1460881334-5ecaa1c8-124e-457f-9df6-858edb153359

1. A computer-implemented method, comprising:
receiving, by one or more processors, a query that includes a set of terms;
identifying, by one or more processors, at least a subset of the set of terms of the received query;
counting, by one or more processors, a number of occurrences per unit time of the particular subset of terms of the received query in a plurality of documents having news content;
graphing, by one or more processors, the number of occurrences of the particular subset of terms in the plurality of documents per unit time over a span of time;
identifying, by one or more processors, on the graph, a particular time at which a local maximum occurs, where the local maximum indicates a local maximum of the number of occurrences of the particular subset of terms;
extracting, by one or more processors, one or more headlines from one or more documents of the plurality of documents, where the extracted one or more headlines each include the particular subset of terms, where the one or more documents are associated with the particular time; and
labeling, by one or more processors, at least one point on the graph, that is associated with the particular time, with one of the extracted one or more headlines.
2. The method of claim 1, where labeling the at least one point comprises:
identifying a headline of the extracted headlines having selected words with a highest frequency of occurrence among a set of words of the extracted headlines; and
labeling the at least one point on the graph with the identified headline.
3. The method of claim 1, further comprising:
calculating a frequency of occurrence of selected words contained in the extracted headlines;
computing a score for each of the extracted headlines based on the frequency of occurrence associated with ones of the selected words included in each headline; and
selecting a headline from the extracted headlines with a highest of the scores.
4. The method of claim 3, where labeling the at least one point comprises:
labeling the at least one point with the selected headline.
5. The method of claim 1, further comprising:
generating, for each of the extracted one or more headlines, a score based at least in part on the number of occurrences of the particular counted term in the plurality of documents; and
ranking the extracted one or more headlines based on the generated score,
where labeling the at least one point on the graph includes labeling the at least one point on the graph with a highest ranked of the extracted one or more headlines.
6. The method of claim 1, where the local maximum is a first local maximum, the method further comprising:
identifying a shadow associated with the first local maximum;
identifying a second local maximum within the shadow of the first local maximum;
determining that the second local maximum is within the shadow of the first local maximum; and
deciding, based on the determining, not to label a point on the graph associated with the second local maximum.
7. The method of claim 1, where the local maximum is represented as a peak on the graph.
8. A system, comprising:
an interface to receive a query that includes a set of terms;
a memory to store information associated with a plurality of documents having news content;
one or more processing units to:
determine a number of occurrences per unit time of at least a portion of the query in the plurality of documents,
graph the number of occurrences per unit time over a span of time,
identify a particular time at which a peak occurs in the graph,
extract headlines from one or more of the plurality of documents, where the one or more documents are associated with the identified particular time, and
label at least one point on the graph associated with the identified particular time with one of the extracted headlines.
9. The system of claim 8, where each document in the plurality of documents is associated with a same timestamp.
10. The system of claim 8, where, when determining a number of occurrences per unit time of the at least a portion of the query in the plurality of documents, the processing unit is further to:
count the number of occurrences per unit time of at least the portion of the query in only the headlines of the plurality of documents.
11. A device comprising:
a memory to store:
first data comprising a first word from a plurality of headlines associated with a set of documents having news content that are hosted at distributed locations in a network, and
second data comprising a first numeric value associated with a number of occurrences of the first word in the plurality of headlines; and

one or more processors to:
plot a number of occurrences of a key, which includes the first word, over a span of time on a graph using the second data;

