Mechanism for enabling full data bus utilization without increasing data granularity
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06F-012/00
G06F-003/06
G11C-007/10
출원번호
US-0965845
(2015-12-10)
등록번호
US-9417800
(2016-08-16)
발명자
/ 주소
Garrett, Jr., Billy
출원인 / 주소
Rambus Inc.
인용정보
피인용 횟수 :
0인용 특허 :
82
초록▼
A memory is disclosed comprising a first memory portion, a second memory portion, and an interface, wherein the memory portions are electrically isolated from each other and the interface is capable of receiving a row command and a column command in the time it takes to cycle the memory once. By int
A memory is disclosed comprising a first memory portion, a second memory portion, and an interface, wherein the memory portions are electrically isolated from each other and the interface is capable of receiving a row command and a column command in the time it takes to cycle the memory once. By interleaving access requests (comprising row commands and column commands) to the different portions of the memory, and by properly timing these access requests, it is possible to achieve full data bus utilization in the memory without increasing data granularity.
대표청구항▼
1. A controller to control the operation of a dynamic random access memory (DRAM) having at least first and second memory portions, wherein each of the first and second memory portions includes a plurality of memory banks, the controller comprising: logic to control the timing of access requests gen
1. A controller to control the operation of a dynamic random access memory (DRAM) having at least first and second memory portions, wherein each of the first and second memory portions includes a plurality of memory banks, the controller comprising: logic to control the timing of access requests generated by the controller, the access requests to be sent to the first and second memory portions, the access requests timed to include commands interleaved between the first memory portion and the second memory portion such that a delay time between consecutive access requests to banks within any one of the first and second memory portions is longer than a time between consecutive access requests to banks within different ones of the first and second memory portions; anda port to send the access requests from the controller to the DRAM. 2. The controller of claim 1, wherein, for each access request, the port sends an indication to the DRAM of which of the first and second memory portions is selected for the access. 3. The controller of claim 1, wherein each of the access requests is a row command which specifies a row for activation within a memory bank of the selected one of the first and second memory portions. 4. The controller of claim 1, wherein the first memory portion is electrically isolated from the second memory portion, and wherein the DRAM comprises a port to receive the access requests from the controller. 5. The controller of claim 1, wherein each of the access requests includes a row command followed by a column command, wherein the row command specifies a row for activation within a memory bank of the selected one of the first and second memory portions, and the column command specifies which storage cells are to be accessed within the row. 6. The controller of claim 1, wherein consecutive access requests to banks in different ones of the first and second memory portions are transferred during consecutive clock cycles of a clock signal. 7. The controller of claim 6, wherein consecutive access requests to banks in any one of the first and second memory portions are transferred in accordance with a delay time that spans longer than a period of the consecutive clock cycles of a clock signal. 8. The controller of claim 1, wherein the access requests to the first and second memory portions are timed to include commands interleaved between the first memory portion and the second memory portion, such that each interleave can occur in an amount of time X, where X is less than or equal to an amount of time T that it takes to cycle the DRAM once, and wherein each access request includes: a row command that specifies a row for activation within a memory bank of the selected one of the first and second memory portions, anda column command specifies which storage cells are to be accessed within the row. 9. The controller of claim 1, wherein each of the access requests is a column command that specifies which storage cells are to be accessed within an activated row for a given selected bank within a selected one of the first and second memory portions. 10. The controller of claim 9, wherein each of the access requests further specifies one of a read operation and a write operation, each specified one of the read operation and the write operation being associated with distinctive ones of the column commands. 11. A method of operation in a controller, wherein the controller operates a dynamic random access memory (DRAM) having at least first and second memory portions, wherein each of the first and second memory portions includes a plurality of memory banks, the method comprising: sending to the DRAM, a plurality of indications for which memory portion, of the first and second memory portions, are to be selected for consecutive accesses; andtiming access requests to include commands interleaved between the first and the second memory portions; such that for each of the consecutive accesses to the first and second memory portions, a delay time between consecutive access requests to banks within anyone of the first and second memory portions is longer than a time between any access request to a bank within the first memory portion followed by an access request to a bank within the second memory portion. 12. The method of claim 11, wherein each of the access requests is a row command which specifies a row for activation within a memory bank of the selected one of the first and second memory portions. 13. The method of claim 11, wherein each of the access requests includes a row command followed by a column command, wherein the row command specifies a row for activation within a memory bank of the selected one of the first and second memory portions, and the column command specifies which storage cells are to be accessed within the row. 14. The method of claim 13, wherein the row command and the column command specify a base granularity of data. 15. The method of claim 11, wherein consecutive access requests to banks in different ones of the first and second memory portions are transferred during consecutive clock cycles of a clock signal. 16. The method of claim 11, wherein consecutive access requests to banks in any one of the first and second memory portions are transferred in accordance with a delay time that is spans longer than a period of consecutive clock cycles of a clock signal. 17. The method of claim 11, wherein each of the access requests comprises a column command, each column command to specify which storage cells are to be accessed within an activated row for a given selected bank within one of the first and second memory portions. 18. The method of claim 17, wherein each of the access requests further specifies one of a read operation and a write operation, each specified one of the read operation and the write operation being associated with a distinctive one of the column commands. 19. The method of claim 11, wherein banks included in the first memory portion are electrically isolated from banks included in the second memory portion, and wherein the DRAM receives the access requests at a port on the DRAM. 20. A controller to control the operation of a dynamic random access memory (DRAM) having at least first and second memory portions, wherein each of the first and second memory portions includes a plurality of memory banks, the controller comprising: means for controlling the timing of access requests to include commands interleaved between the first and second memory portions such that a delay time between consecutive access requests to banks within any one of the first and second memory portions is longer than a time between any access request to a bank within the first memory portion followed by an access request to a bank within the second memory portion; andmeans for sending the access requests to the DRAM, the access request including an indication to the DRAM of which of the first and second memory portions is selected for the access.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (82)
Chiu Edison H. (Richardson TX) Tai Jy-Der (Plano TX) Hsu Te-Chuan (Arlington TX), Architecture for memory multiplexing.
Yumitori Fuminori (Kawasaki JPX) Fujii Yasuhiro (Kawasaki JPX), Dynamic random access memory having sense amplifier control circuit supplied with external sense amplifier activating si.
Foss Richard C. (Kirkcaldy Fife GB6) Gillingham Peter B. (Kanata CAX) Harland Robert (Carp CAX) Mitsuhashi Masami (Gunma JPX) Wada Atsushi (Aichi JPX), Dynamic random access memory using imperfect isolating transistors.
Stamm Rebecca L. (Wellesley MA) Bahar Ruth I. (Lincoln NE) Strouble Raymond L. (Charlton MA) Wade Nicholas D. (Folsom CA) Edmondson John H. (Cambridge MA), Ensuring write ordering under writeback cache error conditions.
Bowater Ronald J. (Romsey NY GB2) Larky Steven P. (New York NY) St. Clair Joe C. (Round Rock TX) Sidoli Paolo G. (Romsey GB2), Flexible dynamic memory controller.
Baydar Ertugrul ; Boudreaux J. Bradley ; Carter Nicholas ; Chen Chung ; Klonsky Steven ; Moran Michael ; Renucci Peter ; Timbs Jeffrey ; Tucker Thomas ; Wardak Waleed, Integrated digital loop carrier system with virtual tributary mapper circuit.
Saito Yoshihiro (Itami JPX) Kittaka Yoshiaki (Itami JPX), Memory control unit and associated method for changing the number of wait states using both fixed and variable delay tim.
Matsuda Yoshio (Hyogo JPX) Fujishima Kazuyasu (Hyogo JPX) Hidaka Hideto (Hyogo JPX), Method and apparatus for driving word line in block access memory.
Stacovsky, Henry; Szabelski, Piotr, Methods and apparatus for reordering of the memory requests to achieve higher average utilization of the command and data bus.
Takemae Yoshihiro (Tokyo JPX), Random access memory device formed on a semiconductor substrate having an array of memory cells divided into sub-arrays.
Choi Yun-Ho (Kyungki-do KRX) Chin Dae-Je (Seoul KRX) Haq Ejaz U. (Seoul KRX) Cho Soo-In (Seoul KRX), Semiconductor memory device having a plurality of row address strobe signals.
Ginter Karl L. ; Shear Victor H. ; Sibert W. Olin ; Spahn Francis J. ; Van Wie David M., Systems and methods for secure transaction management and electronic rights protection.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.