Overclock.net banner

21 - 32 of 32 Posts

·
Registered
Joined
·
762 Posts
How can you have been here for 10 years and still spout off some dumba$$ :poop: like that?
It is painful for my eyes ..... some one registered in this forums .... instead using best System memory for high performance ... to expect 20 cents resistor to do miracles.
No matter how many decades I will wait, some people will never learn neither the basics.
 

·
Registered
Joined
·
762 Posts
But if I mod the card and one of the VRAM chips takes a dirt nap, there is no reason why I shouldent be able to warranty the card as my mod had nothing to do with the fault.
BEWARE of something simple .... The rules are set by the manufacturer
READ THEM WELL ...


Graphic Card Warranty Policy

Graphic card Warranty Policy
A Limited Warranty is offered on Asus branded hardware products. This Limited Warranty does not cover any software applications or programs; non-Asus products or non-Asus branded peripherals.

Asus guarantees that the Asus Graphic Card that you have purchased is free from defects in materials or workmanship under normal use during the Limited Warranty Period. The Limited Warranty Period starts from the date of purchase. Your sales receipt, showing the date of purchase, is your purchase date proof. You may be required by your Purchase Site or Asus Authorized Service Providers to present a proof of purchase as a condition of receiving warranty service. You are entitled to a hardware warranty service if a repair is required within the Limited Warranty Period.

During the Limited Warranty Period, Asus will, at its sole discretion, repair or replace any defective component. All spare-parts or module removed under this Limited Warranty become the property of Asus.

Exclusions from this Limited Warranty Service
ASUS does not warrant uninterrupted or error-free operation of this Product. The warranty only covers technical hardware issues during the Warranty Period and in normal use conditions. It applies to firmware issues but not to any other software issues or customer induced damages or circumstances such as but not limited to:
  1. Damage caused to this Product(s) by you or any non-authorized third party.
  2. The serial number of the Product, components or accessories has been altered, cancelled or removed.
  3. Obsolescence.
  4. Damage (accidental or otherwise) to the Product that does not impact the Product’s operation and functions, such as without limitation to rust, change in color, texture or finish, wear and tear, and gradual deterioration.
  5. Damage to the Product caused by war, terrorism, fire, accident, natural disaster, intentional or accidental misuse, abuse, neglect or improper maintenance, and use under abnormal conditions.
  6. Damage to the Product caused by improper installation, improper connection or malfunction of a peripheral device such as printer, optical drive, network card, or USB device, etc.
  7. Damage to the Product caused by an external electrical fault or any accident.
  8. Damage to the Product resulting from use outside of the operation, storage parameters, or environment detailed in the User’s Manual.
  9. Damage to the Product caused by third party software or virus(es); or there is software loss or data loss that may occur during repair or replacement.
  10. Unusability due to forgotten or lost security passwords.
  11. Unusability of or damage to the Product caused by contamination with hazardous substances, diseases, vermin, or radiation.
  12. Fraud, theft, unexplained disappearance, or damages/detrimental circumstances caused by a willful act of the customer.
  13. Damage to the Product that arises from cryptocurrency mining or related activities unless the model is especially designed for mining.

Returning a Graphic Card to your Purchase Site or an Asus Authorized Service Provider during the warranty period does not automatically mean that it will be repaired free of charge. Upon receiving your product, the Service Centre reserves the right to check the validity of your warranty and your request for warranty service.

Neither Asus, Asus Service Center, Asus Authorized Service Provider , nor your reseller are responsible for any of your confidential, proprietary or personal information contained in the product which you return to Asus , Asus Service Centre or Asus Authorized Service Provider for any reason. You should remove all such information from the product prior to its return Asus.

Asus Service Centre or Asus Authorized Service Provider is not responsible for damage or loss of any personal data, programs, or removable storage media. Asus is not responsible for the restoration or reinstallation of any data or programs other than software installed by Asus when the product was manufactured.

 

·
Senior Overclocker
Joined
·
5,081 Posts
Sorry to hear about your motherboard and 8700k. Were you doing the AVX stress testing with Prime95? I hope not since that's what I use to test my 9700k!
Actually yes, it was during a P95 session. I was first doing a few runs of Ycruncher to test initial stability with AVX, then moved to try a few runs of P95 and at one point the whole PC shut down and wouldn't power on. On close inspection, the CPU had a little "belly" under one of the resistors in the middle, as if it literally blew up inside. There was no sign of damage on the mobo. I still have it in storage. No idea if it still works or not since I didn't want to test it with the new 9900k at the time.
I would advise against P95 AVX stress testing as it pushes the CPU well beyond what it will EVER draw, under ANY circumstance. I now test with Cinebench R23 in a loop and Realbench, both AVX, both pushing the CPU further than any other workload, but not to absurd loads such as Ycruncher and P95 AVX.
 
  • Rep+
Reactions: 8051

·
Registered
Joined
·
150 Posts
Discussion Starter · #24 ·
Ok I finally got my loop together after a few months of trying to get parts.
It turns out I still have some work to do as the flow is only around 110 liters per hour when it should be in the 200s.
IgorsLab noticed improvement up to that in temperature too on the non strix version of my block.

