- Joined
- Jan 20, 2004
- Messages
- 34,822
- Reaction score
- 27,884
The reason is likely they are using the slower DRAM found in the 12 gigabyte updated 3060 board. The reason why the changed the 3060 board from 8 gig to 12 gig was because of ram shortages and they knew the 3060 was going to become their biggest seller. There is a lot of concern over DRAM shortages especially in light of what is happening in China. On a side note Apple for their newest iPhone decided to go with the Chinese made ram because of the scale they are dealing with in shortages. Nvidia decided to go with the slower ram instead of relying on like 1 or 2 vendors stock. I pretty confident even though the performance hit the 4080 overall performance is still going to be a jump over the 3080ti. Not surprising Jensen would make a big deal about the 4090 given it is the top dog of their line to introduce and that the 4080 could be pushed out till end of Nov or Dec. Part shortages are problematic given the push to reduce the alliance on Chinese made parts "Seems counter productive but that is a war room debate".So I came back tonight to look at the 4080 more closely, and I discovered the price hike is even more egregious than previously realized. They pushed the 4090 to the forefront in the hopes people wouldn't notice.
For the 4080, it's a +$200 price hike, and that's only for the 12GB variant (a miserly +2GB VRAM over the launch version of the 3080, and they're both GDDR6X). Meanwhile, the 4080 is unimpressive. It's advantages over the 3080 are far less drastic than the 4090's advantages over the 3090 Ti. In fact, the 3080 actually has higher VRAM bandwidth than the 4080, LOL. The 16GB is absurd at $1199, and so far, they haven't declared any enhancements over the 12GB card other than the VRAM. Everything else is the same. You're expected to pay $300 for 4GB of VRAM. Holy shit. Fuck NVIDIA.
This is exactly what Intel needed. They're opening the door for Intel to become a player with that pricing if Intel aggressively undercuts them (especially if AMD doesn't). Updating my bullet lists below.
MSRP
- RTX 4080* = $1199/$899 (October, 2022: +25 months)
- RTX 3080* = $699 (September, 2020: +24 months)
- RTX 2080 = $799 (September, 2018: +28 months)
- GTX 1080 = $599 (May, 2016: +20 months)
- GTX 980 = $549 (September, 2014: +16 months)
- GTX 780 = $499 (May, 2013: +14 months)
- GTX 680 = $499 (March, 2012: +16 months)
- GTX 580 = $499 (November, 2010)
MSRP (Adjusted for Inflation; in August 2022 USD, the latest available month)
*Unlike past generations, the RTX 3080 & RTX 4080 were not the strongest GPUs at launch, but the stronger 3090/4090 variants merely supplanted the Titan series, even if not in name, so it's still apples to apples.
- RTX 4080 = $1199 / $899
- RTX 3080 = $795
- RTX 2080 = $937
- GTX 1080 = $738
- GTX 980 = $683
- GTX 780 = $634
- GTX 680 = $644
- GTX 580 = $675
Chunky board.
Oh thanks to AMD trolls it is a hip thing to take a dump on Nvidia and Intel lately on social media. I expect Intel will gain a number of customers given the price point and Nvidia will still sell a truck load of 1600 dollar cards at launch.
It does not seem to make sense at first but at the time there was huge chip shortages and dram was getting clobbered so they went with slower Dram in the 3060 instead of going with the faster dram found in the 3060ti and the older 3060 boards. This was part of Nvidia fix on the board shortages at the time with crypto insanity and COVID shortages.
RTX 3060 RTX 3060 Ti
GPU Die GA106 GA104
Architecture Ampere Ampere
VRAM 12GB GDDR6 8GB GDDR6
Last edited:
