Overclock.net › Forums › Graphics Cards › AMD/ATI › Fiji Bios Editing ( Fury / Fury X / Nano / Radeon Pro Duo )
New Posts  All Forums:Forum Nav:

Fiji Bios Editing ( Fury / Fury X / Nano / Radeon Pro Duo ) - Page 85

post #841 of 1780
A couple weeks back I decided to try and tackle negative voltage scaling on the Nano. I decided to try voltage (VID) offset adjustment (via Sapphire Trixx) combined with DPM7 frequency and voltage
adjustments.

Before I get too far with this, if your card doesn't respond positively to -/+ offset voltage, then likely none of this will help you.

My particular Nano responds well to negative voltage offset (-32mv). But this limits me to a max stable overclock of 1100 Mhz. To clock to 1130 with negative scaling I need "0" voltage offset. For higher frequencies I require a "+" offset, with greater resulting negative scaling.
Straight ROM mods with no VID offset present me with a slight but not great improvement and not better performance than 1100 Mhz. frequency with negative offset. I assumed this was still a form of negative scaling.
I decided the key was to find the voltage offset "sweet spot" for DPM's 0-6 (I'll explain how to get this if anyone is interested); with a ROM (Bios) frequency and voltage adjustment to DPM7 (@ stable overclock). One of the problems with making a offset voltage adjustment; is that this effects VID for DPM7 as well. To compensate a farther adjustment to DPM7 VID setting equal and opposite to offset voltage would be required.
Now we enter into a secondary problem. Max stock VID is just shy of 1.3v. This in effect limits Max stable allowable DPM7 frequency to Max applicable voltage. So this application is voltage limited without voltage mods.

With all this said, VID DPM7 @ 1130 is 1.268v (max) via DPM dump on my Nano. So 1.268 + (-1 x -.032) = 1.3v, so problem. But 1125 Mhz. @ 1.292 v is stable with -032 offset (relative actual VID is 1.26).

The result of this (margin of error) VERY slightly better 3DM 11 score and no negative scaling, but power and possibly heat throttling. Air cooled Nano's don't need more heat production so this doesn't help me.

To go father with this a someone with good results to offset voltage, voltage mods and adequate cooling would have to try.

Gupsterg has tried this idea with no improvement, but none of his Fury's respond well to offset voltage.

Side note: With Gusterg's results on offset voltage I'm left to wonder if platform (mother boards) effects results with offset voltage.

Anyhow. This is one of the things I've been up to.

Cheers
post #842 of 1780
Quote:
Originally Posted by gupsterg View Post

Fellow members I would like to discuss HBM clocking smile.gif .

So basically when I got my Fiji cards I noted AMD Matt had stated on a few forums that HBM clocks in discreet steps, 500.00/545.45/600.00/666.66MHz. Now bear in mind this guy is involved with AMD and also that he has no reason to share misinformation. I was surprised by this information, as when HBM overclocking became available in MSI AB, the slider would not increment in such steps as AMD Matt had posted about.

Basically HBM overclocking via MSI AB (other OC tools) will not increment in steps as it has no knowledge how HBM increments.

Next let's look at OverDrive page, the slider for HBM clock became available through bios mod, again OverDrive has no knowledge how HBM increments. If you install older Catalyst driver it will increment in 1MHz steps, with Crimison drivers it will increment in 5MHz steps.

Next VRAM_Info, this section sets up the HBM RAM. In that section we can see there are 100MHz, 400MHz, 500MHz and 600MHz straps/timings, this information is not again solid proof how HBM steps.

So basically for months viewing at all of above I was confused what is going with HBM wth.gif . 3DM13 had been the most sensitive benchmark on Hawaii for RAM tweaks, so I used it on Fiji as well. Now as HBM performance scaling was so small, it become difficult to use this bench data to categorically say how HBM clocks.

Then I thought stability testing is the answer, like highlighted in my previous posts I had assessed 1135/535 was stable for:-

i) lengthy f@h runs (12hrs+).
ii) 3DM13/Heaven/Valley each looped for over 1hrs each.
iii) general gaming.

Where as 540MHz or 540MHz was creating issues.

