Sajin

mylittlepwny2
Had some downtime and ran some superposition.

Generally runs between 700 - 840w depending on the bench (according to the Oled).

700 to 840 watts. 👀

Yeah the oled often reports somewhat higher values than GPUZ or actual hardware does. But for sure it averages well over 600w during 4k native benches, and definitely over 550 or so in lighter stuff like port royal.

But yeah pretty insane how leaky my particular specimen is. Absolutely drinks power. Even my leakiest 3090 doesn't hold a candle to this!

Sajin

mylittlepwny2

Sajin

mylittlepwny2
Had some downtime and ran some superposition.

Generally runs between 700 - 840w depending on the bench (according to the Oled).

700 to 840 watts. 👀

Yeah the oled often reports somewhat higher values than GPUZ or actual hardware does. But for sure it averages well over 600w during 4k native benches, and definitely over 550 or so in lighter stuff like port royal.

But yeah pretty insane how leaky my particular specimen is. Absolutely drinks power. Even my leakiest 3090 doesn't hold a candle to this!

Haha. Yeah. My 3090 Ti FTW3 pulled 628w according to hwinfo in metro exodus benchmark.

Was that average consumption or just peak? Either way you definitely got a leaky specimen as well. And judging by your scores you got a God bin die. Most chips pull in the mid 400s in 4k applications. The God bin chips pull more.

https://www.3dmark.com/pr/1662368

Was working on efficiency. This is the tedious part of benchmarking. Alot of time spent for small gains!

Great score! that card should go above 17,000 if you can find away to cool it .

mylittlepwny2

telehog
Great score! that card should go above 17,000 if you can find away to cool it .

Yeah thats the issue though. If the memory gets below about 35C it starts to lose clocks and efficiency badly. You better be able to heat the memory chips (to preserve as much as you can) while simultaneously freezing the core to try and brute force higher scores!

On plus side still making slight gains!

https://www.3dmark.com/pr/1662861

Nice, now lets see some higher scores on time spy extreme.

newls1
any update on a waterblock release for the FTW3 3090Ti?  EK delayed it once again, and I dont think I want to put on a bykski block on this beautiful card.  Anyone know of a major brand name releasing a block for our cards?

Mine will be here Friday

kraade

newls1
any update on a waterblock release for the FTW3 3090Ti?  EK delayed it once again, and I dont think I want to put on a bykski block on this beautiful card.  Anyone know of a major brand name releasing a block for our cards?

Mine will be here Friday

how did you get the EK Block?? I even called PPCS and even they said it was delayed AGAIN till Mid August.  Im PISSED! but happy for you though.  My loop has been all assembled and just waiting for this damn block.  You think If I just "pre order" it on EK's site it might ship out BEFORE august 18Th like it shows?  Im dying to get this damn block already!!

talon951
Yea some people have already gotten the EK block. I'm interested to see if they have improved their block delta with this new line. The previous FTW3 block seemed to be really poor performing.

Bykski blocks are pretty good. Just complete lack of instructions, so not friendly to inexperienced people. I've seen a lot of people report good results with them. (admittedly I haven't tried one yet)

It's got to be better than the HC 

AIDA64 just updated gpu details for RTX 4090,so if RTX 4090 is coming very soon,maybe evga won't produce more 3090TI KINGPIN,and also hydro copper for 3090TI.
The only way to get a KPE 3090TI seems to be buying the limited E1.But E1 should be more overpriced.

Vince said on Facebook that there were a lot of supply chain issues with the production of the kingpin 3090 Ti so we're lucky we got what we did

mylittlepwny2

tresnugget
Vince said on Facebook that there were a lot of supply chain issues with the production of the kingpin 3090 Ti so we're lucky we got what we did

How's ur card doing? Have you been benching much? Did you get xoc bios and classified tool?

I haven't for a couple days. I got the XOC bios and classified which helped a little. The core is very solid but the vram not so much
I'm getting like 15.8k runs in Port Royal but I feel like I should be getting higher. I'm running at 2250 mhz with an average of 2222 mhz and my VRAM is at +700, +800 with 1.4v on fbvdd. Im going to install and dual boot windows 10 tonight, run as few background processes as possible, and see if that helps. Maybe play with drivers. I feel with what I have I should be getting into 16k at least but maybe it's just the vram holding me back. Also going to tack on 3 noctuas if I can find a fan splitter around here. Pretty sure I've got one laying around just gotta find it lol.

talon951
This stripped down Win10 some of us use works the best to maximize benchmarks. Increases cpu benches the most obviously but still can help gpu benches as well.

https://www.overclock.net...age-1020#post-28974928

Nice, I was looking for something exactly like this! I really appreciate it!

talon951

tresnugget

talon951
This stripped down Win10 some of us use works the best to maximize benchmarks. Increases cpu benches the most obviously but still can help gpu benches as well.

https://www.overclock.net...age-1020#post-28974928

Nice, I was looking for something exactly like this! I really appreciate it!

You might look through KedarWolf's posts after that one too. Also posted a script more recently that disables even more services. I didn't find it when I searched a bit ago, but it's there somewhere in that thread.

Yeah looks like the iso link he posted doesn't work anymore. I'm going to go through his posts and see if I can find a better link but if I can't I'll message him and see if he can send me one

It depends on the state of the OS you're currently benching on. But even comparing a fresh install to a stripped OS, I would always get a significantly higher cpu score in TS (like maybe 200 pts). Graphics score would be higher than my daily driver OS (but usually not compared to a fresh install).

talon951
It depends on the state of the OS you're currently benching on. But even comparing a fresh install to a stripped OS, I would always get a significantly higher cpu score in TS (like maybe 200 pts). Graphics score would be higher than my daily driver OS (but usually not compared to a fresh install).

To be fair modern CPUs score well in excess of 20k in timespy lol. 200 points is like 1% which is within margin of error. Again the stripped OS does help im not arguing that. And i use one myself. But its just important to keep things in context. That's all I was trying to say.