EVGA GeForce RTX 3060 XC

[Luminescent]

Wonder how far this insanity will go, we could be looking at 2000$ for a RTX 4060 in the future, and top cards will go for 5000$.
I don't see them giving this up, why not strike a deal between AMD, Nvidia and possibly Intel to keep this shortage of silicon indefinitely but only for people who want to build their own computer, they can't fool Sony, Microsoft, Apple .....
What can the average user do except pay whatever they charge for the privilege to have a desktop PC, maybe they can make ram expensive again, why not network cards or soundcards, i could even see a shortage of plastic, why is plastic so cheap!

 

[neatfeatguy]

Not sure if what I did was more practical, but sitting on $600 in hopes of picking up a new, decent GPU to replace my 980Ti doesn't ever seem like it'll come to fruition. So, the iPad mini (4th gen, I think it's what the wife and kids use), the screen is starting to crap out on it. The thing has been used and abused for the past 5 or so years and about 1/5th of the screen flickers and has black lines running through it.

I took $400 of my GPU cash and got the wife a new iPad since she uses the old one all the time. Let the kids enjoy the last of the old one until it finally dies and the wife can have hers. At least someone got something they enjoy out of this lack of GPU situation we're in.

 

[Jaywalking]

why there are two reviews about 3060 without any "Performance per watt"?

 

[W1zzard]

why there are two reviews about 3060 without any "Performance per watt"?

we have four RTX 3060 reviews, the section is now called "Efficiency"

 

[cst1992]

Yeah good luck finding an 3060Ti @ 400-500usd, they are all selling for 1000usd now. If you can find an 3060 @ 400-450usd it's actually a good deal .

I used to think we always get a pretty bad deal here in India(a retailer had the original GTX Titan for INR 89,999 - $1,343 - back in the Kepler era, whereas its actual price was $999).
Back in December, I spotted the 3060Ti reference for INR 35,900 - $491. I'd never ordered such an expensive card to date but still I got it, and now it's actually a better deal than in the US or Europe.
Aftermarket cards were available for INR 40,000 - 48,000 ($547 - $657).
All are sold out now.

 

[Jaywalking]

we have four RTX 3060 reviews, the section is now called "Efficiency"

such a pity though, used to have 3 efficiency at different resolution...

 

[W1zzard]

such a pity though, used to have 3 efficiency at different resolution...

explained here https://www.techpowerup.com/forums/threads/performance-per-watt-changes.278988/#post-4466294

 

[Dirtdog]

Why is the VSync 60Hz power so high, like +60w over 2070.

"Board partners tell me that they would be selling at a loss, because just the production cost is that high, before any salaries are paid, or activities like support and marketing are taken into account." Used car salespeople say the same thing ...

I am wondering that too. My 1660 is shown as using 110w but the 3060 uses 176w? And yet in the 'efficiency' graph the 3060 is said to be more efficient? How does that work.

Anyway, with the insane prices of GPUs I see me hanging onto the 1660 for another year or two at this rate. For 1080p it's still good enough and if gaming on PC becomes permanently expensive I will simply stop doing it and either get a console or play my old games. I refuse to feed the greed of companies like nvidia any more.

 

[W1zzard]

I am wondering that too. My 1660 is shown as using 110w but the 3060 uses 176w? And yet in the 'efficiency' graph the 3060 is said to be more efficient? How does that work.

The efficiency chart is based on the normal game power draw, not v-sync, so 181 W and 61.7 FPS.

V-Sync on the GTX 1660 achieved only 54.5 FPS, not 60 like the 3060, so that affects the V-Sync power number, too

 

[Dirtdog]

The efficiency chart is based on the normal game power draw, not v-sync, so 181 W and 61.7 FPS.

V-Sync on the GTX 1660 achieved only 54.5 FPS, not 60 like the 3060, so that affects the V-Sync power number, too

Maybe you should use a game which all cards in the test can achieve 60fps at, even if it's an old one? I don't know. Presumably you used a GPU-limited game which means the 1660 is working flat out, and yet it still uses vastly less power than the 3060 which presumably is working far from flat out.

I think if it was me I would see that graph and realise there was something amiss with my testing methodology or something because it can't be right, it defies logic. Or put it this way, the numbers might be correct but they give a misleading impression. I doubt that all things being equal, a game which runs at 60fps on my 1660 can use 60% more power on a more modern and more efficient card to achieve the same performance. That can't possibly be explained by the fact in this particular test it is running at 54 vs 60, it's only 10%.

 

[W1zzard]

I'm using Cyberpunk 2077. If I pick something that reaches 60 FPS on GTX 1660, it will be CPU limited on 2080 Ti and up

 

[Dirtdog]

Is there something else that is causing the 3060 to work harder I wonder, such as maybe running ray-tracing which the 1660 doesn't do? That's the only sort of thing I could think which would make sense, but I assume you would have thought of that originally and made sure all the cards were doing the same work otherwise the results would be apples and oranges.

