Just before Christmas, the rumor machine went hot on the graphics card side. Every now and then, the market expected a Geforce RTX 3080 Ti, and the specifications were long expected to be nailed to 20 GB of GDDR6 memory, and the card would meet AMD’s recently launched Radeon RX 6900 XT in a similar price range. AMD, on the other hand, had concerns about getting enough cards, and when it was time for the January trade fair, Nvidia’s RX 6900 XT challenger failed.
Now the card is reported once again to be up for launch, this time as soon as in a few weeks. According to Videocardz, a partner manufacturer specifies the release date for the RTX 3080 Ti to “Mid-April”, ie the middle of April. The new model looks to have 12 GB of GDDR6X memory and is then intended to be followed up with another Ti variant of the RTX 3070 at the end of May – also with GDDR6X memory.
Nvidia Geforce RTX 3000 “Ampere”:
RTX 3090 | RTX 3080 Ti | RTX 3080 | RTX 3070 Ti | RTX 3070 | RTX 3060 Ti | |
---|---|---|---|---|---|---|
Technical | 8 nm Samsung | 8 nm Samsung | 8 nm Samsung | 8 nm Samsung | 8 nm Samsung | 8 nm Samsung |
Circuit | GA102 | GA102 | GA102 | GA104 | GA104 | GA104 |
Circuit surface | 628 mm² | 628 mm² | 628 mm² | 392 mm² | 392 mm² | 392 mm² |
Transistors | 28.3 billion | 28.3 billion | 28.3 billion | 17.4 billion | 17.4 billion | 17.4 billion |
Architecture | Ampere | Ampere | Ampere | Ampere | Ampere | Ampere |
CUDA cores | 10 496 st. | 10 240 st. | 8 704 st. | 6 144 st. | 5 888 st. | 4 864 st. |
RT cores | 82 st. | 80 st. | 68 st. | 48 st. | 46 st. | 38 st. |
Tensor cores | 328 st. | 320 st. | 272 st. | 192 st. | 184 st. | 152 st. |
Texture units | 328 st. | 320 st. | 272 st. | 192 st. | 184 st. | 152 st. |
Raster units | 112 st. | 112 st. | 96 st. | 96 st. | 96 st. | 96 st. |
Clock frequency | 1 395 MHz | ? | 1 440 MHz | ? | 1 500 MHz | 1 410 MHz |
GPU Boost | 1 695 MHz | ? | 1 710 MHz | ? | 1 725 MHz | 1 665 MHz |
Computational power | 35 581 GFLOPS | ? | 29 768 GFLOPS | ? | 20 313 GFLOPS | 16 197 GFLOPS |
Memory amount | 24 GB GDDR6X | 12 GB GDDR6X | 10 GB GDDR6X | 8 GB GDDR6X | 8 GB GDDR6 | 8 GB GDDR6 |
Memory frequency | 19 500 MHz | 19 000 MHz(?) | 19 000 MHz | ? | 14 000 MHz | 14 000 MHz |
Memory bus | 384-bit | 384-bit | 320-bit | 256-bit | 256-bit | 256-bit |
Memory bandwidth | 936 GB/s | 912 GB/s(?) | 760 GB/s | ? | 448 GB/s | 448 GB/s |
Power supply | 12-pin | 12-pin | 12-pin | 12-pin | 12-pin | 12-pin |
SLI connection | NVLink 3.0 x4 | ? | – | – | – | – |
TBP | 350 W | ? | 320 W | ? | 220 W | 200 W |
Launch price | 1 499 USD | 999 USD(?) | 699 USD | 599 USD(?) | 499 USD | 399 USD |
Data for RTX 3080 Ti and 3070 Ti are unconfirmed.
The other specifications are not yet officially known, but an announcement from Nvidia is reported to be made in early April. However, the RTX 3080 Ti is expected to get a new circuit called the GA102-225, which like the RTX 3090 will be 384 bits wide with a memory bus and specifications that place the newcomer much closer to the top card in the series than the base model of the RTX 3080.
An upcoming top model in the RTX 3070 family with the suffix Ti also looks to be an interesting story. The card appears to use the circuit GA104-400 known from the portable RTX 3080, which would mean a full-featured circuit with 6,144 CUDA cores. In addition, the card is specified with GDDR6X memory – which means that the memory type is either compatible with existing GDDR6 memory controllers, or that Nvidia has built in memory support in all “Ampere” -based cards.
Whether the company will also continue with its artificial locking for breaking cryptocurrency in the new cards, however, is written in the stars. The original plan was likely that cards in the Geforce family launched after the RTX 3060 would limit performance when the Ethereum algorithm was discovered, something Nvidia themselves later put out of play with a developer-reinforced driver.
The new cards with the Ti suffix at least look to fill the price gaps in Nvidia’s range – we hope that the gaps on the retailers’ store shelves will also be filled.