NVIDIA new Ampere GeForce RTX 3070 should cost $599, and offer RTX 2080 Ti level performance that debuted at $999-$1199.
We are less than a week away from NVIDIA’s huge GeForce RTX 30 series reveal event on September 1, where we should hopefully be introduced to the new Ampere-based graphics cards.
We already know there will be a 12-pin PCIe power connector on the flagship GeForce RTX 3090 Founders Edition graphics card (which requires a huge 850W power supply), but there’s some new dirt on the cheaper GeForce RTX 3070: it should beat the current Turing-based flagship GeForce RTX 2080 Ti.
Tom from Moore’s Law is Dead tweeted out that he can confirm the GeForce RTX 3070 has around the same level of performance of the GeForce RTX 2080 Ti. He adds that “you all can do the math on how much stronger the 3080 & 3060 are relative to this based on the previous gen” with a nice big fat wink face emoticon at the end.
This is about where I’m expecting performance to fall, so it gives the GeForce RTX 3080, GeForce RTX 3080 Ti, and GeForce RTX 3090 some wiggle room to each have their own carved out slices of the Ampere dominance.
NVIDIA should be pricing the GeForce RTX 3070 at $599: The new Ampere-based GeForce RTX 3070 being introduced at $599 would be a gigantic deal, as the GeForce RTX 2080 Ti — which it should have the same performance as — debuted at $999-$1199 in September 2018.
We should expect to see NVIDIA have the GeForce RTX 3070 in two different versions, one offering 8GB of GDDR6 and the other offering 16GB of GDDR6. The GeForce RTX 2080 Ti has 11GB of GDDR6, so maybe we might see a cheaper version with 8GB of RAM — smaller framebuffer, but same GPU performance as the RTX 2080 Ti.
Meanwhile, a higher-end card could offer 16GB of VRAM if you wanted to play higher-res titles or games like Flight Simulator — which is the new Crysis — which chew up 16GB of VRAM at 8K.
Here’s an easier break down for you:
GeForce RTX 3090: $1399
GeForce RTX 3080: $799
GeForce RTX 3070: $599
GeForce RTX 3060: $399
The very latest: RTX 3090 PCB + GDDR6X @ 21Gbps
The PCB of the RTX 3090: A juicy leak of the purported PCB of a custom variant of the GeForce RTX 3090 leaked out a few days ago. It is reportedly the COLORFUL iGame GeForce RTX 3090 Vulcan-X. You can see this model has 3 x 8-pin PCIe power connectors, unlike the RTX 3090 Founders Edition that should pack just a single 12-pin PCIe power connector.
Micron reveals RTX 3090 name, 24GB GDDR6X memory @ 21Gbps: I’m sure this is part of the marketing or else all of us tech media websites would be receiving calls from lawyers, but Micron revealed its next-gen GDDR6X memory that the GeForce RTX 3090 (it admits the card by name) will be clocked at a bonkers-high 21Gbps.
Wait, a 12-pin PCIe power connector?
12-pin PCIe power connector on Founders Edition ONLY: NVIDIA will be using a special new 12-pin PCIe power connector on its GeForce RTX 3090 Founders Edition graphics cards, something that I’ve confirmed with a few industry sources of mine now (to confirm the rumors floating around online).
Many, many 8-pin PCIe power connectors on custom AIB: But, custom GeForce RTX 3000 series graphics cards from AIBs like ASUS, COLORFUL, EVGA, MSI, etc will not be using the 12-pin PCIe power connector and rather multiple (2 and even up to 3 x 8-pin PCIe power connectors).
Traversal coprocessor: We have had more leaks on NVIDIA’s next-gen GeForce RTX 3000 series than any family of graphics cards before it, with an interesting “traversal coprocessor” on the new GeForce RTX 3080 and GeForce RTX 3090 graphics cards.
NVCache: Ampere is meant to have something called NVCache, which would be NVIDIA’s own form of AMD’s HBCC (High Bandwidth Cache Controller, more on that here). NVCache would use your system RAM and SSD to super-speed game load times, as well as optimizing VRAM usage. You can read more on NVCache here
Tensor Memory Compression: NVCache is interesting, but Tensor Memory Compression will be on Ampere, and will reportedly use Tensor Cores to both compress and decompress items that are stored in VRAM.
This could see a 20-40% reduction in VRAM usage, or more VRAM usage with higher textures in next-gen games and Tensor Memory Compression decreasing that VRAM footprint by 20-40%.
How fast is the GeForce RTX 3090? Freaking fast according to rumors, with 60-90% more performance than the current Turing-based flagship GeForce RTX 2080 Ti. We could see this huge performance leap in ray tracing titles, but we’ll have to wait a little while longer to see how much graphical power NVIDIA crams into these new cards. You can read more on those rumors here.
Power hungry: As for power consumption, GA102 reportedly uses 230W — while 24GB of GDDR6X (which we should see on the new Ampere-based TITAN RTX) consumes 60W of power.
You can read more on that here.
Production begins soon: NVIDIA is reportedly in the DVT (or Design Validation Test) range of its new GeForce RTX 3000 series graphics cards. Mass production reportedly kicks off in August 2020, with a media event, benchmarks, and more in September 2020 as I predicted many months ago. More on that here.
I’ve already written about rumors that NVIDIA’s next-gen Ampere GPU architecture would be up to 75% faster than current-gen GPUs such as the Turing architecture, right after rumors that Ampere would offer 50% more performance at half the power of Turing. This is pretty crazy stuff right there.
Not only that, but we’ve got some rumored specs on the purported GeForce RTX 3080 and GeForce RTX 3070 graphics cards, which will both be powered by NVIDIA’s new Ampere GPU architecture.
We’ve already heard that Ampere would offer 50% more performance at half the power of Turing, which sent the hairs on my neck standing up. Better yet, you can read about the leaked specs on the purported Ampere-based GeForce RTX 3080 and GeForce RTX 3070 right here.
Originally published at Tweak Town