Overclock.net › Forums › Graphics Cards › AMD/ATI › [Official] Polaris Owners Club
New Posts  All Forums:Forum Nav:

[Official] Polaris Owners Club - Page 163

post #1621 of 4133
Time to brag, anybody with OCd 480 wanna compare Valley scores?

post #1622 of 4133
@BromoL I find it interesting. In GTAV I get 2 or 3 fps more at 4k with Phase Gain set to 000000 and clocked at 1300 2100 compared to 1400 2135 set to ddd000. I'm between 32 fps and 42 fps depending on the environment. Most of the time in the high 30s. That's playable in this title as the frame timing is very good. I almost never hit 40 fps at ddd000 and was in mid 30s at 1400. (settings are everything turned all the way up, frame scaling off and fxaa on at 2160p. I use v-sync. It seldom tears with it off but it tears just enough that the extra fps aren't worth it to me.

You were right about the 2 or three degrees lower vrm temps. Even on water I get 3 or 4 degrees lower vrm temps at 000000. I haven't looked at the power usage in GTAV at 4k with vsync but I throttle a lot at 1400 but not at 1300 (on stock voltage). With settings set to ddd000 both overclocks reach the same vrm temps. About 62 degrees. I'll look at power usage tomorrow. I'll measure with a killawatt for system power too. I think something is weird with my card. I'm throttling earlier even on the older 16.8.2 driver. I used to not throttle ever in Valley on this driver at 1435. Now i throttle at 1400 in valley. I think I'll run closer to stock so I can risk 000000. At least in the game I'm playing I'm getting better results right now, on my tortured card It's acting like +50% is now stuck at ~+20% power, but I'll measure tomorrow. I initially thought it was an Afterburner bug as i'd experimented with the extend official limits setting and I couldn't get AB to remove it's changes. However, a format and reload didn't give me my power limit back.

My 2nd card goes in under water on Thursday or Friday. Saturday at the latest depending on how busy I am. I'm kinda sick if I've tortured the fun OC madness out of my high asic card, but It's not like I've found too many scenarios where I need the extra performance. You were probably correct to warn me against such a high voltage/offset on a day one early silicon card. BTW this is on a reference Sapphire card under water with a slightly modified bios, based on the one that came with it, set to raise voltage and power limits to the same as the asus unlocked air bios. I'll try embd and see if that makes a difference. EMBD was where I didn't throttle in valley. I haven't run these bios settings for very long. I experimented with them last month, but I don't remember them throttling this much.

I might need to change the thermal pads on my vrm. 58 to 62 degrees on load is fine but a little high for a full cover waterblock. On my test fit of the block the vrm indentations on the pad weren't even. Maybe some of the vrms are getting too hot. I'll see about redoing that Thursday or Friday. This a moderately high leakage card. 81.9%. My 2nd card is ~74 and doesn't oc very well anyway so we'll see but I'll at least be able to objectively see if my power limit is diminished on my Sapphire reference card after I get the other reference card in the system to compare.

Wattmon is still my goto app. Thank you. I never said thanks for the profiles. They work well. Way easier than taking screenshots of different oc settings to recreate. Also, if I reboot before every memory clock setting change and only use wattmon (with monitoring turned off) and never use wattman I don't get the stuck memory performance scaling issue. It seems even the monitoring in Wattman can cause the issue once in a while for me. Since DDU and reinstall is the only way to get my performance back, you've saved me lots of time. Thanks.
post #1623 of 4133
Thread Starter 
Slightly disappointed in the rx 470 power delivery.

In 6 phase its OK but 4 phase due to the similar power characteristics of the 470 and 480 it ends up being slightly weaker than you'd want for heavy overclocking.

It's definitely not an issue in the 480 as its 6 phases, but limits the 470 a bit. If the Asus 470 uses the same power stages as the Strix 480 its actually a better choice for overclocking with ir3555.

Edit: because I'm stupid and don't know datasheets
Edited by Echoa - 9/28/16 at 2:01am
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
post #1624 of 4133
Thread Starter 
Someone tell me if I'm an idiot and reading this wrong lol does it do constant 50a sub 125c or degrade past 90c?


Edited by Echoa - 9/28/16 at 1:50am
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
post #1625 of 4133
Quote:
Originally Posted by Echoa View Post

Someone tell me if I'm an idiot and reading this wrong lol does it do constant 50a sub 125c or degrade part 90c?


It does constant 50A up until 125C as long as heat is effectively dispersed. For how effective cooling is, see graph for ambient temp.
For example:

With ambient 30C, you need 100LFM (linear feet per minute - terrible, terrible unit of measurement) of airflow to make sure it doesn't drop to 46-47A.
With ambient 50C, you now need a heatsink + 200LFM of airflow to keep it at 50A.