Then one day whilst running f@h my 1135 (+31mV) / 535 (+0mV) was failing in f@h sadsmiley.gif, I lowered HBM to 525MHz, which still did not solve the "bad state" in GPU slot. Finally only using 500MHz HBM clock solved the issue, I tested then ~ 60hrs f@h with 1135 (+31mV) / 500 (+0mV), to me this meant GPU OC was fine and HBM OC is the issue. This also meant my idea to test HBM clocking steps by stability testing was a flop doh.gif .

Due to all the wasted time testing HBM OC and lack of performance gain from it I thought just tweak lowering MVDDC below stock whilst being at 500MHz. This exercise showed that 1.2V @ 500MHz is not stable for all uses I had for GPU, ~1.263V was. This again was confusing and a setback IMO thinking.gif . SK Hynix state 1.2V, AMD state 1.3V, I can only conclude from this testing that HBM 1 due to perhaps production variation requires 1.3V.

So after failing to successfully OC HBM with full stability and under volting at stock clock I was no closer to knowing how the steps of HBM clocking were headscratch.gif. Whilst going over past data I came across an AIDA64 GPGPU benchmark which I had done on Fiji, this to me seemed like the EUREKA moment biggrin.gif.

HBM Clock/stepping testing (Click to show)
AMD Fury X 107 ROM used, only mods were:-

i) HBM clock set to 100MHz in PowerPlay.
ii) OD RAM Limit raised to 600MHz so HBM OC'ing available without using "Extend Official Overclocking Limit" in MSI AB.
iii) MVDDC set to 1.325V for testing upto HBM 600MHz.

Note: Due to my i5 rig running stability testing of Fury X No 2 @ 1135/545 I used my Q6600 rig for testing. I will update data ASAP, excluding Memory Read / Write all other bench data is spot on with i5+Fury X run done in the past.

HBM 100MHz (Click to show)

HBM 400MHz (Click to show)

HBM 500MHz (Click to show)

HBM 520MHz (Click to show)

HBM 535MHz (Click to show)

HBM 545MHz (Click to show)

HBM 565MHz (Click to show)

HBM 575MHz (Click to show)

HBM 600MHz (Click to show)


Basically 100MHz vs 400MHz huge difference, then there is large difference between 400MHz vs 500MHz. 520MHz is clocking at 500MHz. 535MHz is clocking at 545MHz and so did 565MHz. 575MHz is clocking to 600MHz and so did 630MHz (due to time limit ran 1 run, thus no screenie). I increased HBM voltage upto 1.35V to test 666MHz but as soon as bench started I got artifact on screen.

So viewing the data above I can see now why 535MHz and 525MHz was failing stability testing recently, I can only conclude why 545MHz and 540MHz artifact quicker in stability testing is due to possibly HBM clock getting close to correct step thus aspects we're unaware of concerning HBM/MC coming into play.

On the i5+Fury X rig I have been testing for just over 61.5hrs continuously using f@h to get 1135/545 stable. Initially I set HBM voltage stock and over a 22.5hr run GPU entered "bad state" twice and did not lose a work unit. So then I upped HBM voltage by +6.25mV, within 12hrs I had 1x bad state. I then increased HBM voltage to +12.5mV, within 5hrs 1x bad state, so then I went +18.75mV so far passed 22hrs f@h and going biggrin.gif . I'm gonna let this card fold until at least another 9hrs wink.gif .

I hope members will share data on AIDA64 GPGPU benchmark for HBM clocks testing.

REP +1.
post #843 of 1780
Quote:
Originally Posted by bluezone View Post

A couple weeks back I decided to try and tackle negative voltage scaling on the Nano. I decided to try voltage (VID) offset adjustment (via Sapphire Trixx) combined with DPM7 frequency and voltage
adjustments.

Before I get too far with this, if your card doesn't respond positively to -/+ offset voltage, then likely none of this will help you.

