Uci e-based data transmission method, artificial intelligence chip, medium and electronic device

Abstract

The application provides a data transmission method based on UCIe, an artificial intelligence chip, a medium and an electronic device. The data transmission method based on UCIe comprises the following steps: communicating through a module end of a first core particle and a module partner end of a second core particle; when the module partner end receives a retraining trigger signal, resetting a circuit of the module partner end, and sending a retraining request signal to the module partner end through the module end; when the module partner end receives the retraining request signal, replying a retraining response signal to the module end through the module partner end; and when the module end receives the retraining response signal, making the module end enter a retraining state. The data transmission method based on UCIe, the artificial intelligence chip, the medium and the electronic device provided by the application can provide good data transmission quality.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of chips, in particular to a data transmission method based on UCIe, an artificial intelligence chip, a medium and an electronic device. BACKGROUND

[0002] Generally, auxiliary channels (such as clock signal lines) between core particles may cause glitches on the clock signal lines during the transmission of signals due to power supply noise and inter-line crosstalk, etc., thereby causing auxiliary channel message parsing errors. In this regard, for example, in the process of tracking real-time calibration or entering a retraining state, if auxiliary channel message parsing errors occur, the message interaction between the module end and the module partner end will timeout, thereby causing a core particle reporting interruption. SUMMARY

[0003] The present application is a data transmission method based on UCIe, an artificial intelligence chip, a medium and an electronic device, which can achieve good data transmission quality.

[0004] According to the embodiment of the present application, the data transmission method based on UCIe includes the following steps: communicating between the module end of the first core particle and the module partner end of the second core particle; when the module partner end receives a retraining trigger signal, resetting the circuit of the module partner end, and sending a retraining request signal to the module partner end through the module end; when the module partner end receives the retraining request signal, replying a retraining response signal to the module end through the module partner end; and when the module end receives the retraining response signal, making the module end enter a retraining state.

[0005] In the data transmission method based on UCIe according to the embodiment of the present application, the following steps are further included: when the module end enters an active state, making the counter start counting; and when the counter counting reaches a default value, in response to an auxiliary channel message parsing error, triggering an interruption and making the counter re-count.

[0006] In the data transmission method based on UCIe according to the embodiment of the present application, the following steps are further included: when the counter counting reaches the default value, sending a re-calibration tracking mode request signal to the module partner end through the module end; and when the module partner end parses the re-calibration tracking mode request signal incorrectly, triggering an interruption through the module partner end and making the counter re-count.

[0007] In the UCIe-based data transmission method according to the embodiment of the present application, the method further comprises the following steps: when the module partner end successfully parses the recalibration tracking mode request signal, the module partner end sends a recalibration tracking mode response signal to the module end; and when the module end fails to parse the recalibration tracking mode response signal, the module end triggers an interrupt and causes the counter to restart counting.

[0008] In the UCIe-based data transmission method according to the embodiment of the present application, the method further comprises the following steps: when the module partner end successfully parses the recalibration tracking mode request signal, the module partner end starts sending a tracking mode signal and a clock mode signal.

[0009] In the UCIe-based data transmission method according to the embodiment of the present application, the method further comprises the following steps: when the module end successfully parses the recalibration tracking mode response signal, the module end executes a tracking recalibration algorithm; and when the tracking recalibration algorithm ends, the module end sends a recalibration tracking mode completion signal to the module partner end.

[0010] In the UCIe-based data transmission method according to the embodiment of the present application, the method further comprises the following steps: when the module partner end fails to parse the recalibration tracking mode completion signal, the module partner end triggers an interrupt and causes the counter to restart counting.

[0011] In the UCIe-based data transmission method according to the embodiment of the present application, the method further comprises the following steps: when the module partner end successfully parses the recalibration tracking mode completion signal, the module partner end sends a recalibration tracking mode completion confirmation signal to the module end; and when the module end fails to parse the recalibration tracking mode completion confirmation signal, the module end triggers an interrupt and causes the counter to restart counting.

[0012] In the UCIe-based data transmission method according to the embodiment of the present application, the method further comprises the following steps: when the module end enters an active state, the module partner end directly starts sending a tracking mode signal and a clock mode signal to the module end and causes the counter to start counting; when the counter counts to a default value, the module end executes a tracking recalibration algorithm; and when the tracking recalibration algorithm ends, the counter restarts counting.

[0013] In the UCIe-based data transmission method according to the embodiment of the present application, the method further comprises the following steps: when a glitch occurs in a clock channel between the module end and the module partner end, an auxiliary channel message parsing error occurs.

