![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
yeah, that is just crazy to me. My 1080 ti would start heating up my room a few degrees, and as of now I have a 4.8ghz FX-8300 and a gtx 680 under my desk (A system I fiddle around with for S&Gs) and that creates a little aura of heat. A pocket, as you will. But ugh, 600w? thats like another 150-250w on top of that rig in the GPU alone!
Yep, had FX as well 5ghz, 250w plus. Then two Radeon 290’s in cfx. Drawing more power then the 5090 system now when running at max. Then 2x 1070Ti’s then 2x Vega’s short period, before going to Ryzen. That is if I remember right. Anyways have a 18000 BTU/hr A/C unit in room, room wired for 220v as well.
I'm not afraid at all. I ran my 4090 for 2 1/2 years with no issue and expect the same from my current 5090. Just plug the damn thing in properly and don't put pressure/bend the cable too much where the plug is.
Like i did with my 4090. Let´s hope it is enough.
Totally agree with this. A lot of problems going back to the 4090 to now is user error whether people want accept that or not is up to them. Some instances that I have seen failures on are people proving common sense isn’t so common.
Well let’s start at the beginning. It’s a design fault before a user error. And Nvidia really doesn’t care since they’ve been doing this since the 4 series.
it just pisses me off that they tried to reinvent the wheel, it clearly failed, and yet, fthey still try to force the shitty design onto everyone! Admit defeat, nvidia, please, it'll be better for everyone.
Poor design I’ll agree with, but still a very low percentage of failures.
My understanding is that this happened when they started chasing the smallest PCB size possible, on newer, gigantic GPUs. But yeah it sucks that they never owned up to it. Will my 4090 end up failing like that before the 6000 cards release? Impossible to be sure but sadly a very realistic and daunting possibility...
would it kill them to add the few mm for a higher gauge and ol' reliable dual, dual 8 pins (wow, 600w is a lot more insane sounding). But, my gtx 680 has the 8+6 pin stacked. I wonder if they could do sonething similar?
Plug cable in so I don’t see any yellow. MSI cables are extremely tight, especially in the power supply side (This MSI PS has two 2x6 plugs on the cable)
Where did that number come from?
guessing ai overview in google shows 0.05%
Nope. I bought a brand new PSU to go along with my 5090 upgrade just to be safe. I'm using the OEM 12V-2x6 cable that came with the new PSU and made sure both ends are plugged in fully and securely and not bent or pulled out of place after cable management.
That's a very interesting question, although I suspect in most cases the fans etc. are on a separate 'rail' so to speak and are in addition to whatever the GPU side draws. I mean it would effectively be like if they were connected directly to the power slot pins.
I don't think it matters where it would draw that power from. The fans/rgb don't take much wattage at all to run.
I've used 4 different 5090s now - no difference performance-wise. Currently using the OLED and blinky lights RGB MASTER ICE and it performs better than the other 3 so far (just silicon lottery in general and of course they're all within a few percentage points of eachother - MSI Gaming OC, ASUS TUF OC and Astral OC).
Except that's easily margin of error. I mean it's cool to do - but I am guessing the PCIe slot power is used for the fans and all of that while the 12VHPWR is for the cards TGP.
I was in the same boat and was against purchasing one, but I'm glad I did also. Enjoy your card..
If you have to enable 3x frame gen to use DLAA I would assume your base frame rate is around 40FPS? Doesn't the low response times bother you? I agree that DLAA is great and I also think the 5090 made a substantial difference in many titles that I own but I pretty much stay away from 3X and 4X frame gen and would only use it if I was getting 60 base FPS and was using it to push frame rates beyond 120FPS on a 240hz monitor. Cyberpunk is a good example, where I can push DLSS quality mode with 2x frame gen and get a solid 60FPSX2 or 120FPS with frame gen. Turning on DLAA makes the game run at 40FPS or lower and 3x and 4x frame gen can push it up to 120FPS but it looks much more blurry and has noticeable artifacts and a lot of noticeable input lag. The difference in latency is very noticeable and feels laggy but feels and looks flawless with 2x frame gen with DLSS quality mode. I could see using DLSS quality mode at 60fps and then using 4x frame gen to get 240FPS on a 240hz monitor but otherwise I don't agree with you on this point and I think it makes the experience worse for both response times and graphics quality. Cyberpunk is the only game I own that I can't push at 4k with DLAA and 2x FG. Many games can run at 2.5DLDSR at 4k with DLAA using only 2x FG. Pretty impressive card either way.
