New Memory Bonding Technique Shows Promise
In a new cross-disciplinary effort, researchers have developed a novel approach to attach bonding wires to stacks of memory chips. The new technique, being called a “breakthrough” by its developers, promises to allow chips to be stacked to several times their current 8-chip and 16-chip heights.
At issue is the challenge of precisely bonding wires a fraction of the diameter of a human hair over great distances without their inadvertently coming into contact with their neighbors to create a short circuit. Such a short could destroy one or more of the chips in the stack, rendering the entire stack useless. The mechanical means of attaching these wires, although highly sophisticated, still has significant issues, that limit the economics of higher stacks.
Researchers at the Berea University of Geology (BUG) in Berea, Kentucky, noticed that certain members of the animal kingdom are capable of very precisely manipulating fine fibers and attaching them to a variety of surfaces. To take advantage of this ability, the university bred spiders to attach bonding wires to stacked memory chips.
“It is very difficult to train a spider,” said BUG’s lead researcher Peter Parker, who has been working over the past 32 years to develop this technique. “We have found that it is more reasonable to find those that already have an inclination to work with wire and selectively breed them in order to refine the process.”
The biggest difficulty with breeding spiders to manipulate wires rather than their own webs is their penchant for laying their eggs in a nest made of metallic filaments. “It’s impossible for the young to emerge from the nest,” says Parker, “so we have to remain with them day and night until they hatch to promptly cut the nest open and free them.” Since the spiders he is working with, a cross between the American black widow (viburnum prunifolium) and a brown recluse (loxosceles reclusa sicariidae), only live for about two weeks, this becomes something of a challenge. “The benefits are so great,” Parker asserts, “that we should get past this hurdle to change the entire semiconductor industry.”