appreciated!

---

There was a whole thread on the 750Ti a while back, The GTX750(Ti) Thread. Perhaps there's some power draw info in there. I'm pretty sure mine are drawing closer to 60 watts when crunching flat out, as I recall that when I replaced a GT 640 with an EVGA 750Ti (in September, 2014), my power use went up by 4-5 watts, instead of going down by about that amount, as I had expected. I also just replaced a passively-cooled GT 630 (very low power usage) with a PNY 750Ti several days ago, and the draw increased by about 45-50 watts.

Just curious, how much power does a GTX 750Ti SC pull when crunching full tilt, has anyone figured that out recently? I think 40-50 watts comes to mind, but I'm not sure.

My 1st Zotac 750 Ti with 3 Cuda50 tasks is at: 53% for Power Consumption & 89% for GPU Load (accroding to GPU-Z)
The 2nd one with 1 NV_SoG is at 64.5% for power & 98% for GPU load
Both cards are overclocked by 150MHz.

So Cuda50 is still king in 3 ways:
1. higher throughput,
2. lower power consumption; and
3. lower GPU load!

Raistmer still has a lot of tweaking to do on NV_SoG!!!

I'm pretty sure mine are drawing closer to 60 watts when crunching flat out, as I recall that when I replaced a GT 640 with an EVGA 750Ti (in September, 2014), my power use went up by 4-5 watts, instead of going down by about that amount, as I had expected. I also just replaced a passively-cooled GT 630 (very low power usage) with a PNY 750Ti several days ago, and the draw increased by about 45-50 watts.

Did you check the older cards, and the current GTX 750Ti, with GPU-Z (or similar) to see what the load on the cards was/is?

---

Grant
Darwin NT