Flexible PCIe risers

Discussion in 'Small Form Factor Systems' started by iFreilicht, Aug 3, 2016.

  1. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    A lot of new case designs are using flexible PCIe risers, and for a good reason. They allow for flexibility when placing the GPU, enabling much more efficient layouts.

    There are many different risers out there, so I thought it would be helpful to have a thread where we can collect information on those risers and discuss recent developments regarding those risers.

    So here are all the ones I have in my private research notes. I'll expand and correct this list until a suitable section on sffwiki.net is made.

    Generic riser:

    [​IMG]

    Price: ~3-20$
    Shielded: No
    Length: ~5-30cm
    Connector: Straight
    Available: ebay, multiple sellers.

    These risers have been used in quite a few builds without causing problems. Hahutzy specifically reported that he still hasn't had an issue using one of these in Hassium and the Hutzy XS. Other users like Runamok81 haven't been that lucky and had to wrap their riser in aluminium foil to make it work.

    In general, I wouldn't recommend using a single one of those, though you can of course try your luck.


    LiHeat:

    [​IMG]

    Price: ~20-50$
    Shielded: Yes, copper tape
    Length: 5-50cm
    Connector: Straight ("D.type"), left-angle ("B.type" - seen in picture), right-angle ("A.type")
    Available: Through ebay-user liheat48 or by e-mail.


    These risers are pretty much glorified generic risers. They are made up of regular ribbon cables with copper tape in between to prevent crosstalk and metallic tape wrapped around the outside. This makes them quite stiff and thick, more than 2mm.

    They perform fairly well, but users on different forums report problems with some of them. These risers are now offered with plastic strain reliefs on both ends which prevent damaging of the solder joints, maybe someone can post a more recent picture.

    They are exclusively sold as leftover stock from contracts with larger companies on ebay, so it might take some time until the desired length and connector option becomes available directly. Sometimes they have a few more options when you contact them via mail, and they also make risers of custom length if you can reach a certain MOQ.

    LiHeat produced the PW-PCI-E and PW-PCI-E38 risers for LianLi, which were used in the PC-O and PC-Y6 cases. These cables have been discontinued for unknown reason, more info below.

    This riser was tested in a chained 100cm configuration and still managed to run Firestrike with minmal performance loss.

    LianLi:

    [​IMG]

    Price: ~70-80$
    Shielded: Quite probably.
    Length: 30cm or 38cm
    Connector: Right-angle
    Available: Retail and online stores.

    LianLi started using a flexible riser of some sort in the PC-C36-Muse a long time ago, but only recently revisited riser-based designs with the PC-O series and the novelty PC-Y6. These cases initially used risers from LiHeat (see above), but those were swapped out later for the PW-PCIE38-1 and PW-PCIE30-1, the supplier of which seems to be TC&C. One might suspect that this switch was caused by either supply or quality issues with the LiHeat cables. dondan claims that these risers are using twin-axial cables.

    Thermaltake:

    [​IMG]

    Price: ~30$
    Shielded: Yes.
    Length: 22cm
    Connector: Straight, very long.
    Available: Retail and online stores.


    Thermaltake started selling the AC-039-CN1OTN-C1 with the release of the not-so-SFF Core P5 Wall-mount chassis. Supplier and quality are unknown, but it might be made by sintech, see below.

    HDPLEX:

    [​IMG]

    Price: 35.5$ (as of 2016-08-03)
    Shielded: No.
    Length: 15+cm
    Connector: Left-angle
    Available: Only from HDPLEX directly.


    This riser is probably the thinnest one in existence at ~0.14mm. Compared to other risers, it is manufactured as a double-sided FPC (flexible printed circuit). Because it isn't multi-layered, it is not shielded whatsoever. It is normally bundled with their H5 chassis, which uses a similar kind of cable for the Front USB3.0 connections.

    The connector on the motherboard side is very different from normal risers, but works perfectly fine and leaves about 2mm of space towards the edge of an ITX mainboard, so no fitment issues arise from this configuration.

    A short review can be read here. It was confirmed by HDPLEX that this riser was tested with GTX 750Ti GPUs without a 6pin connector with no problems. While it is advertised as 15cm, that seems to be the absolute minimum length this riser can have. Mine was more like 16.5cm.

