All major PC vendors are set to launch their next-generation components towards the end of this year. These include chipmakers such as AMD, Nvidia and Intel, who are going to be deploying their in-house proprietary 7nm node for its 13th generation of CPU. Despite this, both AMD and Nvidia will continue to outsource their chip production to TSMC, the world’s leading semiconductor manufacturing company.
Last GPU cycle, however, Nvidia took a 50/50 approach to its outsourcing by relying on both Samsung and TSMC to produce the chips for its RTX 30 series GPUs. This time around, it looks like Nvidia will be relying solely on TSMC for fabricating both the 4nm and 5nm chips that will comprise the H100 and RTX 40 series respectively.
AMD will also be calling solely upon TSMC’s services to fabricate chips for the EPYC Genoa and Ryzen 7000 series CPUs, not to mention the RX 7000 RDNA 3 based GPUs. All of these are to be processed on the same 5nm node, as to be expected.
Interestingly enough, whilst Intel is providing proprietary 7nm nodes for its 13th generation Raptor Lake CPUs, Intel will be outsourcing chip production for its new 1st Gen Alchemist GPUs (2nd and 3rd Gen too) to TSMC to be fabricated on the N6 node.
Over the past couple of years as Hardwaretimes reports, “semiconductor shortages have been the primary bottleneck of industry growth and revenue”. However, with TSMC leading the charge in semiconductor production, they refuse to let history be repeated.
TSMC is apparently going to increase the production output of the 5nm wafer by 25% from 120K to 150K wafers (per month), to combat the horrors that ensued with the 5000 series and RTX 30 series launches respectively. It’s almost been two years since the release of the RTX 30 series and prices still haven’t returned to MSRP, although we are taking steps in the right direction.
This increase is fantastic news for PC building enthusiasts and the power-hungry avid gamer, as there should be a plentiful supply of RTX 40 series and AMD Ryzen 7000 series towards the back end of 2022 and the start of 2023.
What does this mean for Intel?
This is the first news we have had in relation to companies making an effort to combat shortages and prevent the crisis of the last couple of years, and it’s very well received if it is to be true.
We only hope Intel has the facilities to keep up with TSMC when producing its proprietary 7nm node for Raptor Lake. Intel doesn’t benefit at all in this situation, since its Alchemist GPUs are built on the N6 node, they will not see a production increase. Intel already seems to be starting the launch of next-gen on the back foot.