Where ambient is not to be considered the room temperature, but the temperature of air around the chip, which often is higher.
My Rig
(14 items)
 
Ex-wife's Rig
(15 items)
 
 
CPUMotherboardGraphicsRAM
Core i5 4460 AsRock H81M-DG4 Sapphire Rx470 Platinum KVR 1600 16Gb 
Hard DriveHard DriveCoolingOS
2x Seagate 3Tb Samsung 850 EVO 120 Scythe Ninja 3 Rev.B Windows 10 Pro 
MonitorKeyboardPowerCase
Fujitsu Siemens A17-2A Logitech K280e SuperFlower SF-550K12XP Thermaltake Versa H25 
MouseAudio
Logitech G402 Sony MDR XD150 
CPUMotherboardGraphicsRAM
Athlon 750K 4.0Ghz AsRock FM2A75 Pro4+ Sapphire R9 270X Dual-X Kingston 2x4Gb 1600 
Hard DriveHard DriveOptical DriveCooling
Samsung 850 EVO 120  Western Digital 320Gb LiteON DVD-RW CoolerMaster Hyper Z600 
OSMonitorKeyboardPower
Windows 7 Pro x64 Toshiba 32" FullHD TV Logitech FSP Hexa 550 
CaseMouse
DeLUX Logitech 
  hide details  
Reply
My Rig
(14 items)
 
Ex-wife's Rig
(15 items)
 
 
CPUMotherboardGraphicsRAM
Core i5 4460 AsRock H81M-DG4 Sapphire Rx470 Platinum KVR 1600 16Gb 
Hard DriveHard DriveCoolingOS
2x Seagate 3Tb Samsung 850 EVO 120 Scythe Ninja 3 Rev.B Windows 10 Pro 
MonitorKeyboardPowerCase
Fujitsu Siemens A17-2A Logitech K280e SuperFlower SF-550K12XP Thermaltake Versa H25 
MouseAudio
Logitech G402 Sony MDR XD150 
CPUMotherboardGraphicsRAM
Athlon 750K 4.0Ghz AsRock FM2A75 Pro4+ Sapphire R9 270X Dual-X Kingston 2x4Gb 1600 
Hard DriveHard DriveOptical DriveCooling
Samsung 850 EVO 120  Western Digital 320Gb LiteON DVD-RW CoolerMaster Hyper Z600 
OSMonitorKeyboardPower
Windows 7 Pro x64 Toshiba 32" FullHD TV Logitech FSP Hexa 550 
CaseMouse
DeLUX Logitech 
  hide details  
Reply
post #1626 of 4133
Thread Starter 
Quote:
Originally Posted by ronnin426850 View Post

It does constant 50A up until 125C as long as heat is effectively dispersed. For how effective cooling is, see graph for ambient temp.
For example:

With ambient 30C, you need 100LFM (linear feet per minute - terrible, terrible unit of measurement) of airflow to make sure it doesn't drop to 46-47A.
With ambient 50C, you now need a heatsink + 200LFM of airflow to keep it at 50A.

Where ambient is not to be considered the room temperature, but the temperature of air around the chip, which often is higher.

Alright then it's actually at least in the 480s case 10a better than stock @ 100c. The Strix is still even better doing 60a per phase.

Thanks for explaining
Edited by Echoa - 9/28/16 at 2:05am
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
Toxic-DT
(19 items)
 
Toxic-Mini
(6 items)
 
Toxic-SV
(10 items)
 
CPUMotherboardGraphicsRAM
i7 4770k Gigabyte Z87X-UD4H Sapphire Nitro + RX 470 Gskill Ares 
Hard DriveHard DriveHard DriveOptical Drive
Mushkin Eco3 Kingdian  Corsair LE  Lite On 
CoolingOSMonitorKeyboard
Cryorig H5 Windows 10 pro Dell S2415h Azio MK Retro 
PowerCaseMouseAudio
Seasonic G Series 650w Define R5  Logitech G500s Asus Xonar DSX 
AudioAudioAudio
Monoprice Retro Heaphones Monoprice Vocal Dynamic Mic Edifier r980t  
CPUMotherboardGraphicsRAM
Rk3288 Firefly  Mali t760 mp4 Ddr3 1600 
Hard DriveOS
Internal EMMC Android 
CPUMotherboardGraphicsRAM
Xeon e5649 HP z400 FirePro v4800 12gb Gkill Ripjaw 
Hard DriveHard DriveCoolingOS
Hitachi Ultrastar Western Digital 320gb HP High Performance Cooler Ubuntu Gnome 16.10 
PowerCase
HP z400 HP z400 
  hide details  
Reply
post #1627 of 4133
Quote:
Originally Posted by greytoad View Post

@BromoL I find it interesting. In GTAV I get 2 or 3 fps more at 4k with Phase Gain set to 000000 and clocked at 1300 2100 compared to 1400 2135 set to ddd000. ...

