Overclock.net › Forums › Graphics Cards › AMD/ATI › Polaris Bios Editing ( RX5xx / RX4xx )
New Posts  All Forums:Forum Nav:

Polaris Bios Editing ( RX5xx / RX4xx ) - Page 282

post #2811 of 3317
Quote:
Originally Posted by iakoboss7 View Post

what tool is that?
memtestCL. it was suggested to use by someone like 10 pages ago
post #2812 of 3317
Quote:
Originally Posted by Vento041 View Post

MemCL always find errors on that test, use HWiNFO EDC error counter and stress test your VRAM with benchmark and games (careful, you must use both, since they stress VRAM in different ways)

Why the hyperbole?
Roided midgit
(10 items)
 
Midgit
(12 items)
 
 
CPUMotherboardGraphicsRAM
3770k Asus P8Z77I-Deluxe MSI Radeon RX 480 GAMING X Crucial Ballistix XT 
Hard DriveCoolingCoolingCooling
Crucial M500 Noctua NH-D14 Noctua NF-F12 PWM Noctua NF-F12 PWM 
CoolingCoolingPowerCase
Noctua NF-B9 PWM Noctua NF-B9 PWM Corsair SF600 Fractal Node 304 
  hide details  
Reply
Roided midgit
(10 items)
 
Midgit
(12 items)
 
 
CPUMotherboardGraphicsRAM
3770k Asus P8Z77I-Deluxe MSI Radeon RX 480 GAMING X Crucial Ballistix XT 
Hard DriveCoolingCoolingCooling
Crucial M500 Noctua NH-D14 Noctua NF-F12 PWM Noctua NF-F12 PWM 
CoolingCoolingPowerCase
Noctua NF-B9 PWM Noctua NF-B9 PWM Corsair SF600 Fractal Node 304 
  hide details  
Reply
post #2813 of 3317
Anyone know how can i identify in .rom what card is it , like is it 470 480 570 or 580 ?
Is there a unique identifier in all bioses for this ?
I couldn't find it.

Polaris 10 can be 470 or 480
Polaris 20 can be 570 or 580
post #2814 of 3317
should i take the errors serious that I get from that test? HWiNFO shows 0 on stock 2000mhz while benchmarking and games and during that memtestcl too.
post #2815 of 3317
No, memtestcl is stable for me, and if I go past 2020 on uber v3.1 I get errors. Core speed can also affect memtestcl, because it does basic calculations of memory.

The hwinfo errors number is what the card knows is wrong internally. Like artifacts can show without a crash, the undetected errors can still pass through with any instability.

That's why I only trust occt with error counting and memtestcl to determine my safe overclock. Like I would use prime95 and Intel burn in test for a cpu/ram/motherboard test.

Others disagree, but your choice. I don't mine but if you did you could see that some overclock that is thought to be stable makes corrupt hashes, just like using a proper test that checks for errors
post #2816 of 3317
Usually I'll do like you guys said but... That test behave too weird. It's really strange that doesn't matter how many iterations or how much is been memory allocated, if I get errors I get them only there... Like thousand there and 0 everywhere else...
post #2817 of 3317
MSI Gaming X 4G (4GB) elpida+hynix
https://www.techpowerup.com/vgabios/191759/191759

Looks like the waiting is over, i didn't tried it yet.
2500K RE v5
(15 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5-2500k ASRock Z77 Extreme4 MSI RX 480 Gaming X G.Skill TridentX F3-2133C9D-16GTX 
Hard DriveHard DriveCoolingOS
Samsung 840 Pro 128GB Samsung Spinpoint F3 1TB Enermax ETS-T40-TB Windows 10 
MonitorKeyboardPowerCase
Viewsonic VX2253mh-LED Logitech G910v2 Orion Spectrum Corsair Professional HX650 Corsair Carbide Series® 500R 
MouseMouse PadOther
Mionix Avior 7000 HyperX Fury Pro ASUS Cerberus 
  hide details  
Reply
2500K RE v5
(15 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5-2500k ASRock Z77 Extreme4 MSI RX 480 Gaming X G.Skill TridentX F3-2133C9D-16GTX 
Hard DriveHard DriveCoolingOS
Samsung 840 Pro 128GB Samsung Spinpoint F3 1TB Enermax ETS-T40-TB Windows 10 
MonitorKeyboardPowerCase
Viewsonic VX2253mh-LED Logitech G910v2 Orion Spectrum Corsair Professional HX650 Corsair Carbide Series® 500R 
MouseMouse PadOther
Mionix Avior 7000 HyperX Fury Pro ASUS Cerberus 
  hide details  
Reply
post #2818 of 3317
Quote:
Originally Posted by Vento041 View Post

Usually I'll do like you guys said but... That test behave too weird. It's really strange that doesn't matter how many iterations or how much is been memory allocated, if I get errors I get them only there... Like thousand there and 0 everywhere else...


Sorry, I forgot to mention that don't run above 1500 MB test. For some reason windows will bug it up, probably because desktop is hardware accelerated and it swaps out the memory.

1000 mb is good.
Replace 0 with your card to test. Mine is only GPU so 0.

memtestcl -g 0 1000 99999999
Edited by robnitro - 5/4/17 at 3:37pm
post #2819 of 3317
Quote:
Originally Posted by robnitro View Post

Sorry, I forgot to mention that don't run above 1500 MB test. For some reason windows will bug it up, probably because desktop is hardware accelerated and it swaps out the memory.

1000 mb is good.
Replace 0 with your card to test. Mine is only GPU so 0.

memtestcl -g 0 1000 99999999

Lmao Ok now I get it xD I didn't know that thank you for point that out biggrin.gif! REP+
post #2820 of 3317
Quote:
Originally Posted by mynm View Post

Ok, but do you see VDDCI changing at hwinfo?. In my 380 wattman isn't changing VDDCI, only the minimum VDDC VID when memory is at the highest clock, like BromoL says.

I dont trust the software tools to report anything correct other than core voltage, which is why I double check everything on the cards physicaly with a multimeter. hwinfo is probably just reading the PPTable , there is no VDDCI monitoring hardware on Tonga/polaris so there is no way its reading the actual value. Only very few cards have VDDCI on the second loop of the controller.

BTW if you want to check Core/VDDCI physicaly, you can just use the capacitors on the back of the card...the ones clustered in the middle are core voltage, and there are usually 4 capacitors surrounding the central banks which are for VDDCI (youll be able to tell which when you look at it). Just make sure you dont short anything with your multimeter probe!
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: AMD/ATI
Overclock.net › Forums › Graphics Cards › AMD/ATI › Polaris Bios Editing ( RX5xx / RX4xx )