NVIDIA GeForce RTX 5060 graphics card displayed against an orange background.

NVIDIA’s Rubin Enters the Fab, Volume Ramp Set for 2H 2026 as Gaming Smashes Q2 Record with $4.3B

NVIDIA says its next-generation AI platform, Rubin, has already entered fabrication and is on track for volume production in the second half of 2026. The company shared the update during its Q2 2026 earnings call, pairing the roadmap news with record gaming revenue and a rapid ramp of Blackwell Ultra hardware in the data center.

Rubin represents NVIDIA’s third-generation NVLink rack-scale AI supercomputer. Leadership emphasized that this platform will benefit from a fully matured, large-scale supply chain and continue the company’s annual product cadence across compute, networking, systems, and software. Executives also framed Rubin and its successor platforms as key drivers of a multi-trillion-dollar AI infrastructure buildout over the coming years.

The Rubin platform consists of six new chips designed to work together:
– Rubin GPU
– Vera CPU
– CX9 SuperNIC
– NVLink 144 rack-scale switch
– Spectrum-X scale-out and scale-across switch
– Silicon photonics processor

According to the company, all Rubin components are now in fabs and remain on schedule, with a broad rollout expected to begin in 2026. Leadership hinted that Rubin will introduce “a whole bunch of new ideas,” with more details to come at a future GTC event.

While Rubin readies for prime time, Blackwell Ultra is already surging. The GB300 generation is in full production and demand is described as extraordinary. Thanks to a common architecture, software stack, and physical footprint with GB200, major cloud providers were able to transition smoothly to GB300 racks. Factory conversions completed in late July and early August, and production is back to roughly 1,000 racks per week, with output expected to accelerate further as additional capacity comes online in the third quarter.

Gaming also delivered a standout performance. Revenue reached a record $4.3 billion, up 14% sequentially and 49% year over year. NVIDIA credited the ramp of GeForce GPUs based on the Blackwell architecture and stronger availability of the latest GeForce RTX 50 Series.

All told, the company reported $46.7 billion in quarterly revenue. With Blackwell Ultra ramping at speed and Rubin preparing for a 2026 volume launch, NVIDIA sees continued momentum across AI infrastructure and consumer graphics. If the current trajectory holds, the next few quarters could bring even higher output in the data center while the gaming segment benefits from increased supply of next-gen GeForce hardware.