![]() $259, 1600w AND a quality PSU? EVGA and BPC Tech - you are the perfect switch…you turn me on…ĮDIT - There seems to be concern that there's no in-depth reviews out there. I read a forum post about someone complaining about coil whine under heavy load on the 1300w model but YMMV. It's actually cheaper than the 1300w equivalent model which is on sale for $269 at BPC (still a great price) so I figured it's worth future proofing myself (crazy how I thought that getting my existing 850w Tier 1 PSU would be fine for many years when I bought it in 2021). It’s definitely weird, it feels like they bundled D4 npc’s with an integrated gpu burn test, or maybe vendor npc’s are as well crypto-miners hijacking our gpu’s Nothing fishy i guess ofc, but this looks completely abnormal.My 850w PSU wasn't quite up to the task to keep reliably powering my 7900 XTX, after a recent CPU upgrade, so I went on the hunt for a 1000-1300w PSU which led me to stumble across this absolute bargain - a quality 1600w on sale for $259 with free shipping (save $290). When i’m in the vendor/blacksmith menu, these are bumping quickly (and stays stable at the following values in my case) :Īnd i can even hear the fans starting to work harder. to avoid hammering it too much and keeping temps at bay with D4, i’m capping fps to 100 (gsync on), and using balanced DLSS, all that in 1080p, + i’m using an MSI afterburner curve for a slight undervolt. when i use my performance overlay to check power draw / frequencies / temps.Ĭurrently i’m using a Lenovo laptop with a mobile 3060. I noticed as well a strange behaviour with city npc’s, being the weapon vendor or blacksmith. As this is more likely to kill gpus than a vram leak. At least with some lines.ġ.1.1 is addressing the vram end of the issue but I’d like to get a handle on how many other gpu lines are fluctuating on specific npcs. It’s less D4 exposing average cards and more D4 exposing cards that aren’t heavily overkill. I will say that if these cards had overkill pcb components and a solid asic rating, they may have held out. In the cases of the cards I’ve dealt with the issues have been majority failed inductors, with one resistor, and 2 unknown memory related issues. Constantly switching between low and max load is extremely hard on power phasing and can cause things to fail that 3dmark or a stress test will not. For example while sitting at the blacksmith in Zarbinzet, the 3070ti I’m testing on now will constantly jump between 50% and 97% usage with vram maxed as well (for some reason the 7900xtx uses 22gb of vram as well but does not fluctuate). Specific npcs cause a nonstop load fluctuation. You can replicate this, but do so at your own risk. I build computers often so I had a decent amount of test control.ĭiablo 4 is a creating a scenario that is tougher than a 100% load on some GPUs, I’ve confirmed 6k/3k series and suspect the same goes for some others. The 6k amd and 3k nvidia gen cards seem to be specifically vulnerable to this issue. That was until the 8th gpu I worked on that was stated to have “fried from D4” (I’m at 9 now). In the past has this has always been a case of the game exposing a weak part on the pcb, inductor, resistor, chiplet etc. As someone who’s been repairing gpus for over a decade now and works at a repair shop, I’m typically the first one to speak up when people claim “x game fried my gpu”. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |