It seems with each passing day the hype train for NVIDIA's next-gen Ampere GeForce RTX series graphics cards grow, with the Ultimate Countdown... well... counting down until the big special event on September 1.
The latest news on NVIDIA's next-gen GeForce RTX series is that the GA102 GPU board design will pack some delicious power delivery, seeing enthusiasts eyebrows raise (well, at least just one of them). NVIDIA has been slowly providing better and better power delivery designs with its high-end Founders Edition cards, but this is taking it to the next level.
NVIDIA's highest-end Ampere GA102 GPU board design should be the one expected in the GeForce RTX 3080 Ti or GeForce RTX 3090, or GeForce RTX 3090 ULTRA or ULTIMATE -- whatever-the-hell NVIDIA end up calling it.
The upgraded power delivery is reportedly "over 20 chokes", but we don't know which model that the leaker is talking about. It could be any of the flagship Ampere GA102-based cards. How is this any different to the power delivery systems on the previous-gen GeForce GTX and GeForce RTX graphics cards.
- GeForce GTX 1080 Ti: 7 VDDC + 2 MVDD
- GeForce RTX 2080 Ti: 13 VDDC + 3 MVDD
- Ampere GA102 PCB: "over 20 chokes"
Some notes were also compiled from Chiphell, with many references to the GeForce RTX 3090. But the note at the bottom is the best: "It's better than Mrs. Su's HBM card" which seems to be in reference to those leaks about a new Radeon RX 5950 XT "Big Navi" with 24GB of ultra-fast HBM2e memory.
- 1. So-called 24/20GB 384/320-bit Only right on one point.
- 1) 3090 is 24GB this is fine, pushing up KatCorgi also confirmed
- (2) then in order to high-speed video memory is not used 16Gbit particles, this is a double-sided 2Rx32 specifications, positive and negative a total of 24 stars!
- 2, 3080 and 3090, although using the same core, but 3080 seems to be to re-sample the PCB, rather than 3090 cut two particles made out of the
- 3, by the way, 3090 is an interesting arrangement of this particle, there is a particle put to the bottom
- 4, this time the public version of the power supply luxury is unimaginable, or cooling pressure is unimaginable.... NVIDIA SMX kind of card power supply density you all understand, this time is bound to do a dense, a cylinder solid capacitors are not, instead of all the chip capacitors / tantalum capacitors.
- 5, is not a shaped PCB I do not know anyway, but NV this PCB layout merit to be honest!
- It's better than Mrs. Su's HBM card.
We've heard about the power delivery system on the GA102 before, where Igor's Lab reported that the TGP of the purported GeForce RTX 3090 would be 350W. The breakdown of this is 230W for the GA102 GPU, 60W for the 24GB of GDDR6X memory, 35W+ across MOSFETs and other power systems for a total of 350W.
- What's NVIDIA doing with all those mentions of 21? I'm a kinda numerology nerd in some ways, so it helps decode at least some of it -- which I did so here. (GeForce) 256 x 21 (days/years celebration) = 5376 (GPU cores in purported GA102 GPU), and 5 + 3 + 7 + 6 = 21.
- What about "ULTIMATE"? Well that's the other thing -- does NVIDIA mean this will be the "ultimate graphics card", or does it mean we could expect an ULTIMATE branded graphics card? GeForce RTX 3090 ULTIMATE has a nice ring to it... just saying.
- Where can I tune into the GeForce special event? NVIDIA will be premiering the special event broadcast on September 9 at 9AM PT, hosted by none other than NVIDIA CEO and founder Jensen Huang. You can tune in right here.
More reading:
- 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. You can read more on that here.
- 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.
