The RTX 3080 is reportedly using the GA102-200-Kx-A1 GPU, with 4352 CUDA cores and 10 GB of VRAM at 19Gbps, while the RTX 3090 is expected use the GA102-300-A1 GPU, packing 5248 CUDA cores and 12GB of GDDR6X RAM at 21Gbps. The upcoming Titan GPU should utilize the GA102-400-A1, with 5376 CUDA cores and a massive 24GB of VRAM at 17Gbps. The type of VRAM (GDDR6/GDDR6X) used for both the RTX 3080 and the upcoming Titan was not shared.
Regarding the leaked 3DMark Time Spy score of an “Unknown Nvidia Ampere GPU”, rogame from HardwareLeaks spotted a “hidden” database entry detailing an Ampere GPU with a 1935MHz boost clock and 6000MHz memory clock. One thing to note is the odd “6000MHz memory clock” listed in the 3DMark Time Spy database. GDDR6 and HBM2 memory clock values are usually multiplied by 8, which will then result in the “real” memory clock. As per rogame, the displayed 6000MHz memory speed might be “the result of an early/internal driver or a new type of memory that is still unknown to 3dmark detection software”.
This “Unknown Nvidia Ampere GPU” recorded a graphics score of 18257, as seen above. This translates to an improvement of over 30% compared to the stock RTX 2080Ti FE, or 22% when compared to the Nvidia Titan RTX. HardwareLeaks also included a score achieved by KINGPIN through overclocking an EVGA RTX 2080Ti XC with LN2, reaching a 2385MHz boost clock and a 17360Hz memory clock, which beats the Ampere GPU’s score by about 2%.
Rogame confirmed the authenticity of this entry, claiming that it belongs to an Nvidia employee. The 1935MHz boost clock may not seem like a big improvement over the current generation of GPUs, but this entry is likely to be an engineering sample so clock speeds could be adjusted between now and the official RTX 3000 launch.
RTX 3080 release date
For a while, the rumor mill was suggesting that we’re going to see the next GeForce lineup make an appearance at GTC, or the GPU Technology Conference in March. But, Nvidia has since canceled its ‘exciting’ GTC graphics card revelations. Since then, Nvidia has officially announced the Nvidia Ampere for Data Center. And, as expected, it did so before the RTX 3000 lineup, just like when it launched Volta in May 2017 more than a year before Turing.
Luckily, it doesn’t seem like we have to wait until 2021 to get a glimpse of the next generation RTX cards. In the past, Nvidia GeForce lineups have been about two years apart. For instance, the Nvidia GeForce GTX 1080 dropped back in May 2016, with the GeForce RTX 2080 succeeding it in August 2018. The time between the GTX 980 and GTX 1080 was a bit shorter, with the former arriving in September 2014. Still, as a general rule, we’ve been able to historically count on there being a new lineup of graphics cards every couple of years.
Tweaktown has reported that Nvidia plans to unveil the Ampere GPU architecture in August this year and launch it in September at Computex 2020. Similarly, a more recent rumor has Nvidia starting mass production of the GPUs in August, and holding an official media reveal event in September 2020. It’s also likely that Nvidia could show off the GPUs a bit earlier than the media event at a public event, possibly at Gamescom 2020 in August.
This is certainly likely, as it’s similar to how Nvidia launched Turing at Gamescom back in August 2018, and would mark a straight 2-year cycle for the current-generation cards.
We’ve heard that the Nvidia GeForce RTX 3080 Ti, which may be up to 40% faster than its predecessor and run with hugely fast boost clocks of up to 2.2GHz as well as 18Gbps video memory, could also launch by the end of 2020.
In any case, AIB partners are said to be clearing out their stock now, allegedly in preparation for an Nvidia Ampere launch in Q3 2020, so chances of us seeing the RTX 3080, as well as the more powerful RTX 3080 Ti and RTX 3090, this year are high. Though with how these things typically go, we won’t actually know when the RTX 3080 is on its way until Nvidia is ready.