identify a particular time, on the graph, at which a local maximum occurs, where the local maximum indicates a local maximum of the number of occurrences of the key; and
label a point on the graph associated with the particular time, where the label indicates that the particular time is associated with one of the headlines of the plurality of headlines that is more relevant to the key than other headlines of the plurality of headlines, where the plurality of headlines are associated with a plurality of documents that are associated with the particular time.
12. The device of claim 11, where the memory is further to store:
third data comprising a second word from the plurality of headlines associated with the set of documents, and
fourth data comprising a second numeric value associated with a number of occurrences of the second word in the plurality of headlines, and
where the key further includes the second word.
13. A physical memory device that stores computer-executable instructions, comprising:
one or more instructions for receiving a query that includes a set of terms;
one or more instructions for identifying at least a subset of the set of terms of the received query;
one or more instructions for counting a number of occurrences per unit time of the particular subset of terms of the received query in a plurality of documents having news content;
one or more instructions for graphing the number of occurrences of the particular subset of terms in the plurality of documents per unit time over a span of time;
one or more instructions for identifying, by one or more processors, on the graph, a particular time at which a local maximum occurs, where the local maximum indicates a local maximum of the number of occurrences of the particular subset of terms;
one or more instructions for extracting one or more headlines from one or more documents of the plurality of documents, where the extracted one or more headlines each include the particular subset of terms, where the one or more documents are associated with the particular time; and
one or more instructions for labeling at least one point on the graph, that is associated with the particular time, with one of the extracted one or more headlines.
14. The physical memory device of claim 13, where each headline of the one or more headlines corresponds to a news headline.
15. The physical memory device of claim 13, where the one or more instructions for labeling the at least one point comprise:
one or more instructions for identifying a headline of the extracted headlines having selected words with a highest frequency of occurrence among a set of words of the extracted headlines; and
one or more instructions for labeling the at least one point on the graph with the identified headline.
16. The physical memory device of claim 13, where the computer-readable instructions further comprise:
one or more instructions for calculating a frequency of occurrence of selected words contained in the extracted headlines;
one or more instructions for computing a score for each of the extracted headlines based on the frequency of occurrence associated with ones of the selected words included in each headline; and
one or more instructions for selecting a headline from the extracted headlines with a highest of the scores.
17. The physical memory device of claim 16, where the one or more instructions for labeling the at least one point comprise:
one or more instructions for labeling the at least one point with the selected headline.
18. The physical memory device of claim 13, where the computer-executable instructions further comprise:
one or more instructions for generating, for each of the extracted one or more headlines, a score based at least in part on the number of occurrences of the particular counted term in the plurality of documents; and
one or more instructions for ranking the extracted one or more headlines based on the generated score,
where the one or more instructions for labeling the at least one point on the graph includes one or more instructions for labeling the at least one point on the graph with a highest ranked of the extracted one or more headlines.
19. The physical memory device of claim 13, where the local maximum is a first local maximum, where the computer-executable instructions further comprise:
one or more instructions for identifying a shadow associated with the first local maximum;
one or more instructions for identifying a second local maximum within the shadow of the first local maximum;
one or more instructions for determining that the second local maximum is within the shadow of the first local maximum; and
one or more instructions for deciding, based on the determining, not to label a point on the graph associated with the second local maximum.
20. The physical memory device of claim 13, where the local maximum is represented as a peak on the graph.
21. A system comprising:
a memory to store executable instructions;
a processor to execute the instructions, to implement:
means for receiving a query that includes a set of terms;
means for identifying at least a subset of the set of terms of the received query;
means for counting a number of occurrences per unit time of the particular subset of terms of the received query in a plurality of documents having news content;
means for graphing the number of occurrences of the particular subset of terms in the plurality of documents per unit time over a span of time;
means for identifying on the graph, a particular time at which a local maximum occurs, where the local maximum indicates a local maximum of the number of occurrences of the particular subset of terms;
means for extracting one or more headlines from one or more documents of the plurality of documents, where the extracted one or more headlines each include the particular subset of terms, where the one or more documents are associated with the particular time; and
means for labeling at least one point on the graph, that is associated with the particular time, with one of the extracted one or more headlines.
22. The system of claim 21, where the plurality of documents include documents including news content, and
where the one or more headlines correspond to news headlines included within the plurality of documents.
23. The system of claim 21, where each document in the plurality of documents is associated with a same timestamp.
24. The system of claim 21, where the means for labeling the at least one point comprises:
means for identifying a headline of the extracted headlines having selected words with a highest frequency of occurrence among a set of words of the extracted headlines; and
means for labeling the at least one point on the graph with the identified headline.
25. The system of claim 21, further comprising:
means for calculating a frequency of occurrence of selected words contained in the extracted headlines;
means for computing a score for each of the extracted headlines based on the frequency of occurrence associated with ones of the selected words included in each headline; and
means for selecting a headline from the extracted headlines with a highest of the scores.
26. A physical memory device that stores computer-executable instructions, comprising:
one or more instructions for providing an interface to receive a query that includes a set of terms;
one or more instructions for storing, in a memory, information associated with a plurality of documents having news content;
one or more instructions for determining a number of occurrences per unit time of at least a portion of the query in the plurality of documents;
one or more instructions for graphing the number of occurrences per unit time over a span of time;
one or more instructions for identifying a particular time at which a peak occurs in the graph;
one or more instructions for extracting headlines from one or more of the plurality of documents, where the one or more documents are associated with the identified particular time; and
one or more instructions for labeling at least one point on the graph associated with the identified particular time with one of the extracted headlines.
27. The physical memory device of claim 26, where each document in the plurality of documents is associated with a same timestamp.
28. The physical memory device of claim 26, where the one or more instructions for determining a number of occurrences per unit time of the at least a portion of the query in the plurality of documents comprise:
one or more instructions for counting the number of occurrences per unit time of at least the portion of the query in only the headlines of the plurality of documents.

