PCB007 Magazine


Issue link: http://iconnect007.uberflip.com/i/1038326

Contents of this Issue


Page 108 of 125

OCTOBER 2018 I PCB007 MAGAZINE 109 line-based mixture. Perhaps the choline with the potential for a solvent assist is a more cost- effective process for photoresist stripping of outer layers. Meanwhile, what is the best way to control the performance of the resist stripper? Notice in Figure 2 that the stripper containing cho- line starts at a pH over 13. When the pH drops to 11.5, no choline remains, so all remaining stripping occurs with only the MEA. This is a handy way to control this stripping chemistry and a much better approach than the tradition- al titration, which does not address quality or strength of amine present and only looks at the overall quantity of alkalinity present. Figure 2 shows that controlling this type of chemistry by monitoring the pH, not by titra- tion, is the best way to ensure the quality of stripping. This process is set up to run the line until the minimum acceptable quality of strip- ping is observed, followed by checking the pH. This pH becomes the minimum measurement at which the bath gives acceptable results. As one can see, photoresist stripping is a blend of art and science. However, in the end, it is about science. More details about resist strip- ping to come in a future column. PCB007 Michael Carano is VP of technology and business development for RBP Chemical Technology. To read past columns or contact Carano, click here. vice's state dynamics to perform computational tasks in the memory itself, similar to how the brain's memory and processing are co-located. The second level draws on the brain's synaptic network structures as inspiration for ar- rays of phase change memory (PCM) devices to acceler- ate training for deep neural networks. Lastly, the dynamic and stochastic nature of neurons and synapses inspired the team to create a powerful computational substrate for spiking neural networks. IBM researchers designed a new computer architec- ture with co-located memory and processing. In studies, their prototype ran 200 times faster than conventional computers. Unlike the stovepipe components in conven- tional computers, the authors propose that brain-inspired computers could have coexisting processing and memo- ry units. The IBM team drew on three different levels of inspira- tion from the brain. The first level exploits a memory de- A New Brain-Inspired Architecture Could Improve How Computers Handle Data and Advance AI Figure 2: Chart showing pH as monitoring tool for MES/choline-based resist stripper.

Articles in this issue

Links on this page

Archives of this issue

view archives of PCB007 Magazine - PCB007-Oct2018