Monday, December 3rd 2018

NVIDIA Presents the TITAN RTX 24GB Graphics Card at $2,499

NVIDIA today introduced NVIDIA TITAN RTX , the world's most powerful desktop GPU, providing massive performance for AI research, data science and creative applications. Driven by the new NVIDIA Turing architecture, TITAN RTX - dubbed T-Rex - delivers 130 teraflops of deep learning performance and 11 GigaRays of ray-tracing performance.

"Turing is NVIDIA's biggest advance in a decade - fusing shaders, ray tracing, and deep learning to reinvent the GPU," said Jensen Huang, founder and CEO of NVIDIA. "The introduction of T-Rex puts Turing within reach of millions of the most demanding PC users - developers, scientists and content creators."
Ultimate PC GPU
NVIDIA's greatest leap since the invention of the CUDA GPU in 2006 and the result of more than 10,000 engineering-years of effort, Turing features new RT Cores to accelerate ray tracing, plus new multi-precision Tensor Cores for AI training and inferencing. These two engines - along with more powerful compute and enhanced rasterization - enable capabilities that will transform the work of millions of developers, designers and artists across multiple industries.
Designed for a variety of computationally demanding applications, TITAN RTX provides an unbeatable combination of AI, real-time ray-traced graphics, next-gen virtual reality and high performance computing. It delivers:
  • 576 multi-precision Turing Tensor Cores, providing up to 130 teraflops of deep learning performance.
  • 72 Turing RT Cores, delivering up to 11 GigaRays per second of real-time ray-tracing performance.
  • 24GB of high-speed GDDR6 memory with 672GB/s of bandwidth - 2x the memory of previous-generation TITAN GPUs - to fit larger models and datasets.
  • 100GB/s NVIDIA NVLink can pair two TITAN RTX GPUs to scale memory and compute.
  • Incredible performance and memory bandwidth for real-time 8K video editing.
  • VirtualLink port provides the performance and connectivity required by next-gen VR headsets.
Built for AI Researchers and Deep Learning Developers
TITAN RTX transforms the PC into a supercomputer for AI researchers and developers. TITAN RTX provides multi-precision Turing Tensor Cores for breakthrough performance from FP32, FP16, INT8 and INT4, allowing faster training and inference of neural networks. It offers twice the memory capacity of previous generation TITAN GPUs, along with NVLink to allow researchers to experiment with larger neural networks and data sets.

Perfect for Data Scientists
A powerful tool for data scientists, TITAN RTX accelerates data analytics with RAPIDS. RAPIDS open-source libraries integrate seamlessly with the world's most popular data science workflows to speed up machine learning.
Content Creators Create Their Best Work
TITAN RTX brings the power of real-time ray tracing and AI to creative applications, so 5 million PC-based creators can iterate faster. It also delivers the computational horsepower and memory bandwidth needed for real-time 8K video editing.
Available This Month
TITAN RTX will be available later this month in the U.S. and Europe for $2,499.
Add your own comment

184 Comments on NVIDIA Presents the TITAN RTX 24GB Graphics Card at $2,499

#1
btarunr
Editor & Senior Moderator
ROFL

+5% performance over 2080 Ti for +150% price.
Posted on Reply
#2
Noztra
"Driven by the new NVIDIA Turing™ architecture, TITAN RTX — dubbed T-Rex "

Hahahahahaha.
Posted on Reply
#3
diatribe
Well that's what happens when you have a monopoly. I say the government needs to step in a chop nVidia into multiple companies like they have in the past with other monopolies.
Posted on Reply
#5
Vayra86
ELEVEN GIGA RAYS boys. Get in line. +6 FPS in BFV @ 1080p!

This one literally goes up to eleven!

<div class="youtube-embed" data-id="KOO5S4vxi0o"><img src="https://i.ytimg.com/vi/KOO5S4vxi0o/hqdefault.jpg" /><div class="youtube-play"></div><a href="https://www.youtube.com/watch?v=KOO5S4vxi0o" target="_blank" class="youtube-title"></a></div>
Posted on Reply
#6
Vya Domus
Actually I am surprised they chose to put 24 GB in this. I guess you can only get so far in pushing that exponential price increase without any added value.

diatribe said:
I say the government needs to step in a chop nVidia into multiple companies like they have in the past with other monopolies.
The government shouldn't do anything about it. I've always said it, the market is mostly self regulated.
Posted on Reply
#7
legoliveira
Less than 2500 USD? That's not cheap... but I'll give it a try!
Posted on Reply
#8
Cheeseball
It's not worth the price if you're just going to game with this thing.
Posted on Reply
#9
Emanulele
Is the shroud of the heatsink actual solid gold? (serious question)
Posted on Reply
#10
windwhirl
Well, if they are going to call it T-Rex, I hope someone will throw an asteroid at it.
Posted on Reply
#11
TheinsanegamerN
diatribe said:
Well that's what happens when you have a monopoly. I say the government needs to step in a chop nVidia into multiple companies like they have in the past with other monopolies.
Do you think government should be stepping in to regulate the prices of phones after all those iphones are over $1000 now, cant be allowing that right? The price of filet mignon? The price of diamonds? Oh, we should have the government break up mercedes, they have a monopoly on luxury RWD wagons. And dont forget coca-cola, they should be broken up too, they have a monopoly on coca cola after all!

We do not need governments regulating luxuries, we are not a communist society. The whole point of luxuries is that people with money can afford them. There is an argument to be had that the ever rising price of luxury GPUs is cranking the price on the rest, but there is an equally valid counterargument that this is only the result of a single GPU manufacturer being on top, if AMD were competitive prices would be no higher then the FERMI era. Nvidia GPUs are luxury items that currently have 0 competition, of course the prices are going to be ridiculous. That doesnt mean we need the government breaking the company up. Nvidia GPUs are luxuries, they are not necessary for life, your life is not negatively impacted by not having a titan GPU.
Posted on Reply
#12
jabbadap
Noztra said:
btarunr - It's not a gaming card....
So right you are. It's a card that one can game, but so is Quadros. Well even new piece has no single world about gaming and still only thing people care about is gaming performance difference to very expensive Geforce card.
Posted on Reply
#13
SIGSEGV
btarunr said:
ROFL

+5% performance over 2080 Ti for +150% price.
how much price of tesla v100? do they sacrifice tesla card?
576 multi-precision Turing Tensor Cores, providing up to 130 teraflops of deep learning performance.



I just feel pity to people/university/corporate who own tesla based HPC.
Posted on Reply
#14
diatribe
Vya Domus said:
The government shouldn't do anything about it. I've always said it, the market is mostly self regulated.
The market can't self-regulate Monopolies or Oligopolies than conspire to price fix.
Posted on Reply
#15
jabbadap
legoliveira said:
Less than 2500 USD? That's not cheap... but I'll give it a try!
It can do nvlink, buy two and the link only for $5077.99 :nutkick:
Posted on Reply
#16
Robcostyle
OH NO! Now 2080 Ti won't be the most powerfull GPU! What are we 2080 ti owners gonna do?!

ROFL, nice job, fanboys. Retardish maserrace lead to titans being priced as the whole top-tier PC, from now, on.
Keep buying this stuff at that prices - I beleive in you, make huang release Titan $$$ for 10000$!

P.S. However, I beleive, there's still decent amount of smart users - otherwise, it should have been released at 5000$ mark.
Posted on Reply
#17
TheinsanegamerN
diatribe said:
The market can't self-regulate Monopolies or Oligopolies than conspire to price fix.
You do not need governments regulating luxuries. Once you start doing that, the luxury companies refuse to service said market, and the citizens get pissed.

Big government isnt going to fix all of your problems. Sometimes you just have to say "im not going to buy that" and move on. If you are referring to the RAM situation, guess what, regulation didnt fix that, fines did! The US didnt start regulating the price of RAM magically, the RAM manufacturers were investigated and fined for price fixing.

