The $3K Titan V is the fastest graphics card, even though it’s not for gaming

The question now is whether and when Nvidia will release the product to retail, given the effort it expended to create this special edition Titan V card.

1. In return, you are not only getting the awesome new Volta GPU architecture «GV100», buyers also get 12 GB of HBM2 memory.
2. A post over at Reddit shows the card being tested in various benchmarks, beating out every graphics card made to date at stock clocks and delivering unprecedented amounts of performance when overclocked.
3. As for NVIDIA’s intended market of compute and AI users, the Titan V will be supported by NVIDIA GPU Cloud, which includes TensorRT, a number of deep learning frameworks, and HPC-related tools.
4. That bodes well for the true consumer Volta parts that we’ll eventually see some time in 2018.
5. The NVIDIA TITAN V graphics card features the latest 12nm Volta GPU architecture and as such, it is infused with the latest technologies that NVIDIA has to offer.
6. I dare say the idea of the «prosumer» Titan has died with this card, but for the rapidly growing professional compute market, this looks to be exactly the kind of card that a lot of developers have been waiting for.
7. One, I’m sure there is still room for improvement from the software side to extract every bit of the Volta’s super fast Tensor cores.

The use of the Tensor cores or CUDA cores in each SM is exclusive, however, so if you want access to the 110 TFLOPS of FP16 performance from the Tensor cores, that’s not in addition to the CUDA performance. The Tensor cores are also rather specialized, and while it’s possible to use them for certain gaming calculations, don’t count on that happening—Tensor code will need to be specifically written for the processors. The TimeSpy score is the easiest way to see the performance gains and it’s quite noticeable with a score crossing the 10,000 point mark. For reference, the machine used to test these cards is an i7-6700k, 16GB of RAM with Samsung 840 EVO SSD that is running on top of an MSI Z170A KRAIT GAMING motherboard.

When tested on Unigine Heaven’s 1440p preset, the Titan V managed an average of 126 frames per second. This places the new card at a raw performance increase of between 26 percent and 87 percent, more a marginal boost. Nvidia notes that Titan V sports a major overhaul of the streaming multiprocessor at the heart of the card, and it’s twice as energy-efficient as its Pascal predecessor. As mentioned, this is aimed at serious applications by scientific researchers in the fields of AI and high-performance computing in general, with a focus also on energy efficiency as well as power. This card has almost all the features of the 10K$ Tesla V100 at slightly lower performance.

NVIDIA TITAN V Design (Product Shots)

The NVIDIA TITAN V is based on the GV100 GPU architecture and features a total of 5120 CUDA cores and 320 texture units. In addition to the regular cores, the card also packs 640 Tensor Cores inside the Volta GPU. These are geared for maximum deep learning performance as the card can crunch up to 110 TFLOPs of GPU performance for AI related algorithms. Even with such hefty specs, the card only requires an 8 and 6 pin power connector configuration to boot and comes in a 250W package. When we last discussed the NVIDIA Titan V in our preview, it was only a few weeks after its surprise launch at the 2017 Neural Information Processing Systems conference. We came away with the understanding that the Volta-based Titan V was a new breed of NVIDIA’s prosumer line of video cards, one that essentially encapsulated NVIDIA’s recent datacenter/compute achievements and how they got there.

1. Taking a step back, the approach with Volta doesn’t mesh with NVIDIA’s previous approaches with Pascal and others.
2. NVIDIA has just announced their latest Titan graphics card based on the Volta GPU architecture, the Titan V.
3. Google designed these chips mainly for their in-house use, they aren’t planning to sell to consumers any time soon.
4. The Titan V also boasts 12GB of on-board HBM2 memory, with a total memory bandwidth of 652.8GB/s.
5. It represents a more significant leap than most products that have made that claim, however, as it’s the first consumer-grade GPU based around Nvidia’s new Volta architecture.
6. Yup, this is the first TITAN graphics card and also the first NVIDIA line of graphics cards (Non Quadro / Non Tesla) to feature HBM2 memory.

Future of Volta?

Trying to resist the urge to buy one of these, as my only DP capable GPUs are 2 Tesla K20c, and the Titan V is equivalent to at least 4 K20 GPUs in terms of 64-bit compute. Nvidia says the heart of their project is Jetson Xavier, the world’s first computer designed specifically for robotics. Nvidia CEO Jensen Huang has given away 20 Titan V CEO Edition GPUs to AI researchers.

So yeah, the card was faster in stock to stock comparisons and is just a whole another level when overclocked. When NVIDIA announced the graphics card yesterday, they did it in style, telling consumers that the card is ready to purchase and can be yours if you pay the hefty price of $3000 US. I noticed that many people (including our community readers) went ahead to purchase the graphics card on day one.

Add video cards to start comparisons

For NVIDIA, titan v price considering that Pascal has been around for over 2 years now, Volta is conspicuously absent from recent speculation over the next GeForce generation. In looking at the Titan V today, it almost seems that NVIDIA’s divergence is imminent. Even in the case of a Volta-based GeForce launch, the implementation of consumer Volta would be a very big hint at the future direction of GPUs, gaming and compute alike.

NVIDIA TITAN V Volta GPU and HBM2 Memory Powered Graphics Card Announced -…

Little bit of background info before we get into TensorCores, usually people used to train NNs on FP32 (Single Precision) but nowadays single precision also works fine for DL and has led to lot of improvements as Nvidia is exploiting it now. So, all the TeraFlops which they advertise should have a small asterix explaining the context. What they have really improved in this card is exploiting the FP16 by creating the Mixed Precision – a matrix multiplication in FP16 and accumulation in FP32 precision. To be more specific, each Tensor Core performs 64 floating point FMA mixed-precision operations per clock.

From what i’ve seen, its because they can leverage 8-bit calculations in inference mode as you dont need that high precisions. Needless to say, Nvidia’s Titan V spells great news for its upcoming line of Volta cards that are expected to arrive early next year. Fabricated on a new TSMC 12-nanometer FFN high-performance manufacturing process customized for NVIDIA, TITAN V also incorporates Volta’s highly tuned 12GB HBM2 memory subsystem for advanced memory bandwidth utilization. Darren is a freelancer writing news and features for TechRadar (and occasionally T3) across a broad range of computing topics including CPUs, GPUs, various other hardware, VPNs, antivirus and more. He has written about tech for the best part of three decades, and writes books in his spare time (his debut novel – ‘I Know What You Did Last Supper’ – was published by Hachette UK in 2013).

He holds a BS in Computer Science from Brigham Young University and has been working as a tech journalist since 2004, writing for AnandTech, Maximum PC, and PC Gamer. From the first S3 Virge ‘3D decelerators’ to today’s GPUs, Jarred keeps up with all the latest graphics trends and is the one to ask about game performance. We didn’t make Nvidia’s shortlist for an initial Titan V sample—»it’s not for gaming» was the curt response—but Falcon Northwest came through with a drool-worthy PC that kicks practicality to the curb and goes all-in on style and performance. The latest version of the company’s Tiki naturally includes the Titan V, but the luxurious hardware doesn’t stop there. The potent totem includes Intel’s fastest CPU for gaming, the Core i7-8700K, paired with 32GB of DDR CL14 RAM, and then forget about storage bottlenecks with not one but two Samsung 960 Pro 2TB drives, configured in RAID 0. Yes, it is a beast, yes, it is the high watermark for the industry but for the gaming crowd, it is not a good value in any way as the performance gains are not large enough over a 1080Ti at this time to justify the price tag.