Nvidia has made a monster of a chip for its new Blackwell architecture. The B200 chip, which features the supercomputing generative AI named Blackwell, was recently revealed at the GTC Artificial Intelligence Conference. It’s a massive chip, not only in physical size but in power, too.
This massive chip was presented by Nvidia CEO Jensen Huang himself. The Blackwell B200 is a dual-GPU package, which features an astounding 208 billion transistors between the two GPUs. The B200 also features brand-new fifth-generation Tensor Cores, a step up by 30 times when it comes to performance to the fourth-generation Tensor Cores. These are some of the most impressive specs for a GPU ever. The Blackwell B200 is set to be a force of nature.
Nvidia’s impressive Blackwell B200 specs aren’t focused on gaming
However, for gamers, the news might not be so exciting. The B200 chip, along with its B100 counterpart, is designed not for gamers, but rather for enterprises. The Blackwell architecture aligns with Nvidia’s recent goal of improving AI and is meant for large businesses rather than single gaming PCs.
But, with that being said, it’s possible some of this hardware will find its way to Nvidia’s gaming GPUs. Gamers will hardly need most of the performance in the B200, but the new Tensor Cores are another story. While these instruction accelerators are mainly used in AI, the fourth-generation version was built into the Ada Lovelace architecture found in the RTX 40 series. When the next generation of Nvidia gaming GPUs comes out, it wouldn’t be a shock to see the fifth-generation Tensor Cores make the jump, as well.
While the Blackwell B200 is a marvel of technology, it will be a while before any of its impressive power makes its way to gamers. The price of the B200 chip will certainly keep gaming fans from getting their hands on it. Once the Blackwell architecture is more accessible, that is when we’ll likely see it on Nvidia gaming GPUs. Only time will tell for sure, though.