The claims below are in addition to those above.
All refrences to claim(s) which appear below refer to the numbering after this setence.

1. A graphics processing unit, comprising:
a memory controller coupled to a local memory and configured to access data from the local memory that is organized within the local memory as one or more groups of blocks (GOBs), wherein each GOB includes eight sectors and four rows of data such that each row of data traverses four of the eight sectors; and
a display controller coupled to the memory controller and configured to access data from the local memory for display,
wherein the display controller is further configured to transmit a read request to the memory controller to access a first row of data from the local memory, the read request including a command field, a row field, an address field and a sector field, and
wherein the command field indicates a read or write memory access, the address field specifies a GOB within the local memory, the sector field specifies a sector within the GOB, and the row field specifies a row within the GOB, the sector being a vertical portion of the GOB and the row being a horizontal portion of the GOB.
2. The graphics processing unit of claim 1, wherein the memory controller is configured to also read a second row of data from the local memory in response to the read request and to transmit only the first row of data back to the display controller.
3. The graphics processing unit of claim 2, wherein the memory controller is configured to discard the second row of data read from the local memory.
4. The graphics processing unit of claim 2, further comprising a data path that couples the memory controller to the display controller, wherein the memory controller is configured to transmit data read from the local memory to the display controller through the data path, and the data path is sized such that only one row of data read from the local memory may be transmitted through the data path at a time.
5. The graphics processing unit of claim 1, wherein the first row of data includes four pixels, and each pixel is represented using four bytes.
6. The graphics processing unit of claim 1, wherein the intersection of the GOB, the sector within the GOB, and the row within the sector specifies the location of the data within the local memory for display.
7. A computing device, comprising:
a host memory;
a central processing unit coupled to the host memory; and
a graphics processing unit coupled to the central processing unit through a system interface, the graphics processing unit having:
a memory controller coupled to a local memory and configured to access data from the local memory that is organized within the local memory as one or more groups of blocks (GOBs), wherein each GOB includes eight sectors and four rows of data such that each row of data traverses four of the eight sectors, and
a display controller coupled to the memory controller and configured to access data from the local memory for display,
wherein the display controller is further configured to transmit a read request to the memory controller to access a first row of data from the local memory, the read request including a command field, a row field, an address field and a sector field, and
wherein the command field indicates a read or write memory access, the address field specifies a GOB within the local memory, the sector field specifies a sector within the GOB, and the row field specifies a row within the GOB, the sector being a vertical portion of the GOB and the row being a horizontal portion of the GOB.
8. The computing device of claim 7, wherein the memory controller is configured to also read a second row of data from the local memory in response to the read request and to transmit only the first row of data back to the display controller.
9. The computing device of claim 8, wherein the memory controller is configured to discard the second row of data read from the local memory.
10. The computing device of claim 8, further comprising a data path that couples the memory controller to the display controller, wherein the memory controller is configured to transmit data read from the local memory to the display controller through the data path, and the data path is sized such that only one row of data read from the local memory may be transmitted through the data path at a time.
11. The computing device of claim 7, wherein the first row of data includes four pixels, and each pixel is represented using four bytes.
12. The computing device of claim 7, wherein the intersection of the GOB, the sector within the GOB, and the row within the sector specifies the location of the data within the local memory for display.
13. A display controller configured to transmit a read request to a memory controller to access a first row of data from a local memory coupled to the memory controller, wherein the data is organized within the local memory as one or more groups of blocks (GOBs), wherein the GOB includes eight sectors and four rows of data such that each row of data traverses four of the eight sectors, and wherein the read request includes a command field, a row field, an address field and a sector field, wherein the command field indicates a read or write memory access, the address field specifies a GOB within the local memory, the sector field specifies a sector within the GOB, and the row field specifies a row within the GOB, the sector being a vertical portion of the GOB and the row being a horizontal portion of the GOB.
14. The display controller of claim 13, wherein the first row of data includes four pixels, and each pixel is represented using four bytes.
15. The display controller of claim 13, wherein the memory controller transmits data read from the local memory to the display controller through a data path, and the data path is sized such that only one row of data read from the local memory may be transmitted through the data path at a time.
16. The display controller of claim 13, wherein the read request further includes a command field set to indicate that the local memory is being accessed for a read operation.