When writing about the GTX TITAN Z, it must be displayed in all caps. When talking about the GTX TITAN Z – even in a library or during a play – it must be roared out, so loud that the paint on the walls peels and delicate people faint. It’s Nvidia’s new GPU and it’s stupidly powerful.
The GPU is built around two Kepler GPUs with 12GB of dedicated memory – the goal? To fuel next-gen 4K and multi-monitor gaming. Secondary to that is its goal to empty wallets, because this mighty chunk of hardware will set you back about $3,000.
“If you’re in desperate need of a supercomputer that you can fit under your desk, we have just the card for you,” Nvidia CEO Jen-Hsun said. My current rig – which also fits under my desk – is not a supercomputer. It’s not been a problem until now. But I’m starting to resent it.
The TITAN Z is two GK110 GPUs, with both halfs pretty much being a GTX 780 Ti or equivalent. NVIDIA has said it will have 8 TFLOPS of FP32 performance, putting the GPU clockspeed at about 700MHz. Unlike tradition dual-GPU cards, Nvidia has explained, the TITAN Z’s GPUs are tuned to run at the same clockspeed with dynamic power balancing. The result being that neither GPU creates a bottleneck for performance.
It’s impressive, sure, but despite Nvidia’s marketing of the card as gamer hardware, it seems a bit much to expect anyone but the most specialist and flushest of users to drop such a vast amount on it. The GTX TITAN series was already pushing it a bit when it cost $1,000, but the TITAN Z is three times that.
Of course, there are some enthusiasts that will be willing to splash out any amount of cash to have the most advanced, future-proof rig imaginable. This is for them. Damn them and damn my poverty.