That said, water cooling already had a big benefit.
In the uni-engine heaven benchmark 4.0 I was hitting temps of 85ish on the core and 90+ on memory with hotspots over over 100 degrees.
Those are down to 50 degrees on memory and Core, with 60 degrees for hotspot. In other words I almost cut my temperatures in half, that's crazy.

And the cherry on top is then that I got roughly 130 extra Hz with the exact same settings as before.
I now hit a max of 2175Hz. I want to first try and get up to 200 liters and then check again, maybe try some undervolting because I was still hitting voltage/power limits.
But this is already pretty nice. Who knows, maybe I can crack 2200 with the extra flow, but I doubt it.
Right now, shunting is looking less interesting, although it's still scratching at the back of my head :D
 

·
Registered
Joined
·
150 Posts
Discussion Starter · #25 ·
Trying to overlock again was a bit sobering on the other hand.
In heaven benchmark I was able to run a stable 2200 core clock (needed around 1,075 V to run that with 3dmark too) and I would have been happy but I noticed the effective clock actually went down. I was getting only around 2100 despite GPU-Z not showing any limits triggering. I would get higher clocks with less frequency, not evening lowering voltage.

Then on top of that, in heaven this was completely stable and would not budge. In 3dmark the frequency would jump around.
So in the end I settled on +160 in afterburner, giving clocks between 2100-2160 fairly consistently in 3dmark and 2130-2160 in heaven but at a pretty higher 1,094V.
I didn't try going higher than 2200 because it doesn't allow me to go over 1.2V anyway and the effective clocks were going down.
Memory at 1225, 1250 got an artefact and 1300 crashes 3dmark.so that's 10975 in afterburrner, 2740 reported by HWINFO.

So it looks like I am being limited by internal stuff more than power. Which is back at these settings while not seeing it heaven. Though I got pretty close.
I doubt I would be able to get much benefit from a shuntmod at this point?
Temps stay at or below 55 degrees, so that's not an issue in the slightest. This is giving me just above 22000 graphics score in timespy.
 

·
Facepalm
Joined
·
9,966 Posts
Trying to overlock again was a bit sobering on the other hand.
In heaven benchmark I was able to run a stable 2200 core clock (needed around 1,075 V to run that with 3dmark too) and I would have been happy but I noticed the effective clock actually went down. I was getting only around 2100 despite GPU-Z not showing any limits triggering. I would get higher clocks with less frequency, not evening lowering voltage.

Then on top of that, in heaven this was completely stable and would not budge. In 3dmark the frequency would jump around.
So in the end I settled on +160 in afterburner, giving clocks between 2100-2160 fairly consistently in 3dmark and 2130-2160 in heaven but at a pretty higher 1,094V.
I didn't try going higher than 2200 because it doesn't allow me to go over 1.2V anyway and the effective clocks were going down.
Memory at 1225, 1250 got an artefact and 1300 crashes 3dmark.so that's 10975 in afterburrner, 2740 reported by HWINFO.

So it looks like I am being limited by internal stuff more than power. Which is back at these settings while not seeing it heaven. Though I got pretty close.
I doubt I would be able to get much benefit from a shuntmod at this point?
Temps stay at or below 55 degrees, so that's not an issue in the slightest. This is giving me just above 22000 graphics score in timespy.
A 5 mOhm soldered stacked shunt mod on a 3090 Strix will definitely help you maintain better clocks.
 

·
Registered
Joined
·
150 Posts
Discussion Starter · #27 · (Edited)
How much are talking though? Because I am seeing a lot of reliabilty/operating voltage in 3dmark too. that wouldn't change with a shuntmod no?
From what I can tell, Framechasers was running 2160/11000 with shuntmod.
That is basically what I am running too. That said let me try flatteningf the curve after 2160, see if that makes any difference.
 

·
Facepalm
Joined
·
9,966 Posts
How much are talking though? Because I am seeing a lot of reliabilty/operating voltage in 3dmark too. that wouldn't change with a shuntmod no?
From what I can tell, Framechasers was running 2160/11000 with shuntmod.
That is basically what I am running too. That said let me try flatteningf the curve after 2160, see if that makes any difference.
Is your strix stock and on air cooling or water?
Just flash the Strix 1000W vbios and test it for yourself. See if your effective clocks change.
I attached it. Rename it to ROM.
 

Attachments

·
Registered
Joined
·
150 Posts
Discussion Starter · #29 ·
since when is there a higher power strix specific bios Oo. I guess this a modified kingpin?
I'll give it a shot, but if thisis an option, what's the point of shunting then?
 

·
Linux Noob
Joined
·
1,182 Posts

·
Registered
Joined
·
150 Posts
Discussion Starter · #31 ·
Appernetly under the same rock that just made me realize I would need to disable the gpu to flash.
Which I can't do since I don't have any integrated graphics card. What am I missing here?
 

·
Facepalm
Joined
·
9,966 Posts
Appernetly under the same rock that just made me realize I would need to disable the gpu to flash.
Which I can't do since I don't have any integrated graphics card. What am I missing here?
You don't need to disable the GPU.
Just download the newest NVflash from techpowerup.

If the straightforward NVflash biosname.rom doesn't work, try nvflash --protect off
then nvflash biosname.rom or nvflash -6 biosname.rom
replace biosname.rom with the file you renamed. (remove the txt at the end, example: Strix_xoc.rom will work)
 
21 - 32 of 32 Posts
Top