Late last year, John Pitzer, Intel's vice president of corporate planning and investor relations, said that the yield rate of Intel's 18A process is now increasing at a rate of 7% per month. This growth rate is in line with the industry's expectations for a healthy climbing rhythm, and Panther Lake is also on this curve, giving Intel confidence in mass production.
According to TECHPOWERUP, after several months of hard work, Intel seems to have solved the yield problems encountered in the early Intel 18A process, and the yield rate will continue to and predictably improve. If the information is accurate, it means Intel 18A wafers are now more consistent, reducing batch-to-batch variation and making production more predictable. However, this does not mean that the defect density has reached the target level, the parameter yield has reached the optimal level, or that the overall economics has reached the level expected by Intel. It only shows that with continued improvement in yield, the mass production target is expected to be achieved within the foreseeable time.
There are currently two Intel 18A production lines, namely Fab 52 in Octillo Park, Arizona, and another factory in Hillsboro, Oregon. The total monthly production capacity is about 30,000 wafers, which is a good result in the current production ramp-up stage. Intel has started risk production of the Intel 18A-P process at its D1X factory in Hillsboro, Oregon, and plans to move volume production to Fab 62 at its Octillo campus.
In addition, the early samples of Intel 14A also show good prospects. Intel plans to start mass production at the D1X factory in 2029, and then transfer to the new wafer fab at the Ohio One production base in Ohio in the second phase, which will probably wait until 2030 to 2031.