Associate
Same, but mid game something is going weird and my RPM jumps to over 6000. If i manually go in and turn it to 100% it still only got to about 3.5 ... weird fault. Really noisy.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
The R9 Nano Fiji Owners Thread
- Thread starter Kaapstad
- Start date
Same, but mid game something is going weird and my RPM jumps to over 6000. If i manually go in and turn it to 100% it still only got to about 3.5 ... weird fault. Really noisy.
What temps you getting?
According to AMD hardware monitor I get a steady 80C on one card, core clock hovers between 970Mhz-1Ghz and fan speed sits comfortably at 59%. The other card doesn't seem to be doing anything.
I don't seem to have any games which fully utilise the second card unfortunately :/ Even Unigine Valley with a xfire profile and the second card sits there doing nothing. Same with BF1.
.... off to disable ULPS...
Sort of thinking of getting one of these. Is the coil whine as wide spread (almost guaranteed) and horrendous as I've been led to believe?
Sent back 2x Fury Tri-X and 1x Fury Nitro because it sounded like my pc case had been invaded by an incredibly high pitched wasps nest, so a little wary...
Sent back 2x Fury Tri-X and 1x Fury Nitro because it sounded like my pc case had been invaded by an incredibly high pitched wasps nest, so a little wary...
Small sample size of two for me, well just one now but both have coil whine. Latter card was not massively noticeable on stock cooler, but being under water in the PC it is in, it can make a fair bit of a whine.
Par the course for Fury cards I think as like you two Sapphire Tri-X cards prior to the Nanos also had a fair bit of whine.
Par the course for Fury cards I think as like you two Sapphire Tri-X cards prior to the Nanos also had a fair bit of whine.
Soldato
Even the GTX1080s have coil whine, is not exclusive to Furys.
However my Nano was silent after few weeks having coil whine.
However my Nano was silent after few weeks having coil whine.
True enough, but no where near to the extent the Fury cards (or at least all of mine had) even with consistent use. Unless I got unlucky with all 4 Fury based cards. ( possible knowing my luck
)
Soldato
True enough, but no where near to the extent the Fury cards (or at least all of mine had) even with consistent use. Unless I got unlucky with all 4 Fury based cards. ( possible knowing my luck
PSU is responsible for the whine, mainly.
Soldato
- Joined
- 24 Aug 2009
- Posts
- 2,931
I had enough of my problems with my nano, from day 1 the card wasn't running right. It kept throttling no matter how much I raised the power limit, + or -mv voltage. The card wouldn't run over 43c under WC.
I tried updated BIOS from AMD, reseated the card in the motherboard, and tried different power cables. Tried different drivers and OS.
Put air cooler back on today, tested it in brothers computer to make sure the card still had fault so I could raise RMA, worked perfect. Solid 1GHZ core clock, no fluctuation what so ever. I couldn't get my head round it?
My new 1070 works perfectly, so must have been a compatibility issue somewhere with hardware?
I tried updated BIOS from AMD, reseated the card in the motherboard, and tried different power cables. Tried different drivers and OS.
Put air cooler back on today, tested it in brothers computer to make sure the card still had fault so I could raise RMA, worked perfect. Solid 1GHZ core clock, no fluctuation what so ever. I couldn't get my head round it?
My new 1070 works perfectly, so must have been a compatibility issue somewhere with hardware?
True enough, but it's not due to PSU's. I tried them with a pair of 1200 watt platinum units (HX 1200i and EVGA P2 1200) in my main builds which have no issues with non Fiji cards, so leaning towards the Fiji gpus making more noise then the norm (or at least all mine did / still do). Non issue for my use per say as its an awsome performer and is in a HTPC sitting far from me anyways.
Still an element of lottry with whine, so meh.
Last edited:
Decided to try one, decent performance for it's size. Fan isn't too bad, quieter than my MSI 290 was for sure, but far from silent.
Has coil whine almost constantly, even gets it opening up an internet browser with GPU acceleration on... impressive... It's deeper pitched than it was on my Tri-Xs though, so overall probably less annoying.
Going back however, as one of the DP ports is completely dead.
Has coil whine almost constantly, even gets it opening up an internet browser with GPU acceleration on... impressive... It's deeper pitched than it was on my Tri-Xs though, so overall probably less annoying.
Going back however, as one of the DP ports is completely dead.
