摘要:
Disclosed is an article comprising a stacked array of electronic subassemblies, each subassembly including one or more integrated circuits (50) and a thermally conductive base member (15) that is perforated with holes (70). Motivating means (80) are provided for causing a coolant fluid to pass through the holes in a direction substantially parallel to the stacking axis.
摘要:
A new method for fabricating optical assemblies (OAs), which achieves increased throughput and reduced unit cost by mass producing optical subassemblies (OSAs), is disclosed. In accordance with this method, at least two OSAs are simultaneously produced by initially forming at least two corresponding sets of electrically conductive regions on the surface of a substrate. At least two optoelectronic devices are mounted on the surface of the substrate, with each device in electrical contact with one of the two conductor sets. Afterward, the substrate is separated into at least two parts, with each part including one of the optoelectronic devices and one of the conductor sets, constituting an OSA. Each OSA is then incorporated into an OA.
摘要:
The disclosed means for coupling an optical fiber (52) and an opto-electronic device (e.g., LED, laser, or photodetector) (40) comprises a first body (10) having two substantially parallel major surfaces, with a recessed portion (a "well") (11) formed in one surface, and a through-aperture extending from the other surface to the well. Conductive means (21, 24) extend from the former surface onto the bottom of the well, and the opto-electronic device typically is mounted in the well such that the device does not protrude above the plane of the associated surface, such that the electrical contact is established between the device and the conductive means, and such that the active region of the device is centered upon the through-aperture. The first body is advantageously produced from a (100) Si wafer by means of standard Si processing techniques, including selective etching. The assembly can be mounted on a substrate, e.g., a Si wafer with appropriate metallization thereon, and the end of an optical fiber inserted into the through-aperture and secured to the first body. The assembly can be operated at relatively high speed, due to its relatively low parasitic capacitance and inductance, and can be mounted on the substrate in substantially the same was as IC chips are mounted, in close proximity to associated electrical components.
摘要:
Disclosed is a novel "hermetic" optical assembly. A radiation-transmissive member (10) seals an aperture (22) in the assembly housing. The member comprises two major surfaces and a recess (12) that extends from one surface towards the other, with a "septum" (14) between the bottom of the recess and the other surface. An optical or opto-electronic device (23) typically is mounted on the inside surface of the member, and the end of an optical fiber is maintained outside of the assembly in coupling relationship with the device. In a currently preferred embodiment the recess extends from the outward-facing surface of the member and serves to position the end of the optical fiber with respect to a device mounted in a "well" (11) in the inward-facing surface of the member. In another embodiment, the outward facing surface of the member comprises geometrical features (e.g., recesses) that mate with corresponding features (e.g., protrusions) on a guide plate that serves to position the ends of a multiplicity of fibers with respect to devices mounted on the bottom of a recess extending from the inward-facing surface of the radiation-transmissive member. In all embodiments radiation is transmitted through the septum. In preferred embodiments the radiation-transmissive member is (100) Si, with the recess (and other features) produced by means comprising photolithography and anisotropic etching.
摘要:
A system is described for arraying multi-device processing nodes (1,2) in a 3-dimensional computing architecture and for flexibly connecting their ports. The topology of each processing node is of a fixed and constant physical geometry. The nodes may comprise a digital signal processor chip, a static RAM, and a communications and network controller. The nodes or tiles are 4-connected, each having logical north, east, south and west ports. The nodes are mounted on boards (17,18,19). Selective connection of essentially any one port to another on a different board is effected by use of a routing and spacer element (50) having internal routing paths preselected to support a desired node interconnection architecture. Novel multiprocessing architectures and connections to a Host computer are also disclosed.
摘要:
A system is described for arraying multi-device processing nodes (1,2) in a 3-dimensional computing architecture and for flexibly connecting their ports. The topology of each processing node is of a fixed and constant physical geometry. The nodes may comprise a digital signal processor chip, a static RAM, and a communications and network controller. The nodes or tiles are 4-connected, each having logical north, east, south and west ports. The nodes are mounted on boards (17,18,19). Selective connection of essentially any one port to another on a different board is effected by use of a routing and spacer element (50) having internal routing paths preselected to support a desired node interconnection architecture. Novel multiprocessing architectures and connections to a Host computer are also disclosed.