There is none of that here. Nvidia is charging what it is because there is 0 competition. AMD cant compete with the 2080, or 2080ti, or titan RX. Even saying vega 64 is competitive with RTX 2070 is stretching it. So prices go up. It's no different then mercedes and their G-wagon, they have no real competition and thus the price is very high. You want lower GPU prices? Tell AMD to get off of their butts and make a new architecture FFS.
Posted on Reply
#18
Vayra86
SIGSEGV said:
how much price of tesla v100? do they sacrifice tesla card?
576 multi-precision Turing Tensor Cores, providing up to 130 teraflops of deep learning performance.



I just feel pity to people/university/corporate who own tesla based HPC.
How so? That segment is the only one getting a meaningful upgrade this generation. They should be happy, their TFLOPS actually got cheaper! On the gamer side, the opposite is true.
Posted on Reply
#19
Chloe Price
I guess that those deep web marketplaces will be full of kidneys..
Posted on Reply
#20
btarunr
Editor & Senior Moderator
SIGSEGV said:
how much price of tesla v100? do they sacrifice tesla card?
576 multi-precision Turing Tensor Cores, providing up to 130 teraflops of deep learning performance.



I just feel pity to people/university/corporate who own tesla based HPC.
Those have FP64 cores which this card lacks.
Posted on Reply
#21
Vya Domus
diatribe said:
The market can't self-regulate Monopolies or Oligopolies than conspire to price fix.
This has nothing to do with price fixing, the government should not intervene just because some company releases some absurdly high priced product, case closed. The consumers will decided if that's fine or not with their wallet. If you feel like that is the issue then take it to your friends and tell them to self-regulate their pockets.

And by the way Nvidia has done far, far worse things which were on the line of being outright illegal that genuinely hurt the industry. Those should be a concern to you not this.

btarunr said:
Those have FP64 cores which this card lacks.
They all have FP64 cores, not all of them are enabled to the same ratio as the rest of the ALU within the SM though.
Posted on Reply
#22
btarunr
Editor & Senior Moderator
Vya Domus said:
They all have FP64 cores, not all of them are enabled to the same ratio as the rest of the ALU within the SM though.
TITAN RTX has a grand total of 144 FP64 cores (negligible if your application really needs it). Tesla V100 has 2,560 FP64 cores.
Posted on Reply
#23
jabbadap
Vya Domus said:
This has nothing to do with price fixing, the government should not intervene just because some company releases some absurdly high priced product, case closed. The consumers will decided if that's fine or not with their wallet. If you feel like that is the issue then take it to your friends and tell them to self-regulate their pockets.

And by the way Nvidia has done far, far worse things which were on the line of being outright illegal that genuinely hurt the industry. Those should be a concern to you not this.

They all have FP64 cores, not all of them are enabled though.
Titan RTX has all the fp64 cores enabled that tu102 chip has. You can't enable something that does not physically exists.
Posted on Reply
#24
Vya Domus
btarunr said:
TITAN RTX has a grand total of 0 FP64 cores.
That is incorrect, TU102 has FP64 cores , which is the same chip the 2080ti is also based on. Unless of course Nvdia went of their way to disable all of them but leave them on the rest of the RTX line ? Even my 1060 has FP64 capability.

jabbadap said:
Titan RTX has all the fp64 cores enabled that tu102 chip has.
That's what I am saying.

Edit:

btarunr said:
TITAN RTX has a grand total of 144 FP64 cores
Well, so it's not 0 thanks. It's not negligible, a couple of hundred GFLOPs are still a lot compared to most CPUs out there when it comes to double precision floating point arithmetic.
Posted on Reply
#25
XXL_AI
btarunr said:
ROFL

+5% performance over 2080 Ti for +150% price.
try to use GTX/RTX level normal GPU for extensive workload such as deep learning training or machine learning analysis for a long, LONG time. then you'll see the difference.

I have 2 Titan X Pascals and they've been running for 1 year non-stop at fullpower for training and other simulations. no issues whatsoever but normal GTX 1080 failed within 4 months.
Posted on Reply
Add your own comment