NVIDIA will be shifting more of its production capacity to Ampere series
ITHome reports on a post from Board Channel forums, closed forums for people working for distributors and board partners, that NVIDIA has allegedly recently adjusted its strategy when it comes to production capacity and supply.
It is said that NVIDIA has issued an internal notice to board partners (AICs) that it will be adjusting, in this case limiting, the supply of GeForce RTX 2060 graphics cards this month. This is still a very popular SKU in the NVIDIA lineup.
When RTX 30 series launched, NVIDIA attempted to increase the production of this Turing model, due to its popularity production cost. We have seen numerous resurrection attempts of this SKU with slightly modified, often simplified, cooling designs (in other words to make it cheaper). From NVIDIA’s perspective, this model, unlike RTX 2060 SUPER, features TU106 GPU, which is smaller and cheaper to make. It was indeed a good idea to tackle the global GPU shortage.
The internal notice on Board Channels forums allegedly says that NVIDIA will be limiting the supply of this model starting this month. The manufacturer will then shift production capacity to RTX 30 series, the message says. It is probably worth noting that TU106 is made by TSMC, while RTX 30 series GPUs are made by Samsung. But the production capacity likely refers to the whole supply chain, including GPUs, components, and memory bundles (often sold in bulk with GPUs at a lower price). Each of those things still stuffer from global shortages.
A launch of Lite Hash Rate models this month combined with the recent announcement of the Cryptominig Processor (CMP) series, should gradually help reduce the interest in GeForce RTX series from miners.