EVGA GeForce RTX 3090 to Feature 2.1 GHz Boost Clock?

[Tsing]

Image: EVGA



NVIDIA recently updated its GeForce RTX 30 Series specifications chart with boost clock figures, and they seem to be pretty consistent with what its AIC partners are offering with their third-party models – advanced, overclocked models included. Palit’s GeForce RTX 3090 GamingPro OC, for instance, features a 1.73 GHz boost clock, which is barely higher than the GeForce RTX 3090 Founder Edition’s spec (1.70 MHz).



This has gotten enthusiasts wondering whether there’s any serious overclocking potential with the GeForce RTX 30 Series, but things are looking up.



CyberPunkCat has shared a screenshot from EVGA’s recent GeForce RTX 30 Series iCX3...

Continue reading...

 

[Uvilla]

For each Nvidia leak, I only wonder where the hell are my reliable AMD leaks...
Holy mother Lisa is running that ship so tight... Have there EVER been good leaks from NAVI?, it all reads like bull, even from those with " reliable before" sources. I think all AMD is doing is a campaign of miss information, as a consequence I think Nvidia went all out on the attack, (which is a good thing for us) but I don't think Nvidia has much on the tank other than an in between 3080 and 3090, and some reshuffling of the same arc.
If I was guessing 'big navi' will be short lived and will be followed soon by a mcm configuration. AMD is too close to this, for them no to have this cooking already.

 

[kcthebrewer]

Really the only thing that can be assumed about AMD's new GPUs is their stated 50% performance per watt improvement.

Unless they are underselling RDNA2 to their investors (I highly doubt it), performance can be inferred from that.

The only unknowns are how well it scales and how much power it draws.

 

[Brian_B]

Really the only thing that can be assumed about AMD's new GPUs is their stated 50% performance per watt improvement.

When was it, back in February? When the “nVidia Killer” rumor started floating around, I think there my be some credence to that - AMD probably felt they could readily beat Turing. In hind sight though, that may seem foolish, since Navi2 didn’t beat Ampere to market (at least the announcement, due to GDDR6X I still have my doubts on widespread availability), and it really does seem like that AMD rumor pushed nVidia to really push their card TDPs up significantly to push more performance to release something that AMD wouldn’t have a hope of catching. nVidia won’t be caught with their pants down, unlike Intel.

With that in mind, at least with rasterizing gaming performance, I expect Navi2 to beat a 2080Ti, and be on average somewhat better than a 3070, but I am not holding my breath on it trading blows with a 3080, let alone the 3090.

 

[Denpepe]

Well I seen Steve from gamers nexus his vid about steams latest hardware survey and the amount of Turing cards in there are pretty low which I'm pretty sure means they sold pretty badly as the perfomance uplift with the previous generation was pretty bad (which can be considered an intel like move ie incremental improvement, older node)

 

[Armenius]

The Boost clock number is an average over a period of time, if I'm not mistaken, which is still on the conservative side. But people always have a habit of reporting peak Boost before the card has been heat soaked, which is the number they mistakenly go by and expect.

 

[MadMummy76]

The Boost clock number is an average over a period of time, if I'm not mistaken, which is still on the conservative side. But people always have a habit of reporting peak Boost before the card has been heat soaked, which is the number they mistakenly go by and expect.

Yeah, that's why to my surprise I often see clocks over 2000Mhz on my card, despite the boost clock being 1740.