NVIDIA Ampere GeForce RTX 3080 20 GB And RTX 3070 16 GB Graphics Cards Alleged Specifications Leak Out – Up To 3480 Cores, GA103 / GA104 GPUs
The first details for NVIDIA's next-generation Ampere based GeForce RTX 3080 and GeForce RTX 3070 graphics cards have allegedly leaked out. The rumored details come from various sources who have shared the same series of specifications for NVIDIA's upcoming GeForce GPUs.
NVIDIA Ampere GPU Based GeForce RTX 3080 20 GB & GeForce RTX 3070 16 GB Ampere Graphics Cards Alleged Specifications Detailed
From the details by MyDrivers and Twitter user 'KittyCorgi', it seems like NVIDIA is prepping up at least two Ampere GPUs which would go on to power the GeForce RTX 3080 and GeForce RTX 3070 graphics cards. The details cover the full specifications, including core counts and even images of complete block diagrams for NVIDIA's next-generation Ampere powerhouse GPUs.
The NVIDIA GA102, GA103 & GA104 GPUs, Powered by Ampere
First up, it is reported that there are two gaming derived GPUs based on the Ampere architecture, these include the GA103 and the GA104. There's also a GA102 GPU mentioned but there's no card mentioned that would make use of it. It is possible NVIDIA could keep the chip under wraps for a more powerful graphics card.
Like Pascal and Volta, which kept the top tier chips exclusive to HPC / AI / DNN markets, it is also likely that the GA100 GPU would exist but would be developed specifically for the compute GPU market with the more gaming-optimized chips being derived into the GA102/GA103/GA104 tiers. The original Ampere GPU lineup leak goes far back to May of 2019 where a series of NVIDIA GPUs based on the Ampere architecture was hinted (not confirmed), these include (credits: kopite7kimi):
- GA100 8GPC*8TPC*2SM 6144bit
- GA102 7GPC*6TPC*2SM 384bit
- GA103 6GPC*5TPC*2SM 320bit
- GA104 6GPC*4TPC*2SM 256bit
- GA106 3GPC*5TPC*2SM 192bit
- GA107 2GPC*5TPC*2SM 128bit
Now, none of these GPUs have officially been confirmed but we have seen references of NVIDIA's Ampere GA104 GPU popping up in recent ECC listings and this is what these GPU configurations were based on. Like the current Turing and the previous Pascal lineup, the GA102 and the GA104 GPU will make up the higher-end market with GA106 making up the mainstream and GA107 targetted at the entry-level segment. What's interesting is that this leak indicates that the green team might be deciding to base the RTX **70 tier graphics card on the higher tier G**04 chips rather than G**06.
We know that the Turing based GeForce RTX 2070 makes use of the TU106 chip and not the upper-tier TU104 chip like the RTX 2080. At the same time, the RTX **80 tier graphics card would not be using the G**04 chip but rather an even higher tier GPU known as the GA103 which would allow for a wider bus and higher memory configuration alongside more CUDA cores. There's no mention of what fabrication process Ampere GPUs will be based upon, but NVIDIA's CEO stated that the majority of orders would be handled by TSMC for their next-generation 7nm GPUs while Samsung would get a small number of orders. We also know that NVIDIA's Orin SOC which makes use of its next-generation GPU architecture (Ampere) is based upon Samsung's 8nm LPP process node which makes it clear where the smaller numbers of orders are going to. With that said, let's talk about the graphics cards themselves.
The Ampere GPU wouldn't only improve the rasterization and shading performance but also deliver huge uplifts over Turing in ray tracing performance. Since Turing was the 1st generation implementation of ray tracing on NVIDIA graphics cards, Ampere will further optimize it and is expected to offer more ray-tracing hardware onboard in the form of RT and Tensor cores. The Ampere GPUs are also mentioned to leverage performance from PCIe Gen 4.0 which is a given since this is a next-generation GPU architecture tuned for next-generation gaming platforms and the higher bandwidth from the Gen 4.0 protocol, plus NVLINK would be an essential feature.
NVIDIA GeForce RTX 3080 Graphics Card - Up To 20 GB GDDR6, 320-bit Bus, GA103 GPU
As mentioned above, the NVIDIA GeForce RTX 3080 will be the successor to the GeForce RTX 2080. It will feature a total of 60 SMs with 64 CUDA cores each making 3840 CUDA cores. This would be a big leap over the RTX 2080's CUDA core count which was set at 3072 and we aren't even factoring in the clock speeds which NVIDIA will be able to churn up with an advanced 7nm process node.