Was messing around with Trixx / Afterburner and they appear to have completely disabled GPU clock monitoring (always shown at 300Mhz in either them or Wattman) .. which makes tinkering useless as you can't see if there's any clock throttling.
Anyone any idea how to restore the functionality, or do I need to uninstall them?
I would use Wattman, but it doesn't offer over and undervolting, or HBM overclocking.
Anyone any idea how to restore the functionality, or do I need to uninstall them?
I would use Wattman, but it doesn't offer over and undervolting, or HBM overclocking.
The latest drivers changed something so 3rd party tools like Trixx, AB etc don't currently work correctly and running them will lock the GPU clock to 300MHz.
You can have them installed just not run them and then Wattman, other monitoring tools - gpu-z etc - should work ok.
Either need to roll back to previous drivers or wait for the tools to be updated. IIrc this happened before with a driver update and then was eventually resolved.
You can have them installed just not run them and then Wattman, other monitoring tools - gpu-z etc - should work ok.
Either need to roll back to previous drivers or wait for the tools to be updated. IIrc this happened before with a driver update and then was eventually resolved.
Thanks. Having now looked into it, it seems to constantly happen since they introduced Wattman.
Do you know that 17.7.2 (latest) broke it again, or is it just in a semi-permanent state of brokenness now?
It specifically broke again with the latest drivers, from my c:\amd "WHQL-Win10-64Bit-Radeon-Software-Crimson-ReLive-17.7.2-July26".
I was using Sapphire Trixx fine prior to this up to the previous driver "Non-WHQL-Win10-64Bit-Radeon-Software-Crimson-ReLive-17.7.1-July10"
This did happen once before, iirc, months - maybe last year - or so back but got fixed at some point.
Not sure if last time it was a driver fix, Trixx update or both but from a bit of reading around I did when 17.7.2 broke it again recently it looks like this time it will need to be a Trixx/AB update.
If you roll back to 17.7.1, it should work again until the tools get updated
I was using Sapphire Trixx fine prior to this up to the previous driver "Non-WHQL-Win10-64Bit-Radeon-Software-Crimson-ReLive-17.7.1-July10"
This did happen once before, iirc, months - maybe last year - or so back but got fixed at some point.
Not sure if last time it was a driver fix, Trixx update or both but from a bit of reading around I did when 17.7.2 broke it again recently it looks like this time it will need to be a Trixx/AB update.
If you roll back to 17.7.1, it should work again until the tools get updated
Soldato
+1
Also anyone using Nano these days, please for the love of <what ever you believe>, upgrade the BIOS to the AMD UEFI one from April 2016.
Not only you will gain performance, ability to overclock to 1100/550, but UEFI for faster system boot. (my nano was different beast when i did so last year, so is my FuryX compared to stock original BIOS).
Thanks, indeed it worked.
That's interesting. I'd missed that. I assume it has better power management and therefore lower temps and ease of reaching high clocks?
1100 isn't attainable on my current BIOS, but I've chosen 1045 GPU core to stop it thermal throttling under repeated benchmark runs - normal gaming I can easily do 1075. HBM ... I currently have it at 615Mhz, though I'm not sure if it actually applies any increments between 600 and 666Mhz settings, and Afterburner 'only' offers up to 650Mhz.
This is all on air, with the original cooler.
I didn't try overclocking the HBM before, so don't know how much effect the following has had, but this spate of OC'ing interest was predicated by it beginning to thermal throttle more heavily. So I took it apart, cleaned the original TIM off (of which there was FAR too much) and applied Liquid Metal Ultra to the die and HBM. No way to measure HBM temps, but the GPU temps dropped massively, and it stopped throttling at stock clocks. I suspect the reason I can get such an HBM OC (i.e. 600+) is because 1) the original TIM is gone and 2) LMU provides much better contact with the heatsink on those tiny HBM modules than any conventional TIM will.
Anyway, seems to be performing well ahead of stock 1070s / Fury Xs / 980Tis now.
Soldato
Yeah. Is a great card, power starved unfortunately
Mine was plugged for well more than year in a Predator 360 and worked beautifully. It worth to try the UEFI BIOS, it wont do harm, on the contrary. After all they come with 2 BIOS. You flash the second one only.
My FuryX (got second hand in March) initially couldn't break 1070mhz. With the UEFI bios can do 1237 at +96mv, but running it all the time at 1190/600 with +48mv. Ain't worth for 47mhz the extra power and heat.