# US 2012/0254877 A1 · Transferring architected state between cores

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2012/0254877 A1
- Document ID: US20120254877A1
- Status: Published application
- Kind: A1
- Application number: 13/078,263
- Filed: 2011-04-01
- Published: 2012-10-04
- Earliest claimed: 2011-04-01
- Assignees: International Business Machines Corporation
- Inventors: Miguel Comparan; Russell D. Hoover; Robert A. Shearer; Alfred T. Watson, III
- Web page: https://patentgrep.com/patent/US20120254877A1

## CPC

- G06F 9/30043: {LOAD or STORE instructions; Clear instruction}
- G06F 9/3009: {Thread control instructions}
- G06F 9/46: Multiprogramming arrangements
- G06F 9/4856: {resumption being on a different machine, e.g. task migration, virtual machine migration (G06F9/5088 takes precedence)}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A method and apparatus for transferring architected state bypasses system memory by directly transmitting architected state between processor cores over a dedicated interconnect. The transfer may be performed by state transfer interface circuitry with or without software interaction. The architected state for a thread may be transferred from a first processing core to a second processing core when the state transfer interface circuitry detects an error that prevents proper execution of the thread corresponding to the architected state. A program instruction may be used to initiate the transfer of the architected state for the thread to one or more other threads in order to parallelize execution of the thread or perform load balancing between multiple processor cores by distributing processing of multiple threads.

## Claims

### Claim 1

A method of transferring architected state, comprising:
gathering at least a portion of the architected state associated with a thread executing within a source processing core;
transmitting the portion of the architected state directly from the source processing core to a target processing core over a dedicated interconnect that bypasses any memory shared between the source processing core and the target processing core; and
storing the portion of the architected state by the target processing core.

### Claim 2 (depends on 1)

The method of claim 1, further comprising initiating the transfer of the portion of the architected state when an error is detected by circuitry within the source processing core.

### Claim 3 (depends on 1)

The method of claim 1, further comprising initiating the transfer of the portion of the architected state when an instruction specifying the target processing core is received by the source processing core.

### Claim 4 (depends on 1)

The method of claim 1, further comprising:
initiating the transfer of the portion of the architected state when an instruction specifying the target processing core and an additional target processing core is received by the source processing core;
transmitting the portion of the architected state directly from the source processing core to the additional target processing core over the dedicated interconnect; and
storing the portion of the architected state by the additional target processing core.

### Claim 5 (depends on 1)

The method of claim 1, wherein the portion of the architected state is defined by an address range specified by an instruction that is received by the source processing core.

### Claim 6 (depends on 1)

The method of claim 1, wherein the target processing core is configured as a spare processing core that accepts the portion of the architected state.

### Claim 7 (depends on 1)

The method of claim 1, further comprising:
initiating the transfer of the portion of the architected state when an instruction specifying a count of target processor cores is received by the source processing core; and
transmitting the portion of the architected state directly from the source processing core to additional target processor cores over the dedicated interconnect, wherein the quantity of the additional target processor cores and the target processing core equals the count.

### Claim 8

A system, comprising:
a plurality of processing cores, wherein each processing core in the plurality of processing cores is configured to gather at least a portion of architected state associated with a thread executing within the processing core and transmit the portion of the architected state to a different processing core in the plurality of processing cores;
a memory that is shared between the plurality of processing cores; and
a state transfer interconnect that is configured to transmit the portion of architected state between a first processing core and a second processing core included in the plurality of processing cores and bypass the memory.

### Claim 9 (depends on 8)

The system of claim 8, wherein the first processing core configured to initiate transfer of the portion of the architected state when an error is detected by circuitry within the first processing core.

### Claim 10 (depends on 8)

The system of claim 8, the first processing core configured to initiate transfer of the portion of the architected state when an instruction specifying the second processing core is received by the first processing core.

### Claim 11 (depends on 8)

The system of claim 8, wherein the state transfer interconnect separately couples each one of the processing cores in the plurality of processing cores to the other processing cores in the plurality of processing cores.

### Claim 12 (depends on 8)

The system of claim 8, wherein the state transfer interconnect couples each one of the processing cores to a first neighboring processing core and a second neighboring processing core in the plurality of processing cores to form a ring topology.

### Claim 13 (depends on 8)

The system of claim 8, wherein the second processing core is configured as a spare processing core that accepts the portion of the architected state.

### Claim 14

A system, comprising:
a memory that is shared between a source processing core and a target processing core; and
a processor that is coupled to the memory and includes the source processing core and the target processing core,
wherein the source processing core is configured to:
gather at least a portion of architected state associated with a thread executing within the source processing core; and
transmit the portion of the architected state directly from the source processing core to a target processing core over a dedicated interconnect that bypasses the memory, and
the target processing core is configured to store the portion of the architected state.

### Claim 15 (depends on 14)

The system of claim 14, wherein the source processing core is further configured to initiate the transfer of the portion of the architected state when an error is detected by circuitry within the source processing core.

### Claim 16 (depends on 14)

The system of claim 14, wherein the source processing core is further configured to initiate the transfer of the portion of the architected state when an instruction specifying the target processing core is received by the source processing core.

### Claim 17 (depends on 14)

The system of claim 14, wherein the source processing core is further configured to:
initiate the transfer of the portion of the architected state when an instruction specifying the target processing core and an additional target processing core is received by the source processing core; and
transmit the portion of the architected state directly from the source processing core to the additional target processing core over the dedicated interconnect, and
the additional target processing core is further configured to store the portion of the architected state.

### Claim 18 (depends on 14)

The system of claim 14, wherein the portion of the architected state is defined by an address range specified by an instruction that is received by the source processing core.

### Claim 19 (depends on 14)

The system of claim 14, wherein the target processing core is configured as a spare processing core that accepts the portion of the architected state.

### Claim 20 (depends on 14)

The system of claim 14, wherein the source processing core is further configured to:
initiate the transfer of the portion of the architected state when an instruction specifying a count of target processor cores is received by the source processing core; and
transmit the portion of the architected state directly from the source processing core to additional target processor cores over the dedicated interconnect, wherein the quantity of the additional target processor cores and the target processing core equals the count.