My particular Nano responds well to negative voltage offset (-32mv). But this limits me to a max stable overclock of 1100 Mhz. To clock to 1130 with negative scaling I need "0" voltage offset. For higher frequencies I require a "+" offset, with greater resulting negative scaling.
Straight ROM mods with no VID offset present me with a slight but not great improvement and not better performance than 1100 Mhz. frequency with negative offset. I assumed this was still a form of negative scaling.
I decided the key was to find the voltage offset "sweet spot" for DPM's 0-6 (I'll explain how to get this if anyone is interested); with a ROM (Bios) frequency and voltage adjustment to DPM7 (@ stable overclock). One of the problems with making a offset voltage adjustment; is that this effects VID for DPM7 as well. To compensate a farther adjustment to DPM7 VID setting equal and opposite to offset voltage would be required.
Now we enter into a secondary problem. Max stock VID is just shy of 1.3v. This in effect limits Max stable allowable DPM7 frequency to Max applicable voltage. So this application is voltage limited without voltage mods.

With all this said, VID DPM7 @ 1130 is 1.268v (max) via DPM dump on my Nano. So 1.268 + (-1 x -.032) = 1.3v, so problem. But 1125 Mhz. @ 1.292 v is stable with -032 offset (relative actual VID is 1.26).

The result of this (margin of error) VERY slightly better 3DM 11 score and no negative scaling, but power and possibly heat throttling. Air cooled Nano's don't need more heat production so this doesn't help me.

To go father with this a someone with good results to offset voltage, voltage mods and adequate cooling would have to try.

Gupsterg has tried this idea with no improvement, but none of his Fury's respond well to offset voltage.

Side note: With Gusterg's results on offset voltage I'm left to wonder if platform (mother boards) effects results with offset voltage.

Anyhow. This is one of the things I've been up to.

Cheers

What kinda cooling is your Nano on?
8 CAORS
(13 items)
 
Dota 2 Radiator
(5 items)
 
ITX Powerhouse
(8 items)
 
CPUMotherboardGraphicsRAM
AMD FX 8350  Asus M5A99X Evo R2.0 Asus 7970 ROG Matrix Platinum Corsair Vengeance 16GB (2x8GB) 1600Mhz 
Hard DriveHard DriveCoolingOS
Samsung 840 Evo 128GB WD Caviar Green 1TB Coolermaster Seidon 120M Windows 7 Ultimate 
MonitorPowerCaseMouse Pad
BenQ XL2411T Coolermaster V550S  Coolermaster Storm Stryker Steelseries Diablo III QcK Mousepad 
Other
Zalman ZM-MFC3 Controller 
CPUMotherboardGraphicsRAM
Intel Core i5 2410M Some Quanta nvida Geforce GT540M Hyundai 6GB DDR3 
OS
Windows 7 
CPUMotherboardGraphicsRAM
Intel 6700k Asrock Z170 Gaming-ITX/ac R9 Nano Corsair Vengeance LPX 16GB 
Hard DriveCoolingPowerCase
Samsung 850 Evo 250GB M.2 CoolerMaster Seidon 120V Plus Silverstone SX-500 LG Silverstone Sugo SG13B 
  hide details  
Reply
8 CAORS
(13 items)
 
Dota 2 Radiator
(5 items)
 
ITX Powerhouse
(8 items)
 
CPUMotherboardGraphicsRAM
AMD FX 8350  Asus M5A99X Evo R2.0 Asus 7970 ROG Matrix Platinum Corsair Vengeance 16GB (2x8GB) 1600Mhz 
Hard DriveHard DriveCoolingOS
Samsung 840 Evo 128GB WD Caviar Green 1TB Coolermaster Seidon 120M Windows 7 Ultimate 
MonitorPowerCaseMouse Pad
BenQ XL2411T Coolermaster V550S  Coolermaster Storm Stryker Steelseries Diablo III QcK Mousepad 
Other
Zalman ZM-MFC3 Controller 
CPUMotherboardGraphicsRAM
Intel Core i5 2410M Some Quanta nvida Geforce GT540M Hyundai 6GB DDR3 
OS
Windows 7 
CPUMotherboardGraphicsRAM
Intel 6700k Asrock Z170 Gaming-ITX/ac R9 Nano Corsair Vengeance LPX 16GB 
Hard DriveCoolingPowerCase
Samsung 850 Evo 250GB M.2 CoolerMaster Seidon 120V Plus Silverstone SX-500 LG Silverstone Sugo SG13B 
  hide details  
Reply
post #844 of 1780
Thread Starter 
@bluezone

I wouldn't say "none of his Fury's respond well to offset voltage." smile.gif .

