Wednesday, March 15th 2023
Reference NVIDIA GeForce RTX 4070 Rumored to Launch at $749
According to the latest rumor coming from Moore's Law is Dead, the upcoming NVIDIA GeForce RTX 4070 could have a higher MSRP than expected, and could launch at $749. Yesterday, another rumor pointed out that NVIDIA is splitting the reviews for the GeForce RTX 4070 on two dates, with reference models that are sticking to MSRP going out on April 12th, and custom, higher-priced versions, a day later, on April 13th. The same date, April 13th, is when the GeForce RTX 4070 should hit retail shelves.
According to Moore's Law is Dead, the MSRP for reference versions will be set at $749.99, which is a bit more than earlier expected, and just $50 less than the launch MSRP for the GeForce RTX 4070 Ti. The custom versions, or "premium" as some now called them, could start at $799.99.The NVIDIA GeForce RTX 4070 is also based on the AD104 GPU, as the RTX 4070 Ti, but has 46 streaming multiprocessors (SMs), leaving it with 5,888 CUDA cores enabled. On the other hand, it is rumored to come with the same memory configuration as the GeForce RTX 4070 Ti, featuring 12 GB of 21 Gbps GDDR6X memory on a 192-bit memory interface. The TGP is set at 200 W.
The performance is expected to land somewhere between the RTX 3080 Ti and the RTX 3080, or pretty much in line with the RX 6900 XT. Of course, a 50% bump in the launch price compared to the predecessor, the RTX 3070, leaves a rather sour taste.
Sources:
Moore's Law is Dead (Youtube), via Wccftech
According to Moore's Law is Dead, the MSRP for reference versions will be set at $749.99, which is a bit more than earlier expected, and just $50 less than the launch MSRP for the GeForce RTX 4070 Ti. The custom versions, or "premium" as some now called them, could start at $799.99.The NVIDIA GeForce RTX 4070 is also based on the AD104 GPU, as the RTX 4070 Ti, but has 46 streaming multiprocessors (SMs), leaving it with 5,888 CUDA cores enabled. On the other hand, it is rumored to come with the same memory configuration as the GeForce RTX 4070 Ti, featuring 12 GB of 21 Gbps GDDR6X memory on a 192-bit memory interface. The TGP is set at 200 W.
The performance is expected to land somewhere between the RTX 3080 Ti and the RTX 3080, or pretty much in line with the RX 6900 XT. Of course, a 50% bump in the launch price compared to the predecessor, the RTX 3070, leaves a rather sour taste.
191 Comments on Reference NVIDIA GeForce RTX 4070 Rumored to Launch at $749
It didn't go well. Black screen.
Installed older driver, yes screen. yay!
He's returning the card and buying NVIDIA.
Sorry, khmm
ngreedia
My biggest issue with Amd though is they just haven't offered the performance I wanted more so than the drivers being bad.
You made my day.
I owned lots of Nvidia cards dating back from 4200Ti
I never had problems with any of them,
but the other day when I saw that black screen I was truly amazed
My friend was scared though, after all, he paid for the card.
You're so great.
You should be proud of yourself.
Rolls eyes.
What's that? Rolls royce?
I am not into modern tech kiddish types of typing.
Could you not use an emoji?
GPP didn't die, Nvidia just implemented it on the down-low. ASUS ROG 7000 series cards? Nope, Nvidia only. 7900 XTX Master? Sorry, also Nvidia exclusive. 7900 XTX Suprim? Also Nvidia exclusive.
All 3 brand suspiciously stopped offering their top SKUs on AMD starting with the 7000 series.
Nvidia has a huge grip on the market. If only AMD weren't so greedy and priced their GPUs like they priced first gen Ryzen. They seem to be blind to how they achieved their own success.
Still, I agree that they could do much better.
They got half of it right, when they set the strategy for this generation to make a GPU with very good price performance via chiplets.
But when they saw that the performance off by 25% from their target, they kept the same stupid naming and didn't adjust the pricing enough, they effed up bad. Still, it's not too late to keep dropping prices, they can afford it. But they totally missed the opportunity to have good day 1 reviews, which is what normal users read at most, if anything.
A pretty neat feature to help market the cards as being so much better over the previous generation; I find it disingenuous. It doesn't anger me. I do, however, find it humorous that people are okay with the feature or even praise Nvidia for it.
FG wasn't even on my radar as a reason to buy Ada. I still think RT is meh and up-scaling should be burned with fire, but in my experience with FG, it's a feature that has an actual use - higher visible frame rate with lower power consumption and lower CPU requirements.
I guess I don't see what's humorous seeing that as a useful feature. :shrug:
I usually play Fortnite in Performance mode (very low graphical quality) because I want to be at 165FPS locked with my 6900XT for maximum responsiveness.
Today I tested it with epic Unreal Engine settings, which means 40FPS, just to see what my new OLED monitor can do. Well, I didn't mind the 40FPPS at all visually, just the lack of responsiveness in fights. Others, like Steve from Harbor Boxed, say they can't go under 120FPS. I think it's something very personal, I don't mind 40FPS in games at all. I hate 60FPS movies though, I find they lose all theatrical effect, I will get it back to 30 or 24frames ASAP.
So I think you may both be right.
Look at my system specs and the machine I'm just about to finish building. Only my CPU needs to be cleared from customs, everything else is already in hand. Yet even I, an enthusiast who has the resources to build a showcase system like that, can't really spare the money for a 4090. It's too much! What am I supposed to do, sidegrade to a 4070 Ti? Nah, I'm just not buying.