You were right about the 2 or three degrees lower vrm temps. Even on water I get 3 or 4 degrees lower vrm temps at 000000. ...

...

Wattmon is still my goto app. Thank you. I never said thanks for the profiles. They work well. Way easier than taking screenshots of different oc settings to recreate. Also, if I reboot before every memory clock setting change and only use wattmon (with monitoring turned off) and never use wattman I don't get the stuck memory performance scaling issue. It seems even the monitoring in Wattman can cause the issue once in a while for me. Since DDU and reinstall is the only way to get my performance back, you've saved me lots of time. Thanks.

Haven't there been reports that the driver reduces power limits for unsigned bioses? Found it here.

VRM temperatures should not be taken verbatim. There is only on sensor on the PCB in the vincinity of the VRM. "ddd000" obviously should result in different temperatures for the single phases that we cant meassure. It might actually reduce temperatures for the three phases with lower load. Equal load distribution should result in a minor efficiency gain and more equal temperature distribution. Locally it can go in both directions. For high overclocks on reference cards you have do decide whether you want to put the stress on the 12V contacts of the slot or the three VRM phases that are powered from the connector (the connector itself is fine). At +0% power limit i go with 000000 most of the time, because thats my default efficiency profile. When at +30% or more i would shift some load to the 6-pin.

Thanks for your feedback on WattTool. I observed the memory scaling issue only once. I changed memory voltage (not clocks) with WattTool while Luxmark was open (but paused). The GPU was in P6 (1236MHz, not max) and memory also clocked up. Thats probably because Luxmark has OpenCL queues open even when paused. After that i only got half the performance in Luxmark, but everything was fine after a reboot. I think i had HWiNFO open at that time.
Edited by BromoL - 9/28/16 at 2:57am
i5 3470 + RX 480
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5-3470 Gigabyte H77-DS3H AMD Radeon RX 480 2 x 4GB Corsair Vengeance LP DDR3-1600 
Hard DriveHard DriveCoolingOS
128GB Samsung 830 SSD 240GB OCZ Trion 150 SSD Thermalright HR-02 Macho Rev. A (BW) Windows 10 Pro 
MonitorKeyboardPowerCase
Dell U2713HM generic Thermaltake Munic 430W Antec Three Hundred 
Mouse
Roccat Kone Pure Military 
  hide details  
Reply
i5 3470 + RX 480
(13 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5-3470 Gigabyte H77-DS3H AMD Radeon RX 480 2 x 4GB Corsair Vengeance LP DDR3-1600 
Hard DriveHard DriveCoolingOS
128GB Samsung 830 SSD 240GB OCZ Trion 150 SSD Thermalright HR-02 Macho Rev. A (BW) Windows 10 Pro 
MonitorKeyboardPowerCase
Dell U2713HM generic Thermaltake Munic 430W Antec Three Hundred 
Mouse
Roccat Kone Pure Military 
  hide details  
Reply
post #1628 of 4133
neptunex, wrong thread.
Edited by skline00 - 9/28/16 at 3:12am
post #1629 of 4133
Quote:
Originally Posted by skline00 View Post

neptunex, wrong thread.

Why?
post #1630 of 4133
Quote:
Originally Posted by Mergatroid View Post

Sorry for the delay. I have gotten home quite late over the last few days.

Here are a couple of CPUz screenshots while running Firestrike:

GPU1 (top card):




GPU2 (Bottom Card):




Those are all maximum readings during a Fire Strike run. That's a pretty big power difference between the top and bottom cards. The bottom card is getting higher voltage Vddc than the top card. I have no idea why. I have not changed any settings from default. Notice as well that the higher power card has an approximately equal temperature, yet it has a fan speed of about half the other card. This really doesn't make a lot of sense to me.

One last note here, I did a gpu driver update today and it seems like Furmark is no longer seeing my second card. I ran a log during a Heaven benchmark and the log showed nothing for the second card. I have a bad feeling this "update" screwed my system somehow. I tried reinstalling Furmark with a couple of reboots, and made sure I was running it fullscreen, but nothing seems to affect the fact it only shows one GPU. Something screwy going on....

I explain this a few days ago, Wattman not detect at the same way the voltage in the video-cards in CrossFireX - I don't know if this is a bug of the driver or a problem with the pci-express in motherboard - go to wattman check the voltage it's in "Auto" and put in Manual and see the voltages in the 7 stages - And do the same you will se probably a difference of voltage.-
Example in my case in stage 7 the first GPU have 1150 mv and the second 1100 mv - So when I overclock I need to de-sync the video card's and put a "Plus" in the second video card voltage to get the "same".-
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD/ATI
Overclock.net › Forums › Graphics Cards › AMD/ATI › [Official] Polaris Owners Club