Even more reading:
- NVIDIA hosting 'special event' for next-gen GeForce on September 1
- NVIDIA mega rumor: GeForce RTX 2190, GeForce RTX 2180 Ti naming scheme
- NVIDIA is 'not worried' about AMD's next-gen Big Navi GPU
- NVIDIA crazy rumor: RTX 3080 software locked VRAM until RDNA 2 is out
- NVIDIA's flagship GeForce RTX 30 series card could have 24GB VRAM
- Exclusive: AIB custom Ampere cards launch alongside Founders Edition
- NVIDIA should reveal next-gen GeForce RTX 30 series cards on August 31
- NVIDIA GeForce RTX 3000 series announcement rumored for September 9
- NVIDIA reportedly discontinues GeForce RTX 2070 SUPER graphics card
- NVIDIA GeForce RTX 3080 Ti: rumored launch of September 17
- AMD Big Navi: 5120 stream processors teased for RDNA 2 flagship again
- AMD Radeon Instinct MI100 Acturus teased, NVIDIA Ampere destroyer?!
- GeForce RTX 3090: 50% faster than RTX 2080 Ti in early benchmarks
- NVIDIA A100 Ampere benchmarked, is now the 'fastest GPU ever recorded'
- Yeah, NVIDIA should unleash GeForce RTX 3000 series in September
- AMD Big Navi 'NVIDIA Killer' flagship card has 16GB of memory
- NVIDIA GeForce RTX 3080 rumor: 20% faster than GeForce RTX 2080 Ti
- NVIDIA's next-gen Ampere DLSS 3.0 could work on ANY game that uses TAA
- AMD aims Big Navi launch for November as 'show of strength' for RDNA 2
- NVIDIA's next-gen Ampere GeForce RTX 30 series: new 12-pin PCIe power
- AMD's next-gen RDNA 2 rumor: 40-50% faster than GeForce RTX 2080 Ti
- NVIDIA stops making RTX 20 series, ramps to GeForce RTX 30 series
- GeForce RTX 3060 should cost $300-$400, here's some leaked specs
- NVIDIA bundles Death Stranding with GeForce RTX graphics cards
- NVIDIA GeForce RTX 3070 Ti, RTX 3070 rumored specs: two sub $500 cards
- Want NVIDIA's next-gen Ampere GPU now? You can have it... for $12,500
- NVIDIA's Ampere GeForce RTX rumor: built on Samsung 8nm, not TSMC 7nm
- Say hello to the ASUS GeForce RTX 3080 Ti ROG STRIX, maybe
- NVIDIA's new A100 PCIe accelerator: 40GB HBM2e memory, PCIe 4.0 tech
- This new GeForce RTX 3090 leak has it at 26% faster than RTX 2080 Ti
- New GeForce RTX 3090 leaks: 12GB GDDR6X at insane 21Gbps
- GeForce RTX 3080 and RTX 3090 rumored to pack 'traversal coprocessor'
- NVIDIA's next-gen GeForce RTX 3080, RTX 3090 to enter production soon
- GeForce RTX 3090: GA102 consumes 230W, 24GB GDDR6 consumes 60W power
- NVIDIA rumored to use HUGE cooling block on GeForce RTX 3080
- NVIDIA GeForce RTX 3080 cooler: rumored to cost $150 on its own
- GeForce RTX 3090 rumor: 24GB GDDR6X, would annihilate RTX 2080 Ti
- NVIDIA GeForce RTX 3090 rumors: up to 60-90% faster than RTX 2080 Ti
- Check out these awesome renders of NVIDIA's next-gen GeForce RTX 3080
- This could be our first picture of the GeForce RTX 3080 graphics card
- AMD and NVIDIA to both launch next-gen GPUs in September 2020
- NVIDIA reportedly drops Tesla brand, too close to Elon Musk's Tesla
- NVIDIA amps up fight against COVID-19 with Ampere-based supercomputer
- NVIDIA DGX A100: 8 x A100 Ampere GPUs, AMD CPU, 15TB NVMe SSD
- NVIDIA Ampere A100 specs: 54 billion transistors, 40GB HBM2, 7nm TSMC