NVIDIA GeForce RTX 4090 graphics card once again pointed to a launch in October

NVIDIA’s GeForce RTX 4090 graphics card with the flagship AD102 GPU will once again be pointed at a launch in October this fall.

NVIDIA’s GeForce RTX 4090 graphics card with flagship AD102 GPU on its way to launch in October, 24 GB GDDR6X memory on board

Back in early June, we reported that NVIDIA’s GeForce RTX 4090 graphics card could be launched in October. The original plans were to launch it a little earlier, but they changed shortly after NVIDIA decided to clear its existing stock of GeForce RTX 30 Series graphics cards to make room for the new flagships.

Now NVIDIA is trying its best to clear up high-end GeForce RTX 30 storage by offering price cuts and discounted promotions through its partners. The company has recently lowered the prices of various RTX 30 Founders Edition graphics cards, underlining the fact that they seriously want to get rid of these cards as soon as possible before the next generation is ready for launch.

When we return to the GeForce RTX 4090, NVIDIA has not put anything in stone yet and we will not use the word ‘confirmation’ for any launch date yet as they may change weeks before the unveiling. But Moore’s Law is Dead is the fourth source to mention an October launch of the flagship AD102 GPU-based graphics card. Our sources are still confident in the tip they gave us last month and there is a delicious one too Kopite7kimi who believes that the card will be launched in early 4. quarter (October 2022).

It is also mentioned by MLID that although the card would be launched in October, the actual volume is not expected until December, so it looks like there may be an early ownership tax on RTX 4090 and other advanced Ada Lovelace GPU-powered cards being launched in Q4 2022, such as RTX 4080 and RTX 4070.

According to Greymon55The AD102 ‘GeForce RTX 4090’ is also expected to be launched in October, while the other GPUs, AD103, AD104 and AD106 are said to be launched in early 2023. Again, take it all with a pinch of salt.

That said, the NVIDIA GeForce RTX 4090 Ti is also mentioned in the rumor of being a 600W power eater, but it could be launched later. The 3090 Ti was introduced more than a year after the RTX 3090, so the same could happen with the 4090 Ti unless AMD comes up with something fast enough with its own RDNA 3 lineup to justify the release of a more powerful AD102 part, one with features the full GPU kernel configuration.

NVIDIA GeForce RTX 4090 ‘Rumored’ specifications

The NVIDIA GeForce RTX RTX 4090 graphics card is expected to be powered by the top AD102-300 GPU, but only the ‘Ti’ variant will contain the entire chip. The GeForce RTX 4090 will use a slightly cut configuration.

The NVIDIA GeForce RTX 4090 will use 128 SMs out of the 144 SMs for a total of 16,384 CUDA cores. The GPU comes packed with 96 MB of L2 cache and a total of 384 ROPs, which is simply insane. The clock speeds have not been confirmed yet, but considering that the TSMC 4N process is being used, we expect clocks between the 2.0-3.0 GHz range.

In terms of memory specifications, the GeForce RTX 4090 is expected to rock 24 GB of GDDR6X capacity, which will be clocked at 21 Gbps speeds across a 384-bit bus interface. This will provide up to 1 TB / s bandwidth. This is the same bandwidth as the existing RTX 3090 Ti graphics card, and in terms of power consumption, TBP is said to be rated at 450W, which means that TGP can end up lower than that. The card is powered by a single 16-pin connector, which delivers up to 600W of power. It is likely that we can get 500W + custom designs as we saw with RTX 3090 Ti.

NVIDIA GeForce RTX 4090 Ti & RTX 4090 ‘Preliminary’ specifications:

Graphics card name NVIDIA GeForce RTX 4090 Ti NVIDIA GeForce RTX 4090 NVIDIA GeForce RTX 3090 Ti NVIDIA GeForce RTX 3090
GPU name Ada Lovelace AD102-350? Ada Lovelace AD102-300? Ampere GA102-350 Ampere GA102-300
Process Node TSMC 4N TSMC 4N Samsung 8nm Samsung 8nm
Die Size ~ 600 mm2 ~ 600 mm2 628.4 mm2 628.4 mm2
Transistors TBD TBD 28 billion 28 billion
CUDA colors 18432 16128 10752 10496
TMUs / ROPs TBD / 384 TBD / 384 336/112 328/112
Tensor / RT cores TBD / TBD TBD / TBD 336/84 328/82
Basis ur TBD TBD 1560 MHz 1400 MHz
Boost ur ~ 2800 MHz ~ 2600 MHz 1860 MHz 1700 MHz
FP32 Compute ~ 103 TFLOPs ~ 90 TFLOPs 40 TFLOPs 36 TFLOPs
Tensor TOPs TBD TBD 320 TOP 285 TOP
Memory capacity 24 GB GDDR6X 24 GB GDDR6X 24 GB GDDR6X 24 GB GDDR6X
Memory bus 384-bit 384-bit 384-bit 384-bit
Memory speed 24.0 Gbps 21.0 Gbps 21.0 Gbps 19.5 Gbps
Bandwidth 1152 GB / s 1008 GB / s 1008 GB / s 936 Gbps
TGP 600W 450W 450W 350W
Price (MSRP / FE) 1999 USD? 1499 USD? 1999 USD 1499 USD
Launch (availability) 2023? October 2022? March 29, 2022 September 24, 2020

Products mentioned in this post

Leave a Comment