On Tuesday, NVIDIA launched its flagship GeForce RTX 3080 Ti laptop graphics chip. Soon after the launch, Geekerwan, a Chinese YouTuber, released a video showcasing the very first pictures of the Ampere GPU that powers the new laptop powerhouse.
NVIDIA GeForce RTX 3080 Ti GPU's GA103 Chip Pictured, Die Size Measures Around 496 mm²
\Almost three years ago, Twitter leaker koptie7kimi tweeted on an undisclosed future NVIDIA Ampere series that includes GA100, GA101, GA102, GA103, GA104, GA106, and GA107 variants. The NVIDIA GeForce RTX 3080 Ti GA103 was on the complete list of upcoming RTX 30 series graphics cards. But the GPU never saw the light of day until now.
The new laptop graphics processor series, which includes GA103S and GA103M graphics chip variants (model GN20-E8-A1), offers 7424 CUDA cores. The rumored specifications showed the graphics chip to rock 7680 CUDA cores so it looks like we are looking at a slightly disabled SKU.
The standard GA103 card has ten 32-bit memory controllers, implying a full 320-bit memory bus. This makes it the widest memory bus interface available on current mobility GPUs.
NVIDIA's GeForce RTX 3080 Ti laptop graphics card only offers a 256-bit interface. However, this information is not conclusive until NVIDIA reveals the official block diagram of their new GA103 GPU.
Geekerwan uses a uniquely creative way to display the GA103's placement in the unrevealed lineup. It appears the chip is only utilized by RTX 3080 Ti Laptop graphics processor, while the RTX 3070 Ti card still depends on the existing GA104 GPU which powers the RTX 3070 series cards.
Geekerwan states the GA103 graphics card measures 496 mm² — 26.5% larger than the GA104 card offering 25% additional cores. NVIDIA has not disclosed its objectives for the GA103 series for desktop computers. We anticipate that certain GA103 cards will appear in the RTX 3070 Ti series in the future. VideoCardz notes that NVIDIA will need to sell specific chips with somewhat unusable CUDA cores in the future.
Idk what's this lol pic.twitter.com/bQ0d6CJP7s
— Olrak (@Olrak29_) January 27, 2022
Twitter user Olrak also unearthed a picture of the GA107 graphics card earlier today after several months of no one confirming with photographic proof. This specific chip is not the official retail model but an engineering sample because it carries a different moniker.
NVIDIA GeForce RTX 30 Mobility GPU Lineup:
| GPU Name | NVIDIA GeForce RTX 3050 | NVIDIA GeForce RTX 3050 Ti | NVIDIA GeForce RTX 3060 | NVIDIA GeForce RTX 3070 | NVIDIA GeForce RTX 3070 Ti | NVIDIA GeForce RTX 3080 | NVIDIA GeForce RTX 3080 Ti |
|---|---|---|---|---|---|---|---|
| Process Node | Samsung 8nm | Samsung 8nm | Samsung 8nm | Samsung 8nm | Samsung 8nm | Samsung 8nm | Samsung 8nm |
| GPU SKU | GA107 | GA107 | GA106 | GA104 | GA104 | GA104 | GA103 |
| SMs | 16 | 20 | 30 | 40 | 46 | 48 | 58 |
| CUDA Cores | 2048 | 2560 | 3840 | 5120 | 5888 | 6144 | 7424 |
| Base Clock | 1057 MHz | 1035 MHz | 1283 MHz | 1290 MHz | 1035 MHz | 1245 MHz | 1125 MHz |
| Boost Clock | 1740 MHz | 1695 MHz | 1703 MHz | 1620 MHz | 1485 MHz | 1710 MHz | 1590 MHz |
| Memory Clock | 12 Gbps | 12 Gbps | 14 Gbps | 14 Gbps | 14 Gbps | 14 Gbps | 16 Gbps |
| Memory Type | GDDR6 | GDDR6 | GDDR6 | GDDR6 | GDDR6 | GDDR6 | GDDR6 |
| Memory Size | 4 GB | 4 GB | 6 GB | 8 GB | 8 GB | 8/16 GB | 16 GB |
| Memory Bus | 128-bit | 128-bit | 192-bit | 256-bit | 256-bit | 256-bit | 256-bit |
| Bandwidth | 192 GB/s | 192 GB/s | 336 GB/s | 448 GB/s | 448 GB/s | 448 GB/s | 512 GB/s |
| TGP | 35-95W | 35-95W | 60-115W | 80-125W | 80 - 125W | 80-150W+ | 175W |
| Configurations | Max-Q Max-P |
Max-Q Max-P |
Max-Q Max-P |
Max-Q Max-P |
Max-Q Max-P |
Max-Q Max-P |
Max-Q Max-P |
| Launch | Q2 2021 | Q2 2021 | Q1 2021 | Q1 2021 | Q1 2022 | Q1 2021 | Q1 2022 |
The post NVIDIA GA103 GPU That Powers The Flagship GeForce RTX 3080 Ti Mobility Graphics Chip Has Been Pictured by Jason R. Wilson appeared first on Wccftech.