    3M:

    [​IMG]

    Price: ~90-80$
    Shielded: Yes, using twin-axial cables.
    Length: 25cm or 50cm
    Connector: Straight
    Available: digikey

    Most famously used in the DAN A4-SFX, this riser seems to be the holy grail in terms of quality. 3M has already confirmed it to be compatible with PCIe 4.0 specs, which is a very impressive feat. Unfortunately, it is one of the most expensive risers to date and is not available with angled connectors. The riser uses twin-axial cables to achieve its high performance and is very flexible. Each ribbon is 0.75mm thick for a total of 1.5mm.

    This riser is also available for 8x and 4x links.

    ModDIY/sintech:

    [​IMG]

    Price: ~20-40$
    Shielded: Yes, probably using metallic tape.
    Length: 5-30cm
    Connector: Straight
    Available: modDIY (19cm and 30cm), ebay sintech.cn (5+cm), aliexpress (5+cm)

    Not a lot to say about this one. A few people have used it and so far I haven't heard of any problems. It seems very similar to the one Thermaltake is using, so maybe sintech is actually producing the Thermaltake riser.

    Internally I would suspect it to be similar to the LiHeat riser, though it looks thinner on pictures. Sintech also makes custom risers to order and has a website where a right-angle riser with shielded ribbon cables is shown.

    Adexelec:

    Price: ~80$
    Shielded: Yes and no.
    Length: Varies:
    Connector: Straight, Left-angle
    Available: Only from manufacturer.

    This company made the risers for the Orthrus Prototype, in which they seem to work great. It seems like their risers are only available made to order.

    Samtech:

    Price: ?
    Shielded: Yes, using twin-axial cables.
    Length: 30cm, other lengths available.
    Connector: Straight
    Available: Only from manufacturer

    [​IMG]

    This riser seems to be very similar to the 3M one, but with thicker cables and longer PCBs. Firewolfy tested it here.
     
    Last edited: Oct 10, 2016
  2. Hahutzy

    Hahutzy Limp Gawd

    Messages:
    169
    Joined:
    Sep 4, 2015
    I'm actually looking into getting a different riser.

    I finally ran into a problem, where if I run Win7 on Skylake, the system boots but displays a blackscreen. In safemode it displays everything fine. I suspect the cable isn't actually supporting the full PCI-E 3.0 16x.
     
    Last edited: Aug 3, 2016
  3. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    Good to know it's not too good to be true :)
     
  4. Firewolfy

    Firewolfy Limp Gawd

    Messages:
    469
    Joined:
    Sep 11, 2015
    Do we know if safemode makes the pcie run at less than 3.0?
    Or did you run safemode with the integrated graphics maybe?

    Almost sounds like a driver conflict maybe?
     
  5. Firewolfy

    Firewolfy Limp Gawd

    Messages:
    469
    Joined:
    Sep 11, 2015
    iFreilicht--

    Can you elaborate on your statement: "They perform fairly well, but users on different forums report problems with some of them." I hadn't heard of problems with performance, so it would be great if you have some info on that. Thx.
     
  6. thehack

    thehack [H]Lite

    Messages:
    92
    Joined:
    Jan 16, 2016
    I want to say thanks for the post. This showed me a couple more options for pcie risers!
     
  7. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    Well, on SFN, one user said that half of the risers he got from LiHeat were not working with some combinations of components. Some didn't work at all, others worked only after rebooting multiple times. That last one actually happened to me with my first LiHeat riser as well. Got through post, just had a black screen afterwards. After unplugging it and plugging it back in, rebooting a few times, it worked and after that it didn't show any hiccups anymore. I think what I'm trying to say is, they're great if you get them for a private rig, but if you're selling a product and include them, some users might have issues, even if you tested them all with your components.

    I don't know whether these issues were due to the unprotected solder joints or something like that, though.

    You're welcome!
     
  8. Marstg

    Marstg Limp Gawd

    Messages:
    176
    Joined:
    Jul 31, 2016
    I had a generic 1x to 16x riser and used to run a gtx 470 on a 2P socket F mobo and i had no issues in gaming, ever.
     
  9. Hahutzy

    Hahutzy Limp Gawd

    Messages:
    169
    Joined:
    Sep 4, 2015
    I used safemode to disable the driver. I haven't done the research, but I assume disabling the driver makes it run at lower than 3.0, or lower than 16x. Because it never black screens when in safemode, and it never black screens when I uninstall the Nvidia driver.

    I know that those risers run fine otherwise, because my work machine uses one. It is running on a i5-4440S power my 980Ti through full PCIE 2.0 16x

    I'm fine with using these risers myself because I am ready to go in and tinker with it if it fails, but as a product it won't be fine if people run into issues like that.
     