Let's say my stock VID is 1.212V, I require +32mV in MSI AB to achieve 1135MHz GPU and have no negative scaling in bench. Then I manually set VID to 1.243V in PowerPlay the bench result will be the same as test 1. Then if I increase VID in PowerPlay to 1.275V and set a negative offset using MSI AB or ROM, so final VID is 1.243V, the result will be the same as test 1 & 2.

I also conducted testing where GPU/HBM clock was stock and using the various methods final VID would be equal to stock EVV VID and all benches were equal, with only run to run variance.

On another note just about to wrap up f@h smile.gif , done 48hrs solid error free on 1135 / 545 biggrin.gif .





log.txt 192k .txt file
Edited by gupsterg - 6/10/16 at 6:58am
The Stilt Machine
(14 items)
 
XPS - i5 4690K
(12 items)
 
XPS - Q6600
(8 items)
 
CPUMotherboardGraphicsRAM
Ryzen 7 1700 Asus Crosshair VI Hero Sapphire Fury X@1145/545 G.Skill Trident Z 2x 8GB 3200MHz C14 
Hard DriveCoolingCoolingCooling
Samsung Evo 840 ThermalRight Archon IB-E X2 + 2x TY143 ThermalRight TY-143 2x front case intake Arctic Cooling F12 + 2x F9 as rear case exhaust 
OSMonitorKeyboardPower
Win 7 Pro x64 / Win 10 Pro x64 Asus MG279Q Cherry MX-Board 3.0 Cooler Master V850 
CaseMouse
SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
i5 4690K@4.9GHz 1.255V Asus Maximus VII Ranger Sapphire Fury X@1145/545 HyperX Savage 2x8GB 2400MHz C11 
Hard DriveCoolingOSMonitor
Samsung Evo 840 TR Archon SB-E X2 Win 7 Pro x64 / Win 10 x64 Asus MG279Q 
KeyboardPowerCaseMouse
Cherry MX-Board 3.0 Cooler Master V850 SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
Intel Q6600 G0 Asus P5K Premium Black Pearl Sapphire Toxic HD5850 Corsair Dominator 4GB 
Hard DriveCoolingOSPower
Crucial MX 100 256GB TR TRUE Rev.A Win 7 Pro x64 Cooler Master V650 
  hide details  
Reply
The Stilt Machine
(14 items)
 
XPS - i5 4690K
(12 items)
 
XPS - Q6600
(8 items)
 
CPUMotherboardGraphicsRAM
Ryzen 7 1700 Asus Crosshair VI Hero Sapphire Fury X@1145/545 G.Skill Trident Z 2x 8GB 3200MHz C14 
Hard DriveCoolingCoolingCooling
Samsung Evo 840 ThermalRight Archon IB-E X2 + 2x TY143 ThermalRight TY-143 2x front case intake Arctic Cooling F12 + 2x F9 as rear case exhaust 
OSMonitorKeyboardPower
Win 7 Pro x64 / Win 10 Pro x64 Asus MG279Q Cherry MX-Board 3.0 Cooler Master V850 
CaseMouse
SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
i5 4690K@4.9GHz 1.255V Asus Maximus VII Ranger Sapphire Fury X@1145/545 HyperX Savage 2x8GB 2400MHz C11 
Hard DriveCoolingOSMonitor
Samsung Evo 840 TR Archon SB-E X2 Win 7 Pro x64 / Win 10 x64 Asus MG279Q 
KeyboardPowerCaseMouse
Cherry MX-Board 3.0 Cooler Master V850 SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
Intel Q6600 G0 Asus P5K Premium Black Pearl Sapphire Toxic HD5850 Corsair Dominator 4GB 
Hard DriveCoolingOSPower
Crucial MX 100 256GB TR TRUE Rev.A Win 7 Pro x64 Cooler Master V650 
  hide details  
Reply
post #845 of 1780
Thread Starter 
I have come to the conclusion HBM overclocking is not at all worthwhile gains for me on day to day basis sadsmiley.gif , this is considering the increased MVDDC required for small performance gain redface.gif .