RTX 3080 specs
We have seen a lot of really vague speculation and rumors surrounding supposed RTX 3080 specs, but there are some constants: that it’ll be 7nm and will be faster than the RTX 2080. That second one should be super obvious, of course – why would Nvidia put out new graphics cards if they’re not significantly faster?
The other rumor, the 7nm manufacturing process, is much more interesting. If you’ve been following the AMD vs Intel smackdown over the last year, you’re probably aware that one of the things that’s kept AMD on top of the CPU world is the fact that it’s been able to fully adopt 7nm while Intel, at least when it comes to desktop, being stuck at 14nm. If Nvidia is able to adopt a 7nm manufacturing process for the RTX 3080, AMD won’t have such an easy time dunking on it.
If these rumors are actually true, it would mean that Nvidia could significantly boost performance – up to 50% according to some rumors – while boosting power efficiency significantly. When combined with DLSS, which will absolutely be a part of whatever graphics cards Nvidia puts out next, could potentially lead to 4K graphics cards being available in the mid-range sector for the first time.
All that power requires a much more efficient cooling system, so recently leaked images showing a prototype of the Nvidia GeForce RTX 3080 with a striking new fan design is naturally making rounds.
Hardware leaker kopite7kimi has also speculated the RTX 3080 might use the GA102-200 GPU, and that there may just be an RTX 3090 graphics card backing it up. If this educated guesswork holds water, we could be looking at an AMD Big Navi killer.
Exactly how much faster and more powerful the RTX 3080 would be, it’s too early to tell. There’s an Nvidia GPU with a whopping 33 teraflops, 24GB of memory, and 1.11GHz clock that was recently spotted in a benchmark, as well as another one with 124 compute units. While these may be too powerful to be the RTX 3080, they could also hint towards what the Nvidia GeForce RTX 3080 may look like.
We have even seen some more specific spec leaks, notably one from hardware leaker KittyCorgi, who suggests that Nvidia’s next-generation flagship could potentially come with up to 20GB of VRAM. Of course, this is a leak we’d take with a hefty dose of salt, but it could potentially prove our theory that the next graphics architecture Nvidia announces will in fact be for professionals. The Quadro RTX 6000, for instance, packs 24GB of GDDR6 – but it also costs $4,000 (£4,066, AU$6,999).
Right now, if we had to put our money down, we’d suggest that these leaked specs are for an Nvidia Quadro or Tesla card, or maybe even the follow-up to the Nvidia Titan V. Whatever the GeForce RTX 3080 looks like, we seriously doubt we’ve seen it yet.
Meanwhile, a massive leak from Moore’s Law is Dead reveals a GA102 GPU that will theoretically be Nvidia’s top-end Ampere consumer offering. It’s highly likely going to be named the RTX 3080 Ti, which we may also have already caught a glimpse of, thanks to Asus.
This graphics cards looks to be an out-and-out beast of a graphics card with hugely fast boost clocks of up to 2.2GHz, offering 21 TFlops of raw performance, 18Gbps video memory – 864GB/s of memory bandwidth, around 40% more than the 2080 Ti (at 616GB/s), and PCIe 4.0 support.
More recent rumors, however, do point towards the Nvidia GeForce RTX 3080 Ti having a massive 627mm² die, adding weight to the rumors that Nvidia Ampere GeForce GPUs will be based on Samsung’s 8nm node, instead of TSMC’s 7nm process.
Other lesser graphics cards in the 3000 range have been revealed as well, and they may push those boost clocks to 2.5GHz. The RTX 3070 and RTX 3070 Ti are allegedly going to be built on the cut-down GA104-300 and the GA104-400 respectively. The RTX 3070 will purportedly have 2,944 CUDA cores (the same as the existing RTX 2080), along with 8GB of GDDR6 video memory. On the other hand, the RTX 3070 Ti will up the ante to 3,072 CUDA cores and have GDDR6X memory.
Moore’s Law is Dead argues that 4K graphics will become the mainstream standard with these new Ampere models, and we might be looking at the future of gaming on PC coming in 4K.
Don’t worry, though. We’ll keep our eyes open for the next generation of GeForce, and we’ll be sure to update this article as soon as we see any juicy RTX 3080 spec rumors.