Nvidia showed off a bunch of new GPUs at CES earlier this year, including the very powerful RTX 3090 Ti. Right now, we’re interested in the 4000 Series GPUs and want to know when they’ll be available.
People have very high hopes for the next generation of graphics cards, so we’ve put together all the key information you need to know what to expect when they finally come out.
GPUs are going through a very interesting time. Because of problems with supply around the world, it is hard, if not impossible, to find graphics cards. But people still have high hopes that the RTX 4000 series will be coming soon. Nvidia hasn’t said anything official about it. But there is still a lot to sort through because of leaks, rumors, and an unfortunate data breach when Nvidia was hacked. So far, here’s what we know about Nvidia’s RTX 4000 series.
- Tesla Phone Release Date: How Much Will The Tesla Pi Phone Cost?
- Here’s How Companies Get Started with Blockchain Technology
Nvidia RTX 4000 Release Date
The A100 was made public in May 2020, and the RTX 3080 and RTX 3090, which are Ampere GPUs for consumers, came out about four months later.
If Nvidia follows a similar release schedule with Ada Lovelace GPUs, we can expect the RTX 40-series to come out between July and September. This would suggest that Nvidia will set its own launch window for Lovelace instead of piggybacking on an event like Computex or CES, but that’s not the only thing we can assume from previous releases.
The Nvidia GeForce RTX 3080, which is the best of the RTX 3000 line, came out on September 17. The Nvidia GeForce RTX 3090 came out on September 24. On October 28, the RTX 3070 came out.
This staggered release is likely to continue with Lovelace, so while the range could grow to have as many models and variations as Ampere, we expect only a few cards to be available on launch day, with new models coming out in the months after to fill any gaps in the market.
Nvidia GeForce RTX 4000 Price
The cost for any of the Lovelace GPUs slated to debut the series has yet to be formally disclosed, but we expect it to be similar to the existing Ampere line.
The MSRP of the Nvidia GeForce RTX 3080 is $699 (£649, about AU$975), whereas the Nvidia GeForce RTX 3070 is just $499 (£469, about AU$680). The RTX 3090, which was the family’s flagship at the time before the RTX 3090 Ti was released, cost $1,499 (£1,399, AU$2,100).
Unfortunately, due to the chip scarcity, many of these GPUs were artificially inflated, and many of us were unable to acquire one at this price unless we were fortunate enough to get a founders edition straight from Nvidia at launch.
Need to upgrade your GPU? Check out @sabrepc for PNY NVIDIA Quadro RTX 4000 8GB GDDR6 Graphic Card:
👉https://t.co/LZFqSKzv4z#AI #MachineLearning #ML #DeepLearning #DataScience #GPU #GraphicCard #PNY #NVIDIA #RTX4000 #Hardware #VR pic.twitter.com/sjKV59HRVx
— SabrePC (@sabrepc) April 14, 2022
Though enough stock is available for this launch, inflation and price scalping may be avoided, thus even if the GeForce RTX 4080 has the same MSRP as its younger sibling, these GPUs may be more inexpensive if history repeats itself.
There have also been fears that, because this series is expected to have double the power and performance of its predecessor, it would cost twice as much. We’ve seen no sign that this would be the case, but anything might happen given the previous several years.
Nvidia RTX 4000 GPUs Specs And Performance
The most important thing to know is, of course, how good will the 4000 series be? Maybe the most exciting thing about the new set of cards is that they will use a new design called Lovelace. These are likely to use a 5-nm (nanometer) process instead of the 8-nm process used in Ampere GPUs right now.
Based on what has been leaked, there are several models with the name “4090.” Most likely, the AD102 will be the most powerful version, and the AD103 will be the next most powerful. The AD104-106 will likely be the mid-range option, and the AD107 will be for the entry-level market.
- 12 Graphics Processing Clusters
- 18432 CUDA Cores
- 96MB L2 cache
- Potentially up to 92 Teraflops of computing performance
- 2.5GHz clock speed
Compared to the RTX 3090, which is the most powerful card in the current generation and offers:
- 36 Teraflops of computing performance
- 10496 CUDA cores
- 7 Graphics Processing Clusters
- 6MB L2 cache
- 37.6 Tflops compute performance
- 1.4GHz clock speed
It looks like the top-of-the-line AD102 GPU will have 144 SMs on a single die, while Ampere’s GA102 only has 84 SMs. This is a 71 percent increase and one of the biggest in a generation. The AD102 could have up to 18,432 CUDA cores, which is a huge 75 percent increase over Nvidia’s upcoming RTX 3090 Ti, which will only have 10,752 CUDA cores.