Not had time to create the table of AIDA64 GPGPU Benchmark but will be doing so soon. All the HBM investigation will form part of "How to edit HBM clock" section in OP. I never added this in OP previously as needed to get things together like now, hopefully some of this "stuff" will be handy to ref when we have HBM 2 cards biggrin.gif.

I now think HBM overclocking was not accessible so easily by an OverDrive slider as in previous gen cards due to:-

a) very minuscule gains in performance from OC'ing it.
b) the MC/HBM steps.

Benches (Click to show)
3DM FS 3 runs of 500MHz vs 545MHz, link. Next out of those results the best run of 500MHz vs 545MHz, link. Next the worst one of 3 runs of 500MHz vs 545MHz, link. Finally the worst run of 500MHz vs best run of 545MHz, link.

Next 400MHz timings in 500MHz & 600MHz strap, 545MHz vs 545MHz+TMOD, link. Next out of those results the best run of 545MHz vs 545MHz+TMOD, link. Next the worst 545MHz vs 545MHz+TMOD, link. Finally the worst 545MHz vs best 545MHz+TMOD, link.

AIDA64 GPGPU Benchmark 545MHz vs 545MHz TMOD (Click to show)


I have previously posted compare of 400MHz vs 500MHz, link. At the time I had been wondering effect of tighter timings in 400MHz strap, now I reckon ever so slightly to none at all. I may setup a ROM where 400MHz has 500MHz timings in and retest.

So next stock ROM vs 1135/545+TMOD, link. For that compare I've taken the closest physics scoring 3DM FS bench for 1050/500 & 1135/545+TMOD, this will hopefully mean CPU was performing equal for those runs redface.gif . Anyone who notes PE=Off in the 1050/500 bench, I'd just like to state it has no effect on 3DM benches from what I've seen.


Yesterday when I commenced 3D load testing of 1135 / 545 I required +25mV MVDDC, a single step increase over f@h testing. Just for research I'm continuing to use GPU: 1135 (~+31mv) HBM: 545 (+25mV) with TMOD for next few days to see how it shapes up in normal uses / gaming, etc.
The Stilt Machine
(14 items)
 
XPS - i5 4690K
(12 items)
 
XPS - Q6600
(8 items)
 
CPUMotherboardGraphicsRAM
Ryzen 7 1700 Asus Crosshair VI Hero Sapphire Fury X@1145/545 G.Skill Trident Z 2x 8GB 3200MHz C14 
Hard DriveCoolingCoolingCooling
Samsung Evo 840 ThermalRight Archon IB-E X2 + 2x TY143 ThermalRight TY-143 2x front case intake Arctic Cooling F12 + 2x F9 as rear case exhaust 
OSMonitorKeyboardPower
Win 7 Pro x64 / Win 10 Pro x64 Asus MG279Q Cherry MX-Board 3.0 Cooler Master V850 
CaseMouse
SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
i5 4690K@4.9GHz 1.255V Asus Maximus VII Ranger Sapphire Fury X@1145/545 HyperX Savage 2x8GB 2400MHz C11 
Hard DriveCoolingOSMonitor
Samsung Evo 840 TR Archon SB-E X2 Win 7 Pro x64 / Win 10 x64 Asus MG279Q 
KeyboardPowerCaseMouse
Cherry MX-Board 3.0 Cooler Master V850 SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
Intel Q6600 G0 Asus P5K Premium Black Pearl Sapphire Toxic HD5850 Corsair Dominator 4GB 
Hard DriveCoolingOSPower
Crucial MX 100 256GB TR TRUE Rev.A Win 7 Pro x64 Cooler Master V650 
  hide details  
Reply
The Stilt Machine
(14 items)
 
XPS - i5 4690K
(12 items)
 
XPS - Q6600
(8 items)
 
