If you were hoping for a surprise GeForce launch or unveil at E3 or Computex this year, Nvidia CEO Jen-Hsun Huang has some bad news. While the company is aggressively moving forward with new embedded hardware, including Jetson Xavier, a new, robotics-focused development platform with a Volta Tensor Core GPU, an eight-core ARMv8.2 CPU of undetermined vintage, and what Nvidia calls a “Vision Accelerator” (a 7-way VLIW processor), that platform is intended strictly for robotics development and isn’t coming to consumer platforms. And as for a new GPU, well, that’s not in the immediate cards either.
When asked when a next-generation GeForce might tip up, Nvidia CEO Jen-Hsun Huang responded with “It’s a long time from now.”
Now, the question is, is Jen-Hsun being honest, or playing a bit of a game? It’s always hard to call these situations, but if we had to guess, we’d say the cards still point to a GPU launch in the next few months. There are several reasons why. First, Jen-Hsun isn’t going to say anything to sabotage the sale of existing Pascal cards, which are still moving off store shelves briskly as the cryptocurrency market cools down and the gaming market likely picks up some steam as a result. It wouldn’t be surprising for GPU sales in Q2 to be above standard seasonality as a result of deferred demand from the holiday and spring, where prices were so extraordinary high end GPUs were selling for more money in the retail channel than the cost of an entire self-built gaming PC.
Rumors we’ve heard around the industry always pointed to two potential launch windows — either one this spring (proved false) or a later, mid-to-late summer launch that would position GPUs for the back-to-school and holiday cycles. The age of Pascal is itself another argument in favor of a new product family; at over two years old, Pascal is now downright toothy by GPU standards. In fact, Nvidia could wind up matching or breaking the record AMD set between the June 2015 launch of its Radeon Fury and the lengthy delay until Vega debuted in late 2017. Both companies, it seems, now see good reason to take their time in between GPU launches.
In the meantime, Nvidia is pushing ahead with its plans to bring a new series of massive, 65-inch, BFGD (Big Format Gaming Displays, obviously). These panels are 4K, HDR-enabled (still a rarity on PCs), support G-Sync, and are expected to sell for at least $2,000 according to data we’ve seen online. That actually might be at the lower end of the scale; price leaks have suggested a range between $2,000 and $4,000. Now, bringing all of these features — 4K, HDR, 65-inch screens, and G-Sync — together likely costs a pretty penny. But at the same time, I’m not sure if a panel priced that high is going to move the needle much with cutting-edge feature adoption.
As for when we might see a next-generation Nvidia launch, we’re still thinking end of the summer sounds most likely, but it’s not beyond the realm of possibility that the company could push things back farther. With relatively limited competition in the GPU market these days, Team Green simply isn’t under a lot of pressure to introduce a new round of consumer cards. We still think Pascal’s age will likely tip the scales in favor of an updated card, but if things should slip into the fall, it wouldn’t really represent a challenge to the status quo.
AMD Targets Nvidia’s Controversial GeForce Partner Program
AMD lines up new partners for RX Vega and takes on Nvidia's GeForce Partner Program in the process.
Nvidia Has Canceled the GeForce Partner Program
Nvidia has announced the end of the GeForce Partner Program, effective immediately.
Do FreeSync Displays, Nvidia GeForce GPUs Play Nice Together?
When Nvidia announced its support for FreeSync, it raised the question that some displays might not play nicely with its GPUs. Early testing suggests this may not be a major issue.
Custom GeForce GTX 1660 Ti Model Information Leaks From Multiple Retailers
Retailer leaks have confirmed that the GTX 1660 and GTX 1660 Ti are coming, but details on their specs and pricing are still pointing in different directions. This could reflect Nvidia's intent to support a variety of configurations with the same par, as it did with the GTX 1060.