-
公开(公告)号:DE112016000381B4
公开(公告)日:2020-06-18
申请号:DE112016000381
申请日:2016-02-15
Applicant: IBM
Inventor: COLGAN EVAN GEORGE , DENNEAU MONTY MONTAGUE , KNICKERBOCKER JOHN
IPC: H01L23/367 , G06F1/18 , G06F1/20 , H01L23/473 , H01L23/52
Abstract: Halbleiterstruktur, aufweisend:mindestens einen Halbleiterprozessor-Wafer, der starr an einer von Einheiten freien Seite an einem flüssigkeitsgekühlten Substrat angebracht ist, dessen Wärmeausdehnungskoeffizient dem des Halbleiterprozessor-Wafers gleich ist, wobei der Halbleiterprozessor-Wafer zwei oder mehr Chips enthält, die durch eine On-Chip-Verdrahtungsebene miteinander verbunden sind, wobei Substrate jedes Chips an einzelnen Chips auf einer Einheitenseite des Halbleiterprozessor-Wafers angebracht sind, wobei die Chip-Substrate eine kleinere Fläche einnehmen als die Chips auf dem Wafer, undeine oder mehrere an jedem Chip-Substrat angebrachte Karten, wobei eine Hauptfläche jeder Karte senkrecht zu einer Fläche des Halbleiterprozessor-Wafers steht.
-
公开(公告)号:DE112016000381T5
公开(公告)日:2017-10-05
申请号:DE112016000381
申请日:2016-02-15
Applicant: IBM
Inventor: COLGAN EVAN GEORGE , DENNEAU MONTY MONTAGUE , KNICKERBOCKER JOHN
IPC: H01L21/70
Abstract: Eine Halbleiterstruktur enthält ein Substrat mit Kühlschichten, Kühlkanälen, Kühlmittelzuleitungen und -ableitungen in Fluidverbindung mit den Kühlkanälen, und eine Einheitenschicht auf den Kühlschichten mit einem oder mehreren Verbindungspunkten und eine Einheitenschichtfläche. Der Wärmeausdehnungskoeffizient der Einheitenschicht ist im Wesentlichen gleich dem der Kühlschichten. Eine Mehrzahl von Laminatsubstraten sind auf der Einheitenschicht angeordnet und elektrisch mit dieser verbunden. Der Wärmeausdehnungskoeffizient des Laminatsubstrats ist von dem der Einheitenschicht verschieden, jedes Laminatsubstrat ist kleiner als ein Bereich der Einheitenschicht, auf dem es angebracht ist, und jedes Laminatsubstrat enthält Lücken zwischen Seiten benachbarter Laminatsubstrate. Die Laminatsubstrate sind über die Lücken zwischen ihnen hinweg nicht elektrisch oder mechanisch miteinander verbunden, und die Laminatsubstrate sind klein genug, um ein Aufwölben der Einheitenschichten, der Verbindungsschichten und der Kühlschichten aufgrund von Wärmeausdehnung zu verhindern.
-
公开(公告)号:GB2357412B
公开(公告)日:2003-12-10
申请号:GB0019066
申请日:2000-08-04
Applicant: IBM
IPC: G06T1/20
Abstract: The geometric processing of an ordered sequence of graphics commands is distributed over a set of processors by the following steps. The sequence of graphics commands is partitioned into an ordered set of N subsequences S0 . . . SN-1, and an ordered set of N state vectors V0 . . . VN-1 is associated with said ordered set of subsequences S0 . . . SN-1. A first phase of processing is performed on the set of processors whereby, for each given subsequence Sj in the set of subsequences S0 . . . SN-2, state vector Vj+1 is updated to represent state as if the graphics commands in subsequence Sj had been executed in sequential order. A second phase of the processing is performed whereby the components of each given state vector Vk in the set of state vectors V1 . . . VN-1 generated in the first phase is merged with corresponding components in the preceding state vectors V0 . . . Vk-1 such that the state vector Vk represents state as if the graphics commands in subsequences S0 . . . Sk-1 had been executed in sequential order. Finally, a third phase of processing is performed on the set of processors whereby, for each subsequence Sm in the set of subsequences S1 . . . SN-1, geometry operations for subsequence Sm are performed using the state vector Vm generated in the second phase. In addition, in the third phase, geometry operations for subsequence S0 are performed using the state vector V0. Advantageously, the present invention provides a mechanism that allows a large number of processors to work in parallel on the geometry operations of the three-dimensional rendering pipeline. Moreover, this high degree of parallelism is achieved with very little synchronization (one processor waiting from another) required, which results in increased performance over prior art graphics processing techniques.
-
公开(公告)号:GB2357412A
公开(公告)日:2001-06-20
申请号:GB0019066
申请日:2000-08-04
Applicant: IBM
IPC: G06T1/20
Abstract: The geometric processing of an ordered sequence of graphics commands is distributed over a set of processors by the following steps. The sequence of graphics commands is partitioned into an ordered set of N subsequences S 0 .. S N-1 , and an ordered set of N state vectors V 0 .. V N-1 is associated with said ordered set of subsequences S 0 .. S N-1 . A first phase of processing is performed on the set of processors whereby, for each given subsequence S j in the set of subsequences S 0 .. S N-2 , state vector V j+1 is updated to represent state as if the graphics commands in subsequence S j had been executed in sequential order. A second phase of the processing is performed whereby the components of each given state vector V k in the set of state vectors V 1 .. V N-1 generated in the first phase is merged with corresponding components in the preceding state vectors V 0 .. V k-1 such that the state vector V k represents state as if the graphics commands in subsequences S 0 .. S k-1 had been executed in sequential order. Finally, a third phase of processing is performed on the set of processors whereby, for each subsequence S m in the set of subsequences S 1 .. S N-1 , geometry operations for subsequence S m are performed using the state vector V m generated in the second phase. In addition, in the third phase, geometry operations for subsequence S 0 are performed using the state vector V 0 . Advantageously, the present invention provides a mechanism that allows a large number of processors to work in parallel on the geometry operations of the three-dimensional rendering pipeline. Moreover, this high degree of parallelism is achieved with very little synchronization (one processor waiting from another) required, which results in increased performance over prior art graphics processing techniques.
-
-
-