CPUMotherboardGraphicsRAM
Ryzen 7 1700 Asus Crosshair VI Hero Sapphire Fury X@1145/545 G.Skill Trident Z 2x 8GB 3200MHz C14 
Hard DriveCoolingCoolingCooling
Samsung Evo 840 ThermalRight Archon IB-E X2 + 2x TY143 ThermalRight TY-143 2x front case intake Arctic Cooling F12 + 2x F9 as rear case exhaust 
OSMonitorKeyboardPower
Win 7 Pro x64 / Win 10 Pro x64 Asus MG279Q Cherry MX-Board 3.0 Cooler Master V850 
CaseMouse
SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
i5 4690K@4.9GHz 1.255V Asus Maximus VII Ranger Sapphire Fury X@1145/545 HyperX Savage 2x8GB 2400MHz C11 
Hard DriveCoolingOSMonitor
Samsung Evo 840 TR Archon SB-E X2 Win 7 Pro x64 / Win 10 x64 Asus MG279Q 
KeyboardPowerCaseMouse
Cherry MX-Board 3.0 Cooler Master V850 SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
Intel Q6600 G0 Asus P5K Premium Black Pearl Sapphire Toxic HD5850 Corsair Dominator 4GB 
Hard DriveCoolingOSPower
Crucial MX 100 256GB TR TRUE Rev.A Win 7 Pro x64 Cooler Master V650 
  hide details  
Reply
post #846 of 1780
Quote:
Originally Posted by Flamingo View Post

What kinda cooling is your Nano on?

I'm using a modified air cooled Nano. I have stacked fans for greater air flow and reduced fan noise and heat.
Quote:
Originally Posted by gupsterg View Post

@bluezone

I wouldn't say "none of his Fury's respond well to offset voltage." smile.gif

Sorry I misunderstood our PM's. But it's all good and thank you for the correction.smile.gif
Quote:
Originally Posted by gupsterg View Post

I have come to the conclusion HBM overclocking is not at all worthwhile gains for me on day to day basis sadsmiley.gif , this is considering the increased MVDDC required for small performance gain redface.gif .

Not had time to create the table of AIDA64 GPGPU Benchmark but will be doing so soon. All the HBM investigation will form part of "How to edit HBM clock" section in OP. I never added this in OP previously as needed to get things together like now, hopefully some of this "stuff" will be handy to ref when we have HBM 2 cards biggrin.gif.

I now think HBM overclocking was not accessible so easily by an OverDrive slider as in previous gen cards due to:-

a) very minuscule gains in performance from OC'ing it.
b) the MC/HBM steps.

Benches (Click to show)
3DM FS 3 runs of 500MHz vs 545MHz, link. Next out of those results the best run of 500MHz vs 545MHz, link. Next the worst one of 3 runs of 500MHz vs 545MHz, link. Finally the worst run of 500MHz vs best run of 545MHz, link.

Next 400MHz timings in 500MHz & 600MHz strap, 545MHz vs 545MHz+TMOD, link. Next out of those results the best run of 545MHz vs 545MHz+TMOD, link. Next the worst 545MHz vs 545MHz+TMOD, link. Finally the worst 545MHz vs best 545MHz+TMOD, link.

AIDA64 GPGPU Benchmark 545MHz vs 545MHz TMOD (Click to show)


I have previously posted compare of 400MHz vs 500MHz, link. At the time I had been wondering effect of tighter timings in 400MHz strap, now I reckon ever so slightly to none at all. I may setup a ROM where 400MHz has 500MHz timings in and retest.

So next stock ROM vs 1135/545+TMOD, link. For that compare I've taken the closest physics scoring 3DM FS bench for 1050/500 & 1135/545+TMOD, this will hopefully mean CPU was performing equal for those runs redface.gif . Anyone who notes PE=Off in the 1050/500 bench, I'd just like to state it has no effect on 3DM benches from what I've seen.


Yesterday when I commenced 3D load testing of 1135 / 545 I required +25mV MVDDC, a single step increase over f@h testing. Just for research I'm continuing to use GPU: 1135 (~+31mv) HBM: 545 (+25mV) with TMOD for next few days to see how it shapes up in normal uses / gaming, etc.

I had hoped that you would succeed where I was having no luck. Thank you for all for all of your hard work on this.thumb.gif

REP +1.

I made up a Bios tool to help determine optimum voltage per DPM. It worked so well I decided to test out negative voltage scaling effects. The results were interesting. Running 3DM 11.

