NVIDIA announced its Ampere GPU architecture with the introduction of Ampere A100 throttle earlier this year, the company’s first 7nm GPU, and also its first PCIe 4.0 card. But what good is a next-generation GPU if it hasn’t been compared yet?
This is where Jules Urbach, the CEO of OTOY, a cloud graphics company famous for its Octane Render software, has compared NVIDIA’s new A100 accelerator. The new Ampere-based NVIDIA A100 accelerator was compared to OctaneBench, a benchmark designed to test performance on OctaneRender.
OTOY compared the GA100 based card consisting of 6912 CUDA cores and 40GB of super fast HBM2 memory.
Ampere is approximately 43% faster than Turing on OctaneRender, and that’s with RTX disabled. It will be impressive to see how those radically tuned Tensor and RT cores rip out Ampere silicon, compared to its Turing brothers.
More reading:
SEE GALLERY – 3 IMAGES
- Cross Coprocessor: We’ve had more leaks in NVIDIA’s next-generation GeForce RTX 3000 series than any previous graphics card family, with an interesting “cross-coprocessor” in the new GeForce RTX 3080 and GeForce RTX 3090 graphics cards. You can read more about it here.
- NVCache: Ampere is meant to have something called NVCache, which would be NVID’s own form of AMD’s HBCC (high-bandwidth cache controller, more about that here) NVCache would use your system’s RAM and SSD to speed up game load times as well as optimize VRAM usage. You can Read more about NVCache here.
- Memory compression tensor: NVCache is interesting, but Tensor Memory Compression will be in Ampere and will reportedly use Tensor Cores to compress and decompress items 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?? Quickly freaking out according to rumors, with 60-90% more performance than the current Turing-based flagship GeForce RTX 2080 Ti. We could see this huge performance jump in ray tracing titles, but we’ll have to wait a little longer to see how much graphics power NVIDIA has in these new cards. You can Read more about those rumors here.
- Hungry for power: As for power consumption, GA102 supposedly uses 230W, while 24GB of GDDR6X (which we should see in the new Ampere-based TITAN RTX) consumes 60W of power. You can read more about it here.
- Production starts soon.: NVIDIA is in the DVT (or Design Validation Test) range of its new GeForce RTX 3000 series graphics cards. Mass production reportedly begins 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-generation Ampere GPU architecture would be up to 75% faster than current-generation GPUs like Turing architecture, right after rumors that Ampere would offer 50% more performance at half the power of Turing. This is pretty crazy here.
Not only that, but we have some rumored specifications on the alleged GeForce RTX 3080 and GeForce RTX 3070 graphics cards, which will feature NVIDIA’s new Ampere GPU architecture.
We have already heard that Ampere would offer 50% more performance at half the power of Turing, which made my hair stand on end in my neck. Better yet, you can read about the leaked specs in the supposedly based on Ampere GeForce RTX 3080 and GeForce RTX 3070 here.
Even more reading:
- NVIDIA’s next-generation Ampere DLSS 3.0 could work in ANY game that uses TAA
- AMD Targets Big Navi Launch for November as ‘Show of Strength’ for RDNA 2
- The next generation of NVIDIA Ampere GeForce RTX 30: new 12-pin PCIe power
- AMD’s Next Generation RDNA 2 Rumor: 40-50% faster than GeForce RTX 2080 Ti
- NVIDIA Stops Manufacturing RTX 20 Series, Moves to GeForce RTX 30 Series
- GeForce RTX 3060 should cost $ 300- $ 400, here are some leaked specs
- NVIDIA combines Death Stranding with GeForce RTX graphics cards
- NVIDIA GeForce RTX 3070 Ti, RTX 3070 Rumored Specs – Two Cards Under $ 500
- Do you want the next-gen Ampere GPU from NVIDIA now? You can have it … for $ 12,500
- NVIDIA Ampere GeForce RTX rumor: built in Samsung 8nm, not TSMC 7nm
- Say hi to the ASUS GeForce RTX 3080 Ti ROG STRIX, maybe
- NVIDIA’s New PCIe A100 Accelerator – 40GB HBM2e Memory, PCIe 4.0 Technology
- This new GeForce RTX 3090 leak has it 26% faster than RTX 2080 Ti
- New GeForce RTX 3090 Leaks: 12GB @ 21Gbps GDDR6X
- GeForce RTX 3080 and RTX 3090 rumored to include ‘cross-coprocessor’
- The next generation of NVIDIA GeForce RTX 3080, RTX 3090 will go into production soon
- GeForce RTX 3090: GA102 consumes 230W, 24GB GDDR6 consumes 60W of power
- NVIDIA Rumored To Use Huge Cooling Block In GeForce RTX 3080
- NVIDIA GeForce RTX 3080 cooler: rumored to cost $ 150 on its own
- Rumor GeForce RTX 3090: 24GB GDDR6X, would annihilate RTX 2080 Ti
- NVIDIA GeForce RTX 3090 rumors: up to 60-90% faster than RTX 2080 Ti
- Check out these amazing next-gen renderings of NVIDIA’s GeForce RTX 3080
- This could be our first image of the GeForce RTX 3080 graphics card
- AMD and NVIDIA to Launch Next-Generation GPUs in September 2020
- NVIDIA reportedly drops Tesla brand too close to Elon Musk’s Tesla
- NVIDIA Expands Fight Against COVID-19 With Ampere-Based Supercomputer
- NVIDIA DGX A100: 8 x A100 Ampere GPU, AMD CPU, 15TB NVMe SSD
- NVIDIA Ampere A100 Specifications: 54 Billion Transistors, 40GB HBM2, 7nm TSMC
.
