Alright. The 1 card is in. I am so peeved. I was going to put 16 gb of matching ram in and forgot. The 1 card is up.
Dare me. I'll try 2 on this psu. Regular clock.
Alright. The 1 card is in. I am so peeved. I was going to put 16 gb of matching ram in and forgot. The 1 card is up.
Dare me. I'll try 2 on this psu. Regular clock.
Yeah they hate me already.
Good stress test, bad frame rates. The power maxxed only seems to draw 180 maybe 195. GPU draw is about 180. VDDC stays at 5-8. I didn't think this card would draw that much. VDDCI draws about 13 as well.
New power supply. I didn't want to do all this.
On a plus side? These pcie' are 3.0....Lucky I guess.
Board specs say 2.0
I think I overkilled it.
Well, I did warn you. :) Now at least you can add the extra RAM, too. :D
Ok. Good and bad. Amazingly the psu handles both cards pulling about 300w (more than I OP'd) per card. I think my 480's ran better. I am having low frame rates after numerous drivers. I also cannot give them more power due to the constraints.
Help??
And the extra light wasn't needed. Just my own personal complaint.
No darkness here photolady.
Ram goes in as well.
I don't know what to tell you about frame rates. But I've found that parts I buy at Amazon are sometimes, not first runs but perhaps being sold as new when they're actually products that have been returned. I've gotten motherboards there that died right out of the box, Video cards that go bad within a few months, to less than that. I've stopped buying computer parts from them, and now buy from Newegg. And I make sure they don't come from any third party seller, too.
Last one I think.
The cards temps seem to stick around 80c under load. Is that normal on these or should I do the obvious (reseat heat sinks)? Pretty sure I've got some NT-H1 around.
photolady. These were bought used. I was ok with that considering paying 80 usd per card.
For RX 480 cards, 80 C at full load would be about right. RX 580 cards should probably be more like 69 C at full load.
Ugh. Okay I resat both heatsinks. My temps dropped in a big way. Under stress loads I am keeping both cards at about 64c. I also ended up getting better frame rates by disabling ram interleaving. (odd). I understand what the channel inerleaving is for but I didn't expect it to effect the video. Stranger things? No crashes yet. I just may be happy. Also I seem to only experience a 2-4% bottleneck from the cpu. I think I have pushed this rig pretty far.
On a side note: My OC is great. 5.1 GHz and air cooled in ambient of 21c.
I'll post back. And they are ROG cards. I guess I did make out buying them.
Nevermind. Shadow of the tomb raider fixed that. Temps ramped back up to 80c. Hard crash too. Power draw? I maxxed at 200w and 185 vddc. And to think that this thing started as just a video server.
Oh that was a single card run.
Power supply change and my frame rates are up. I guess it did matter that much. I was getting random web pages locking up too.
2 cards are running healthy now and I'm running x-fire when I can. I haven't had an issue since (after a couple of days of testing).
I think I got it now.
Were the higher temps due to lack of power?
I don't know about lack of power and higher temps but it could have been, since it was lagging it was trying to pull more power than the psu had to give.Quote:
Originally Posted by Ashalon1
Power supplies matter a lot more than we think they do. That's why we buy them in a higher power than is needed than is usually recommended.
Is it ok to undervolt these differently? On that same deal, these things just run hot so I added some more cooling which seems to help. CPU temps are stable at about 31c. Since I am always trying to get a higher clock as well I pose another question. Would it be safe to add a NO type electrical button to the cmos jumper so I don't have to constantly open my case?
CMOS reset switch options:
https://www.techspot.com/community/t...-switch.68155/