  10. Boil

    Boil Gawd

    Messages:
    971
    Joined:
    Sep 19, 2015
    Seems like the safest bet (although also the most expensive) is the 3M riser…

    I would rather pay the extra for a high-quality riser than have to replace a lesser quality riser after purchasing a chassis that requires said riser…
     
  11. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    To be fair, though, that card was only PCIe 2.0, which had much lower signal integrity requirements. I don't know about the bus-speed requirements for decent gaming at that time, but 1x seems a bit extreme of a limitation.

    Yeah the 3M is pretty much undisputed to be the best riser out there, but the one from LianLi seems promising as well, and it could be quite a bit cheaper. The thing is that such a riser raises the cost of the case significantly, which makes the case less accessible.

    Just found a thread about the Thermaltake one, which seems to have issues as well.
     
  12. Soul_Est

    Soul_Est n00bie

    Messages:
    9
    Joined:
    Sep 12, 2014
    Thank you for this excellent overview. It has definitely helped me with the research part of a couple projects I'm working on.
     
  13. dondan

    dondan [H]ard|Gawd

    Messages:
    1,482
    Joined:
    Apr 15, 2013
    iFreilicht Good overview. Maybe some of you wonder why the 3M one works stable in Gen 3 while normal ribbon or ribbon shielded have troubles. The trick is twin axial cables. Risers like 3M, Lian Li and two more that are not in this overview using this typ of cable. Every wire of twin acial is shielded and the shield is set to ground. The next important thing is the resist of the cable. You need 85 ohm for PCIe Gen3+.

    The special thing on 3M version is the thickness of the cables and the tiny bending radius. The cables are 0.75mm thick and you could bend them 1000 times without breaking the imner wire and loosing Performance.


    By the way I think I forund the manufacture of the LianLi cables:
    http://www.tcac.com.tw/pics/16X_gen3b.jpg
     
    Last edited: Aug 6, 2016
    TheHobbyist, Boil, illram and 3 others like this.
  14. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    Would you mind sharing those? ;) I'll update the post with the information you provided.

    So that means overall the riser is 1.5mm thick because there are two ribbons atop each other. Is that correct?
     
  15. Soul_Est

    Soul_Est n00bie

    Messages:
    9
    Joined:
    Sep 12, 2014
    Samtec is very designer/engineer friendly going by that webpage.
     
  16. Boil

    Boil Gawd

    Messages:
    971
    Joined:
    Sep 19, 2015
    Definitely looks like the 3M riser cable is the hands down no compromise winner in the PCIe riser cable battles; curious if the Lian-Li twin-axial design meets or exceeds that benchmark…!
     
  17. Carbon

    Carbon n00bie

    Messages:
    2
    Joined:
    Jan 24, 2016
    Just as a warning for those looking at these for the NCase i could not fit the 3M riser in order to move the graphics card to the 2nd and 3rd slots
     
  18. dondan

    dondan [H]ard|Gawd

    Messages:
    1,482
    Joined:
    Apr 15, 2013
    This is right because the the pcie port is not angled.
     
  19. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    The HDPLEX riser should work nicely for this, but it's a little bit longer than needed and you should probably protect it from being punched by solder-tails on the GPUs backside if it doesn't have a backplate.
     
  20. NameAlwaysInUse

    NameAlwaysInUse n00bie

    Messages:
    6
    Joined:
    Aug 16, 2016
    Just to report: I have 5 LiHeat (2x A-Type, 2x B-Type, 1x D-Type) risers and all work perfectly! They are really nice and answer really fast on Facebook AND they do custom lengths.
     
  21. KazeoHin

    KazeoHin [H]ardness Supreme

    Messages:
    6,618
    Joined:
    Sep 7, 2011
    I've tried two different risers in The Wall (specs below) and both get issues with no video at startup, its a PITA because you have to hard reboot the machine a few times to get the video card outputting.

    Could this be a Shielding issue?
     
  22. NameAlwaysInUse

    NameAlwaysInUse n00bie

    Messages:
    6
    Joined:
    Aug 16, 2016
    Can be, try to change the PCIE to GEN2 in the BIOS... maybe they cant do GEN3
     
  23. Firewolfy

    Firewolfy Limp Gawd

    Messages:
    469
    Joined:
    Sep 11, 2015
    What riser brands?
     
  24. KazeoHin

    KazeoHin [H]ardness Supreme

    Messages:
    6,618
    Joined:
    Sep 7, 2011
    One is the generic Ebay riser, the other LOOKS like the Thermaltake one, however it is not a Tt branded one. Funny enough, the Ebay one works more consistently.