The card is also said to feature a 320-bit bus interface which means that it could support either 10 or 20 GB of GDDR6 memory. If NVIDIA really wants to blow the audience away with Ampere, then 20 GB would be a perfect choice. However, we have to take into consideration the high costs that are associated with GDDR6 memory and the looming rumors of a memory shortage due to higher demand from the console market.
So let's be conservative and say 10 GB of GDDR6, which would still be 2 GB more than the current GeForce RTX 2080. If we are looking at 16 Gbps modules on the upcoming RTX 3080, it would give us 640 GB/s bandwidth which would be insane over the 496 GB/s of the RTX 2080 SUPER which features 15.5 Gbps memory alongside a 256-bit bus interface.
NVIDIA GeForce RTX 3070 Graphics Card - Up To 16 GB GDDR6, 256-bit Bus, GA104 GPU
The second card is the GeForce RTX 3070 and as stated above, it will feature the GA104 GPU. The successor to the TU106 based GeForce RTX 2070 is suggested to feature the same CUDA core count as the RTX 2080, at 3072, all rounded up in its 48 streaming multiprocessor units. The card is expected to feature a 256-bit bus interface which once again would allow for 8 GB or up to 16 GB GDDR6 memory, but you have to take into consideration what I just said above about the GDDR6 supply and cost factor.
NVIDIA Ampere GPU 'GeForce RTX 3080 & RTX 3070' Rumored Specs Comparison:
|Graphics Card Name||NVIDIA GeForce RTX 2070||NVIDIA GeForce RTX 2080||NVIDIA GeForce RTX 3070||NVIDIA GeForce RTX 3080||NVIDIA GeForce RTX 2080 Ti|
|GPU Architecture||Turing TU106||Turing TU104||Ampere GA104||Ampere GA104||Turing TU102|
|Process||12nm FinFET NVIDIA (TSMC)||12nm FinFET NVIDIA (TSMC)||7nm TSMC?||7nm TSMC?||12nm FinFET NVIDIA (TSMC)|
|CUDA Cores||2304 CCs||3072 CCs||3072 CCs||3840 CCs||4608 CCs|
|Memory||8 GB GDDR6||8 GB GDDR6||8/16 GB GDDR6||10/20 GB GDDR6||11 GB GDDR6|
|Memory Speed||14 Gbps||14 Gbps||16 Gbps?||16 Gbps?||14 Gbps|
|Memory Bandwidth||448 GB/s||448 GB/s||512 GB/s||640 GB/s||640 GB/s|
What's interesting is that these two cards leave so much room open for a GeForce RTX 3080 Ti to just blow away the enthusiast market segment again. If the reports of NVIDIA's overall Ampere architecture gaining 50% higher performance than Turing while offering twice the efficiency are correct, than given those specs, an RTX 3070 would be able to deliver the performance or even outperform the RTX 2080 Ti at a much lower price point.
According to the Taipei Times, NVIDIA's Ampere GPU is going to deliver seriously impressive figures in both performance and efficiency. The overall performance gain is expected to be 50% higher than existing Turing GPUs while offering twice the power efficiency. It is clearly mentioned that Ampere would halve the power consumption compared to existing Turing GPUs. This means that a GPU with a lower TDP figure than the RTX 2060 (125-150W) would be able to outperform the 2080 Ti which is seriously impressive. One can only expect the kind of performance which a high-end GPU based on the Ampere architecture would offer as we can definitely expect true 4K 60 FPS ray-traced gaming from the updated architecture.
We were told NVIDIA will launch a new GPU in the second quarter.
AMD is also expected to have Big Navi around Computex.
Info from board partners.
— VideoCardz.com (@VideoCardz) January 20, 2020
There's no word on an exact launch plan for Ampere based GeForce GPUs, but rumors have hinted at a 2H 2020 launch, however, it could be much earlier than that. NVIDIA is expected to host their annual GTC 2020 conference in March, which would mean that we could finally get to hear more about the next-generation green team powerhouse chip, even if it's an HPC card announcement first. This was the case with Turing, which got to see its first public appearance in the form of a workstation card or Pascal which was first unveiled as an HPC oriented compute chip. We will definitely keep you updated in case we hear more about NVIDIA's next-generation GPUs.