You did exactly what predicted and %100 wrong. I posted explanations why it would take at least 12-18 months for any new equipment qualification and it was not disputed.
Here they are...
https://defence.pk/pdf/threads/excl...al-chip-suppliers.666451/page-6#post-12331330
https://defence.pk/pdf/threads/excl...al-chip-suppliers.666451/page-8#post-12333562
Divestiture of established hardware is not as simple as ignorant and non-experience people like you think it is.
When I say 'new', it is not meant to be literally new on the market, never seen before. The word 'new' mean: Never used in production capacity in the company current process before. But I will use a smaller example to primarily inform the silent readers out there and less to educate resistant fools like you.
Take this
USED semicon related component selling on ebay...
https://www.ebay.com/itm/93k-Verigy-Probe-Card-Stiffener-12-WPI-E7018LA-/323567626612
If I buy this for my own site and to modify it to suit my own needs, at that time, the tester, or the word 'tool' as some company would say, would be immediately disqualified from Production use. This item stabilizes the probe card against seismic (vibration) influences and secure card planarity during operation. We are dealing with atomic scaling so even footsteps one meter away can knock the probe card out of alignment, losing (failure to test) some dies or even zero yield the entire wafer.
To re-qualify the tester, I would have to first run engineering wafers then normal Production wafers thru the tester. I have to review the data with the Equipment Support Engineer. If no deviations from normal Production wafers are found, the tester is released to Production use. The entire re-qualification process, assuming no major issues are found, will have the tester at least one day out of capacity. In that time, Planning/Shipping will adjust their capacity projections at least 30 days into the future. For one tester, it will be a minor adjustment. If the tester was running DRAM, another tester that was used to run NAND or wafer level reliability (WLR) testing can take its place. One day out of Production use affected 30 days of capacity projection.
Even
ONE modification rendered the current tester as figuratively
'NEW' or
'SUSPECT' hardware. Think about that.
The (re)qualification and/or correlation process is completely transparent to the clients. Generally, the clients (plural) will inquire the company as to what involves suspect hardware and the company will say 'We do A, B, C, and so on'. If one client reject, the company can either create a custom process for that client, or create a new process that will affect all clients, or tell that one client that nothing can be done. The company can also do nothing and lie to the clients.
At my company, I designed the Probe hardware qualification/correlation process. The Equipment group can make any modification or bring in any suspect hardware, but the moment they are ready, the suspect hardware falls to me and my group. We keep 10 yrs documentation on everything as to what happened and what was done to be in ISO compliance and we are ISO audited twice per yr.
Just in case you think I made up the 'WLR' bit...
https://www.mpi-corporation.com/ast/applications/wlr/
http://www.lricks.com/wlrt.htm
https://www.tek.com/document/brochure/wafer-level-reliability-systems
At the Production level, anything that can affect wafer yield will require WLR testing. It means: Does that item, hardware or software, affects the die over its projected lifespan? For this particular probe card item, WLR testing is not required. But if the item is an entire tester or a new test program revision, we will take some Production wafers that tested under the new item and run those wafers thru WLR regimes. It means higher amplitude, voltage, and temperatures. We will stress test literally to die structures destruction. If the test data is within statistical parameters, the new item, suspect hardware or software, is released to Production use.
There are three PDF Chinese members who claimed to have semicon experience. Most likely they will remain silent on what I said, or they will make some minor comments but not really make any correction. So if they keep silent, either everything I said is true in principles and that they could not find anything wrong, or they are frauds. We know of at least one fraud among the Chinese supporters who claimed to have 'aviation experience'.
Now here you are telling the world that it is simple for SMIC to switch over to new hardware in the entire fab processes, from wafer start to Probe (my dept)? Without
NEGATIVELY affecting their output and their customers demands?
Ultimately, SMIC can do whatever it want to support Huawei, but I will not allow liars and propagandists (like you) to mislead the silent readers out there as to the technical truths and their consequences.