    I'll get back to you with the brand of the other one when I get home.
     
  25. Hahutzy

    Hahutzy Limp Gawd

    Messages:
    169
    Joined:
    Sep 4, 2015
    It's possible you're experiencing the same issues as my testing.

    It's not EMI. If you boot in Safe Mode, the display should come on properly.

    The issue is the riser not providing the speeds for PCIE 3.0.
     
    KazeoHin likes this.
  26. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    Nice! What lengths do those have?
     
  27. Boil

    Boil Gawd

    Messages:
    971
    Joined:
    Sep 19, 2015
    Something that has been passed down for generations in my family, sage words of wisdom to be shared with all…

    "Never cheap out on the PCIe riser cable, son…!"

    ;^p
     
    Firewolfy likes this.
  28. NameAlwaysInUse

    NameAlwaysInUse n00bie

    Messages:
    6
    Joined:
    Aug 16, 2016
    1x 300mm, 2x 350mm, and two shorter ones
     
    iFreilicht likes this.
  29. KazeoHin

    KazeoHin [H]ardness Supreme

    Messages:
    6,618
    Joined:
    Sep 7, 2011
    Okay, so, This worked flawlessly. Changed my PCI-E Link Speed to 2.0 in the BIOS, now it displays perfectly every time. Thanks so much!
     
    DrLobotomy likes this.
  30. Firewolfy

    Firewolfy Limp Gawd

    Messages:
    469
    Joined:
    Sep 11, 2015
    Speaking of running 2.0, I remember reading a comparison test that looked at the gpu performance reduction (in game i think) of x16, x8, x4 in 2.0 and 3.0. Hopefully someone else will remember where that was.

    Anyway, the difference was almost negligible, with the 2.0 giving more than 95% performance of the 3.0.

    I'll have to dig that up.

    EDIT -- Here is one comparo
    www.pugetsystems.com/labs/articles/Impact-of-PCI-E-Speed-on-Gaming-Performance-518/
     
    Last edited: Aug 22, 2016
  31. RazorWind

    RazorWind 2[H]4U

    Messages:
    2,885
    Joined:
    Feb 11, 2001
    I've now owned four PCI-E risers:

    1. The original thermaltake design from the Core P5 - This was garbage. It didn't work with my 290X at all, and was flaky with even an older card.
    2. The revised thermaltake design from the Core P5 - Worked with my 290X, but with some apparent loss. Every benchmark I tried had a 10ish percent lower score while using it, vs. plugging the card directly into the motherboard.
    3. The ModDIY one the OP posted above - Finally, this one worked as intended. Edit: Tried with 290X and GTX 1080. Works properly with both.
    4. A generic PCI-E 1x cable with a 16X slot (I guess meant for using a graphics card for bitcoin mining?) - This one works as intended as well, with the Meinberg radio clock card I have plugged into it.
     
  32. DG25

    DG25 Gawd

    Messages:
    520
    Joined:
    Dec 15, 2006
    Came across this article about PCI-E 4.0: PCI Express 4.0 Brings 16 GT/s And At Least 300 Watts At The Slot

    "When we asked the PCI-SIG, we received the news that for the first time, PCIe will get a massive power increase at the connector. Solomon couldn't recall the exact ceiling because member companies have proposed several options. Solomon stated that the minimum would be 300W, but the ceiling “may be 400 or 500W."

    Even with the minimum 300 watts, there is more than enough power provided by the edge connector to run a GeForce GTX 1080 (reference design 180 watts) and Titan X (250 watts) without an external power source. PCI Express 3.0 provides a maximum of 75 watts thus far, though many slots support the default 25W. The rating for a x16 lane card comes from 3 amps at 3.3 volts and 5.5 amps at 12 volts, which is a rating that dates back to the introduction of PCI Express 1.0a in 2003."


    Do you think this high power rating will hinder the existence of longer extension cables? Currently they support 75w, but 300 watts or more sounds a bit worrying to me.
     
  33. RazorWind

    RazorWind 2[H]4U

    Messages:
    2,885
    Joined:
    Feb 11, 2001
    Distinctly possible, but I'd imagine that the folks making the cables will just have to make compliant cables, with thicker, stranded conductors or something. It's not like creating a cable to handle 300 watts is a new concept.

    I would be far more worried about EMI causing excessive noise in the signal portion, which already seems to be a persistent problem with inexpensive cables.
     