The following is using Crimson 16.6.1.

Frist, relatively stock Bios @ 1100 Mhz. with 50% PL -24 mv offset results for reference. Pay attention to the graphics score only.



Now Bios tool 1100 Mhz. @ 0 PL -24mv offset..



Now Bois tool 1100 Mhz @ 50% 0 offset.



They are all very close. PL adjustment appears to be almost ineffective because of the locked frequency. Remember PL is port of the control system.
The next snapshots are 50% PL with increasing Mhz. and voltage offset. I adjusted voltage offset to the min. required to get through a run. So likely not fully stable.

1150 Mhz. @ +24 mv offset.



1165 Mhz. @ +48 mv offset. This was barley stable but I'm out of voltage adjustment at this point.



Notice no negative voltage scaling. I think the negative voltage scaling (performance) we are seeing is an artifact of the power and control software/hardware on the Fury (maybe just the Nano).
I'm thinking that there is hard to observe down clocking happening with normal Bios and control systems using + voltage offset and higher frequencies.

Looks like we need to crack the Bios and hardware further. Anyone have any ideas?
Edited by bluezone - 6/11/16 at 5:19pm
post #847 of 1780
Quote:
Originally Posted by bluezone View Post

I'm using a modified air cooled Nano. I have stacked fans for greater air flow and reduced fan noise and heat.

Ooh I remember the Silent Wings mod? Or is it different? Do you have a picture, can you share please? What about the VRMs? Dont they get too hot?
8 CAORS
(13 items)
 
Dota 2 Radiator
(5 items)
 
ITX Powerhouse
(8 items)
 
CPUMotherboardGraphicsRAM
AMD FX 8350  Asus M5A99X Evo R2.0 Asus 7970 ROG Matrix Platinum Corsair Vengeance 16GB (2x8GB) 1600Mhz 
Hard DriveHard DriveCoolingOS
Samsung 840 Evo 128GB WD Caviar Green 1TB Coolermaster Seidon 120M Windows 7 Ultimate 
MonitorPowerCaseMouse Pad
BenQ XL2411T Coolermaster V550S  Coolermaster Storm Stryker Steelseries Diablo III QcK Mousepad 
Other
Zalman ZM-MFC3 Controller 
CPUMotherboardGraphicsRAM
Intel Core i5 2410M Some Quanta nvida Geforce GT540M Hyundai 6GB DDR3 
OS
Windows 7 
CPUMotherboardGraphicsRAM
Intel 6700k Asrock Z170 Gaming-ITX/ac R9 Nano Corsair Vengeance LPX 16GB 
Hard DriveCoolingPowerCase
Samsung 850 Evo 250GB M.2 CoolerMaster Seidon 120V Plus Silverstone SX-500 LG Silverstone Sugo SG13B 
  hide details  
Reply
8 CAORS
(13 items)
 
Dota 2 Radiator
(5 items)
 
ITX Powerhouse
(8 items)
 
CPUMotherboardGraphicsRAM
AMD FX 8350  Asus M5A99X Evo R2.0 Asus 7970 ROG Matrix Platinum Corsair Vengeance 16GB (2x8GB) 1600Mhz 
Hard DriveHard DriveCoolingOS
Samsung 840 Evo 128GB WD Caviar Green 1TB Coolermaster Seidon 120M Windows 7 Ultimate 
MonitorPowerCaseMouse Pad
BenQ XL2411T Coolermaster V550S  Coolermaster Storm Stryker Steelseries Diablo III QcK Mousepad 
Other
Zalman ZM-MFC3 Controller 
CPUMotherboardGraphicsRAM
Intel Core i5 2410M Some Quanta nvida Geforce GT540M Hyundai 6GB DDR3 
OS
Windows 7 
CPUMotherboardGraphicsRAM
Intel 6700k Asrock Z170 Gaming-ITX/ac R9 Nano Corsair Vengeance LPX 16GB 
Hard DriveCoolingPowerCase
Samsung 850 Evo 250GB M.2 CoolerMaster Seidon 120V Plus Silverstone SX-500 LG Silverstone Sugo SG13B 
  hide details  
Reply
post #848 of 1780
Quote:
Originally Posted by Flamingo View Post

Ooh I remember the Silent Wings mod? Or is it different? Do you have a picture, can you share please? What about the VRMs? Dont they get too hot?