 

[bug]

I'm using Cyberpunk 2077. If I pick something that reaches 60 FPS on GTX 1660, it will be CPU limited on 2080 Ti and up

Plus, you really want something that exercises modern feature, you wouldn't want to estimate power draw using a game that only does pixel shader 2.0 or so.

 

[W1zzard]

Is there something else that is causing the 3060 to work harder I wonder, such as maybe running ray-tracing which the 1660 doesn't do? That's the only sort of thing I could think which would make sense, but I assume you would have thought of that originally and made sure all the cards were doing the same work otherwise the results would be apples and oranges.

Nope, it's all the same settings and everything. This looks to be a peculiarity of the new GPU, RTX 3060 Ti and 3070 go to lower power states just fine

 

[xorbe]

Okay wow that's misleading at best. So the card that doesn't reach 60 perhaps runs at 30 with vsync ... drastically cutting the power. Aight, I'll skip that useless chart in the future, heh.

 

[W1zzard]

Okay wow that's misleading at best. So the card that doesn't reach 60 perhaps runs at 30 with vsync ... drastically cutting the power. Aight, I'll skip that useless chart in the future, heh.

It's actually a 60 FPS cap, so weak cards run as fast as they can, not 60/30/20. "V-Sync" is part of the title to help people understand the usage model

VSync: 110.3 W
V-Sync FPS below 60: 54.5 FPS

 

[cst1992]

It's actually a 60 FPS cap, so weak cards run as fast as they can, not 60/30/20. "V-Sync" is part of the title to help people understand the usage model

Is it worth it to include a separate chart for that, though? I'm guessing in GPU limited games it's the same as regular efficiency.

 

[xorbe]

It's actually a 60 FPS cap, so weak cards run as fast as they can, not 60/30/20. "V-Sync" is part of the title to help people understand the usage model

VSync: 110.3 W
V-Sync FPS below 60: 54.5 FPS

Why not just label the chart correctly "60 fps cap", instead of incorrectly "v-sync".

 

[W1zzard]

Is it worth it to include a separate chart for that, though? I'm guessing in GPU limited games it's the same as regular efficiency.

Every card except for the slowest bunch will be at exactly 60 FPS, so no point really

Why not just label the chart correctly "60 fps cap", instead of incorrectly "v-sync".

To make it easier for the less experienced readers to understand the reasoning behind this test. I can add a note in the testing details though that this is not 60/30/20, but as close to 60 as possible, good idea

 

[bug]

To make it easier for the less experienced readers to understand the reasoning behind this test. I can add a note in the testing details though that this is not 60/30/20, but as close to 60 as possible, good idea

As long as you don't change your benchmarking monitor with a high-refresh one

 

[Sp33d Junki3]

I got one which I'll take any GPU right now from my GTX1060.
Price is not nice, $529 CAD + Tax. It was the cheapest one.
Value wise, not good. Performance great with my i5 10400F. I can actually play Cyberpunk 2077 (very High/Ultra adjusted settings) at 2560x1440 55-60fps and many other games Like Outer Worlds and Outriders (maxed 60fps)
Happy I got a GPU at least.

 

[neatfeatguy]

Figure a good spot to post.

I got my chance to purchase this card from EVGA. Using the associate's code to save 3% and cost of shipping, card ran me $400. I should hopefully see it within a few days.

I didn't want to settle for a 3060, but my 980Ti is still having high fan spin ups. Some days I may not get one and others it could happen a dozen times or more. I have a gentle fan curve setup on MSI Afterburner. The fan speed should slowly ramp up, but the fans sometimes just go crazy fast for 3-4 seconds and then slow back down. When the fans spike there is no indication of increased fan speed on GPUZ or MSI Afterburner and there are no spikes for power or temps.....

Guess I'll have to do with this card over the 3070 or 6800 I originally was hoping to get. But I refuse to spend $800+ on one of those cards from a retailer and my wait on the list for EVGA isn't moving for those cards where I'd see one anytime soon.

I'm content I finally found something new, just wasn't what I was looking for. I'll just keep on gaming and enjoy what I have.

 

[cst1992]

I got my chance to purchase this card from EVGA. Using the associate's code to save 3% and cost of shipping, card ran me $400.

Not a bad deal, considering the current climate.

And EVGA make good cards.

 

[TechnicalPiglin]

It seems buying the 3060 is akin to buying a 2070 when you can get a 2080 Super for almost the same price. Decent performance but extremmmmmmmmely bad value. This card should really cost $279.

 

[SSGBryan]

You will never see this card new for $279 - those days are gone.

I am happy I got this model for $369 - the 12Gb buffer is excellent for 3d rendering, and I suspect I'll be gaming at 1440p on high, and that is what I am after.