  34. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    There will be a new connector, so no, power shouldn't be a problem. But the increased frequency will cause all sorts of problems. If it comes to the worst, fiber optic cable might be required for distances longer than a few inches.
     
  35. Boil

    Boil Gawd

    Messages:
    971
    Joined:
    Sep 19, 2015
    With the pre-emptive statement that I may not know what I am talking about at all…

    I would assume that Fiber would carry the data signal, but a separate cabling (meaning, metal) would be needed to carry the actual power (could be combined into one cable assembly, obviously)…

    So, that said, might this lead to a MUCH smaller overall cable assembly (just the Fiber Optic date & the metallic power) BETWEEN the portion that slots into the MB & the portion that the GPU slots into…?!?

    That could be a Good Thing…!
     
  36. Hahutzy

    Hahutzy Limp Gawd

    Messages:
    169
    Joined:
    Sep 4, 2015
    Following your success, I changed the speed to Gen2 as well in the BIOS for my Hutzy XS prototype. It is now running flawlessly on the same generic riser that kept failing before.

    I'm coming to a hypothesis here, which will suggest that the generic risers are still useful:

    1) A lot of people have tried and failed to use generic risers. I think it's safe to say that the issue was their system was trying to run the connection in PCIE 3.0 x16, and the riser cannot provide that speed.

    2) I don't know exactly the reason for why it can't do PCIE 3.0 (cable impedance?), but I am quite certain it is not EMI, as in all scenarios where I have failed in using the riser, adding some sort of Faraday cage to the ribbon has not changed the result.

    3) When running in Gen2, I have still not had my risers fail on me.
    a) In my older test system, I was using a Haswell chip that can do Gen3 (i5-4440S), but my mobo -- Gigabyte H81N -- was only providing PCIE 2.0 (all H81 boards only did PCIE 2.0)
    b) When I switched to my Skylake system (i5-6400 + MSI H110I PRO AC), it started failing continuously, until today when I set it to run Gen2 in BIOS
    c) All the while, my i5-4440S and Gigabyte H81N have been running in my current workstation build -- Hutzy HS (Hassium) -- with my GTX 980 Ti, and I have not had a problem. This is the system that I do all my gaming / design / rendering / browsing on.

    So my hypothesis is this: for any build where it is instructed specifically to run at PCIE 2.0 x16, the generic riser (up to 300mm that I tried) will perform correctly and successfully.
    Any failure of performance from using these risers should result in a first-step diagnosis of checking whether the system is running at PCIE 2.0, or is it automatically trying to achieve PCIE 3.0, latter of which will make the riser fail.




    Now a follow-up for discussion:

    PCIE 2.0 x16 has the same bandwidth as PCIE 3.0 x8. So we have to ask ourselves: Is that enough? Or do we have something to gain from being able to achieve PCIE 3.0 x16?

    The current trend on the other side of the playing field is using external GPU enclosures that connect with Thunderbolt 3.
    Alienware Amp, Razer Core.. the TB3 connection they use carry a PCIE 3.0 x4 (PCIE 2.0 x8) connection on them, which in itself is another topic of discussion.
    But that's half the bandwidth the generic riser provides, and as far as I know, a GTX 1070 can't even saturate the PCIE 3.0 x4 (PCIE 2.0 x8) connection on there. GTX1080 gets close I believe.
    It might take a while for a consumer card to saturate PCIE 2.0 x16.

    So to sum it up: as far as the state of current technologies is concerned, PCIE 2.0 x16 is way more than enough, especially in the context of SFF builds.
    Achieving PCIE 3.0 x16 feels great of course, but it is essentially gaining you zero real life performance.

    So if it really is true that the generic riser can handle PCIE 2.0 x16 without fault, to me personally it is quite tempting to settle at that than paying $80 for a cable that adds headroom nothing uses and no performance gain.
     
    Last edited: Aug 23, 2016
  37. Tony Ou

    Tony Ou SilverStone Tech Representative

    Messages:
    486
    Joined:
    Jun 12, 2009
    We've tested many solutions extensively and tried to develop riser ribbons on our own for a while now. We can also vouch for 3M riser's superiority over other solutions.
     
  38. iFreilicht

    iFreilicht [H]ard|Gawd

    Messages:
    1,348
    Joined:
    Sep 23, 2014
    Well if that isn't a statement of reassurance, I don't know what is. Thanks for your input, Tony, much appreciated!
     
  39. NameAlwaysInUse

    NameAlwaysInUse n00bie

    Messages:
    6
    Joined:
    Aug 16, 2016
    3M says their riser can also do PCIe 4.0