1. Home
  2. >>
  3. rtx
  4. >>
  5. [HOT] : NVIDIA GeForce RTX 3060 graphics cards with Ampere GA104 GPU spotted

[HOT] : NVIDIA GeForce RTX 3060 graphics cards with Ampere GA104 GPU spotted



[HOT] : NVIDIA GeForce RTX 3060 graphics cards with Ampere GA104 GPU spotted


NVIDIA GeForce RTX 3060 graphics cards were spotted using the GA104 GPU by Matthieu Smith (TechPowerUp GPU Database Editor). The entries are very recent, which means the AIBs use faulty Ampere dies to power the mainstream gaming graphics card.

NVIDIA GA104 GPU-powered GeForce RTX 3060 graphics cards spotted

The NVIDIA GeForce RTX 3060 graphics card is based on the Ampere GA106 GPU, which is a completely different SKU than the GA104 that powers the GeForce RTX 3060 Ti and GeForce RTX 3070.

NVIDIA GeForce RTX 2060 12GB graphics card to launch in Q1 2022 to address GPU uptime issues

Based on the new entry, it looks like NVIDIA has a surplus stock of faulty GPU chips that didn’t make it into the high-end graphics card, but are still good enough for RTX 3060 cards. This means that Even though the graphics card will retain its base specs such as 3,584 CUDA cores and 12GB of GDDR6 memory running on a 192-bit bus interface, the chip itself will be a much higher-end SKU than the GA106 GPU d origin of the RTX 3060. core.

NVIDIA Ampere GA104 and GA106 GPUs are shown together. (Image credits: TechPowerUp)

This won’t be the first time NVIDIA has used 104 faulty dies to replace 106 SKUs. The RTX 2060 which was originally based on the TU106 GPU received several configurations, one of which featured a TU104 GPU. Other cards such as the Turing TU117 powered GeForce GTX 1650 have also received updated configurations in the TU116 and TU106 flavors. The latter offered better support for NVENC and slightly faster content creation capabilities, although we are unlikely to see those on the TU104-based RTX 3060 cards, as Turing had many architectural differences in the SKUs. High-end and mainstream GPU.

What we could see is the GeForce RTX 3060 powered by the GA104 to allow stable boost clocks as well as better power control than its GA106 constrained variant. So overall, most of you will never know what’s under the hood of your NVIDIA RTX 3060 graphics card, whether it’s a GA104 or GA106 GPU based variant. It was recently reported that the GA106 GPU supply may be limited, which could also explain why these defective GA104 dies are being reused for the RTX 3060.

NVIDIA GeForce RTX 30 Series ‘Ampere’ graphics card specifications:

Graphics card name NVIDIA GeForce RTX 3050 NVIDIA GeForce RTX 3050 Ti NVIDIA GeForce RTX 3060 NVIDIA GeForce RTX 3060 Ti NVIDIA GeForce RTX 3070 NVIDIA GeForce RTX 3070 Ti NVIDIA GeForce RTX 3080 NVIDIA GeForce RTX 3080 Ti NVIDIA GeForce RTX 3090
GPU name Ampere GA107 Ampere GA107 Ampere GA106-300 Ampere GA104-200 Ampere GA104-300 Ampere GA104-400 Ampere GA102-200 Ampere GA102-225 Ampere GA102-300
Process node Samsung 8 nm Samsung 8 nm Samsung 8 nm Samsung 8 nm Samsung 8 nm Samsung 8 nm Samsung 8 nm Samsung 8 nm Samsung 8 nm
Die size To be determined To be determined To be determined 395,2 mm2 395,2 mm2 395,2 mm2 628,4 mm2 628,4 mm2 628,4 mm2
Transistors To be determined To be determined To be determined 17.4 billion 17.4 billion 17.4 billion 28 billion 28 billion 28 billion
CUDA cores 2048 ? 2560 ? 3584 4864 5888 6144 8704 10240 10496
TMU / ROP 64 / 40 80 / 48 112 / 64 152 / 80 184 / 96 192/ 96 272 / 96 320 / 112 328 / 112
Tensor / RT Cores 64 / 16 80 / 20 112 / 28 152 / 38 184 / 46 192/ 48 272 / 68 320 / 80 328 / 82
Basic clock To be determined To be determined 1320MHz 1410 MHz 1500 MHz 1575 MHz 1440 MHz 1365 MHz 1400 MHz
Boost the clock To be determined To be determined 1780 MHz 1665 MHz 1730 MHz 1770 MHz 1710 MHz 1665 MHz 1700 MHz
FP32 calculation To be determined To be determined 12.7 TFLOP 16.2 TFLOP 20 TFLOP 22 TFLOP 30 TFLOP 34 TFLOP 36 TFLOP
RT TFLOP To be determined To be determined 25.4 TFLOP 32.4 TFLOP 40 TFLOP 42 TFLOP 58 TFLOP 67 TFLOP 69 TFLOP
Tensor-TOP To be determined To be determined 101 TOP 129.6 TOP 163 TOP 174 TOP 238 TOP 273 TOP 285 TOP
Memory capacity 4 Go de GDDR6? 4 Go de GDDR6? 12 Go de GDDR6 8 Go de GDDR6 8 Go de GDDR6 8 Go de GDDR6X 10 Go de GDDR6X 12 GB GDDR6X memory 24 Go de GDDR6X
Memory bus 128 bits 128 bits 192 bits 256 bits 256 bits 256 bits 320 bits 384 bits 384 bits
Memory speed To be determined To be determined 15 Gbit/s 14 Gbit/s 14 Gbit/s 19 Gbit/s 19 Gbit/s 19 Gbit/s 19,5 Gbit/s
Bandwidth To be determined To be determined 360 Gbit/s 448 Gbit/s 448 Gbit/s 608 Gbit/s 760 Gbit/s 912 Gbit/s 936 Gbit/s
TGP ~75W ~100W 170W 200W 220W 290W 320W 350W 350W
Prix ​​(PDSF/FE) 149 $ ? 199 $ ? 329 $ 399 $ US 499 $ US 599 $ US 699 $ US 1199 $ 1499 $ US
Launch (Availability) 2021 ? 2021 ? February 2021 december 2020 October 29, 2020 June 10, 2021 September 17, 2020 June 3, 2021 September 24, 2020