It's not the Silent Wings. I thought of doing this awhile ago, but took me a while to get around to. The fan is just one off an old master cooler 101. With the brackets left on it for easy mounting. lol biggrin.gif



I'm not using the Bios I mentioned above as a daily driver. It was for test purposes only. It's far too hot running, but not as much heat production as I first thought it would have.
Staying stuck at 1165 Mhz. is a bit of a buzz kill for power usage. devil.gif
Edited by bluezone - 6/11/16 at 7:57pm
post #849 of 1780
Hi there. I'm using the R9 Nano with watercooling. With Evga Precision 1060 MHz at -48mV and +26% PL. I wish to use this seetings in a bios, but it's too complex for me to edit the bios. Also I have the problem in Idle/Browser/Office the Nano clocks very high, up to 900 MHz and higher.
post #850 of 1780
Hey Guys,

A tip if you are struggling with your GPU temps in the summer, or just need to lower temps for overclocking smile.gif .. A "fix" can be to place 3 x 120mm SP fans in a PCI FAN bracket above your GPU - pulling air away from the backplate. In my case this lowers the temps by +10% . In my case from 78c to 70c at full load with card overclocket. (1030Mhz/545Mhz mem/3840 shaders)+ i think it looks awesome thumb.gif

(GPU fans at 60%)

BlingCode
(12 items)
 
VR Plex Machine
(11 items)
 
 
CPUMotherboardGraphicsGraphics
Intel i7-7700 ASUS Maximus IX Code EVGA GTX1080 FTW with EK FWT Acetal+Nic​kel Block EVGA GTX1080 FTW with EK FWT Acetal+Nic​kel Block 
RAMHard DriveHard DriveOS
Corsair Dominator Platinum Chrome 32GB (4 x 8GB... Samsung 960 Evo 1TB M.2-2280 Solid State Drive Samsung 850 EVO-Series 1TB 2.5" Solid State Drive Windows 10 Pro 
MonitorPowerCaseMouse
Asus PG279Q Corsair RM1000i Thermaltake Suppressor F51 Logitech G602 
CPUMotherboardGraphicsRAM
AMD Ryzen 5 1600 Biostar X370GTN itx EVGA GTX1080ti FTW3 Corsair Vengeance LPX  
Hard DriveCoolingOSMonitor
Samsung 960 EVO  Noctua NH-D15 SE-AM4 Windows 10 Pro 64 Bit LG OLED 65C6V 
KeyboardPowerCase
Logitech K400 wireless EVGA Supernova 750 G3 Fractal Design Node 304 
  hide details  
Reply
BlingCode
(12 items)
 
VR Plex Machine
(11 items)
 
 
CPUMotherboardGraphicsGraphics
Intel i7-7700 ASUS Maximus IX Code EVGA GTX1080 FTW with EK FWT Acetal+Nic​kel Block EVGA GTX1080 FTW with EK FWT Acetal+Nic​kel Block 
RAMHard DriveHard DriveOS
Corsair Dominator Platinum Chrome 32GB (4 x 8GB... Samsung 960 Evo 1TB M.2-2280 Solid State Drive Samsung 850 EVO-Series 1TB 2.5" Solid State Drive Windows 10 Pro 
MonitorPowerCaseMouse
Asus PG279Q Corsair RM1000i Thermaltake Suppressor F51 Logitech G602 
CPUMotherboardGraphicsRAM
AMD Ryzen 5 1600 Biostar X370GTN itx EVGA GTX1080ti FTW3 Corsair Vengeance LPX  
Hard DriveCoolingOSMonitor
Samsung 960 EVO  Noctua NH-D15 SE-AM4 Windows 10 Pro 64 Bit LG OLED 65C6V 
KeyboardPowerCase
Logitech K400 wireless EVGA Supernova 750 G3 Fractal Design Node 304 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD/ATI
Overclock.net › Forums › Graphics Cards › AMD/ATI › Fiji Bios Editing ( Fury / Fury X / Nano / Radeon Pro Duo )