Live data from Hacker News

Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

theverge.com

81–90 of 392 posts

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#81

Is anything known about the number of tensor cores in the new 2080 Ti? What about the float16 performance using the Tensor cores? I couldn't find this on the spec page [1]. [1] https://www.nvidia.com/en-us/geforce/graphics-cards/rtx-2080...

A couple of sites have said 384 TC's, which would be approximately ~50% of the v100/new high-end Quadro card. Would like to see a confirmed source as well.

Yeah, it seems at the moment that there's no confirmation on to what degree they're actually enabled either. It may well be that NVidia will make them available for inference via DirectX so they can do AI based denoising and anti aliasing in games but artificially limit support in frameworks for training so as not to eat into the Quadro / Tesla markets. I'm holding off on ordering one until I see more info on that.

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#82
post #66

Earlier quoted context omitted.

These will make a good card to experiment with. I am going to purchase one when I can. More RAM which is nice for larger models, but you can do a lot with 8GB of ram, especially if you are just getting going. Do not let worrying about having the latest and greatest hardware keep you from getting started. Time is the most important resource to have, not spec sheets. These cards look like they will have Tensor Cores, w…

> These cards look like they will have Tensor Cores, How they look like that? Do those cores support FP32 precision? I'm genuinely interested.

No, from INT4 to FP16. Inferencing can get nice boost.

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#83
post #78
post #6

2080 Ti FE RTX 2080 Ti GTX 1080 Ti Price $1,199 $999 $699 GPU Architecture Turing Turing Pascal Boost Clock 1635 MHz 1545 MHz 1582 MHz Frame Buffer 11 GB GDDR6 11 GB GDDR6 11 GB GDDR5X Memory Speed 14 Gbps 14 Gbps 11 Gbps Memory Interface 352-bit 352-bit 352-bit CUDA Cores 4352 4352 3584 TDP 260W 250W 250W Giga Rays 10 10 ? 2080 FE RTX 2080 GTX 1080 Price $799 $699 $549 GPU Architecture Turing Turing Pascal Boost Clo…

Your TDP figures are off * 2080 TI FE is 260W * 2080 TI is 250W https://www.nvidia.com/en-us/geforce/graphics-cards/rtx-2080...

Fixed, thanks!

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#84

Putting AI/ML aside, I have a question for graphics experts or/and gamers: what is the practical [1] ceiling of the current champion, the 1080ti? I was under impression that that card already was much more than most people needed. In other words, is the new generation only useful for 4K gaming? [1] By "practical" I mean on 1440p@144Hz monitors.

For me, it's around ~45-50fps@4k for a good set of games, and often ~100-120fps@144hz for me.

1000$ is a steep price for the 2080TI, although depending on the 2080 performance it might be the only card which can and will continue to reliably hit 60fps@4k.

Not upping ram with such a huge price increase is absolutely absurd.

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#85
post #6

2080 Ti FE RTX 2080 Ti GTX 1080 Ti Price $1,199 $999 $699 GPU Architecture Turing Turing Pascal Boost Clock 1635 MHz 1545 MHz 1582 MHz Frame Buffer 11 GB GDDR6 11 GB GDDR6 11 GB GDDR5X Memory Speed 14 Gbps 14 Gbps 11 Gbps Memory Interface 352-bit 352-bit 352-bit CUDA Cores 4352 4352 3584 TDP 260W 250W 250W Giga Rays 10 10 ? 2080 FE RTX 2080 GTX 1080 Price $799 $699 $549 GPU Architecture Turing Turing Pascal Boost Clo…

2080 Ti FE, 1259 Euros. All US companies follow this simple formula: $ to Euro 1:1, add 25% and then add 50-60 Euros for Free Shipping.

It's the 2-year warranty, translation of manuals and a need to have office in all major EU markets due to different legal stuff that makes it all pricier.

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#86

Earlier quoted context omitted.

Yeah, they were. NVidia's lesson from the Crypto-boom seems to be: "Some gamers are willing to pay >$1000 for their cards". EDIT: To be fair, NVidia is still on 14nm or 12nm class lithography (I forgot which one). So the increased die size of these new chips will naturally be more expensive than the 10xx series. Bigger chips cost more to produce after all. So if you fail to shrink the die, the economics demand that y…

Could you cite this? I've heard the opposite: that Nvidia didn't chase after the crypto market /because/ a crash would cause problems (if they overproduced). Besides, they make plenty of money everywhere else. Furthermore, Intel has charged (edit: consumers) $1000+ for a chip before. The market will bear it, crypto or no crypto.

Intel has $10k xeons but that's no consumer grade hardware.

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#87
post #6

2080 Ti FE RTX 2080 Ti GTX 1080 Ti Price $1,199 $999 $699 GPU Architecture Turing Turing Pascal Boost Clock 1635 MHz 1545 MHz 1582 MHz Frame Buffer 11 GB GDDR6 11 GB GDDR6 11 GB GDDR5X Memory Speed 14 Gbps 14 Gbps 11 Gbps Memory Interface 352-bit 352-bit 352-bit CUDA Cores 4352 4352 3584 TDP 260W 250W 250W Giga Rays 10 10 ? 2080 FE RTX 2080 GTX 1080 Price $799 $699 $549 GPU Architecture Turing Turing Pascal Boost Clo…

Wow, were the founders editions really that cheap on the last gen? Guess the inflated prices from all the crypto mining really threw me off.

I paid $699 for my 1080TI FE card in early 2017.

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#88
post #12

Hmm, 24%-42% performance increase (14/16TFlops vs 11.3TFlops) for 70-90% price increase... And prices were already inflated by crypto that is now collapsing. Not sure who is the target market for this tech honestly. Still only 11GB RAM, even if 50% faster, making it a nonsense purchase for Deep Learning enthusiasts (state-of-art models are already larger). Unless somebody invented RTX-based coin of course, then this…

You can always decrease a batch size to fit in smaller RAM.

Unless you are already on a batch size 1 and TensorFlow greets you with a nice OOM message... Try some Wide-ResNet for state-of-art classification and play with its widening parameter.

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#89

Earlier quoted context omitted.

Wow, were the founders editions really that cheap on the last gen? Guess the inflated prices from all the crypto mining really threw me off.

Yeah, they were. NVidia's lesson from the Crypto-boom seems to be: "Some gamers are willing to pay >$1000 for their cards". EDIT: To be fair, NVidia is still on 14nm or 12nm class lithography (I forgot which one). So the increased die size of these new chips will naturally be more expensive than the 10xx series. Bigger chips cost more to produce after all. So if you fail to shrink the die, the economics demand that y…

Balance between area, yields and tooling - a mature process with established tooling and strong yields can offset some of the additional wafer costs required by larger area.

Re: Nvidia announces RTX 2000 GPU series with ‘6x more performance’ and ray-tracing

#90
post #16

Does anyone know how they are offering better raytracing performance? Are there new api’s or something?

my understanding could be wrong, but I believe doing raytracing well requires better branching architecture, which traditionally gpus are pretty bad at.
Post reply on HN