[0014] According to an embodiment of the present application, an artificial intelligence chip includes a first die and a second die. The first die has a module end. The second die is coupled to the first die and has a module partner end. When the module partner end receives a retraining trigger signal, a circuit of the module partner end is reset, and the module end sends a retraining request signal to the module partner end. When the module partner end receives the retraining request signal, the module partner end sends a retraining response signal to the module end. When the module end receives the retraining response signal, the module end enters a retraining state.

[0015] According to an embodiment of the present application, a computer readable storage medium stores a computer program. The computer program is executed by a processor to implement the UCIe-based data transmission method.

[0016] According to an embodiment of the present application, an electronic device includes a storage unit and a processor. The storage unit stores a computer program. The processor is coupled to the storage unit and executes the computer program stored in the storage unit to cause the electronic device to perform the UCIe-based data transmission method.

[0017] Based on the above, the UCIe-based data transmission method, the artificial intelligence chip, the medium, and the electronic device of the present application can effectively ensure the data transmission quality of the auxiliary channel between the dies (Die-to-Die).

[0018] In order to make the above features and advantages of the present application more apparent, the following embodiments are described in detail below, and the accompanying drawings are described in detail as follows. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 A schematic diagram of a data transmission system of an embodiment of the present application;

[0020] Figure 2 A flowchart of a UCIe-based data transmission method of an embodiment of the present application;

[0021] Figure 3 A schematic diagram of a data transmission process of an embodiment of the present application;

[0022] Figure 4 A flowchart of a UCIe-based data transmission method of another embodiment of the present application;

[0023] Figure 5 A flowchart of a UCIe-based data transmission method of another embodiment of the present application;

[0024] Figure 6 A schematic diagram of an artificial intelligence chip of an embodiment of the present application.

[0025] BRIEF DESCRIPTION OF DRAWINGS

[0026] 100: data transmission system;

[0027] 110: first chiplet;

[0028] 111: module end;

[0029] 120: second chiplet;

[0030] 121: module partner end;

[0031] 600: artificial intelligence chip. DETAILED DESCRIPTION

[0032] Reference will now be made in detail to the exemplary embodiments of the present application, examples of which are illustrated in the accompanying drawings. Wherever possible, the same reference numbers will be used in the different drawings and the description to refer to the same or like parts.

[0033] Figure 1 is a schematic diagram of a data transmission system according to an embodiment of the present application. Referring to Figure 1 , the data transmission system 100 comprises a first chiplet 110 and a second chiplet 120. The first chiplet 110 and the second chiplet 120 are configured to implement inter-chiplet communication based on Universal Chiplet Interconnect Express (UCIe) technology. In this embodiment, the first chiplet 110 comprises a module end 111, and the second chiplet comprises a module partner end 121. In this embodiment, a sideband can be established between the module end 111 and the module partner end 121, which is independent of a main data channel between the module end 111 and the module partner end 121. In this embodiment, the sideband can be composed of a clock lane (i.e. clock signal line) and a data lane (i.e. data signal line). In this embodiment, the module end 111 and the module partner end 121 can exchange information between them through sideband messages. Based on the UCIe protocol, the module partner end 121 can refer to a corresponding module end on a remote die to which the module end 111 is connected.

[0034] In this embodiment, the sideband is used to transmit control signals or management information, and to ensure efficient cooperation between chiplets. For example, control commands can be transmitted between the module end 111 and the module partner end 121 through the sideband to help coordinate the operation of the main data channel. The status information of the chiplets, such as error detection or performance monitoring, can be transmitted between the module end 111 and the module partner end 121 through the sideband.

[0035] In one embodiment of the present invention, the data transmission system 100 may be implemented in an electronic device, and the electronic device includes a storage unit and a processor. The storage unit is used to store a computer program. The processor is coupled to the storage unit and is used to execute the computer program stored in the storage unit to cause the electronic device to perform the UCIe-based data transmission method as described in the embodiments of the present invention.

[0036] Processors may include, for example, a central processing unit (CPU) or other programmable general-purpose or special-purpose microprocessor, digital signal processor (DSP), programmable controller, application-specific integrated circuit (ASIC), programmable logic device (PLD), other similar processing devices, or combinations thereof.

[0037] Storage units may include, for example, random access memory (RAM), non-volatile memory, hard disk drive (HDD), or solid state drive (SSD). Random access memory may include, for example, dynamic random access memory (DRAM) or static random access memory (SRAM). Non-volatile memory may include, for example, flash memory or read-only memory (ROM).

[0038] Figure 2 This is a flowchart illustrating a UCIe-based data transmission method according to an embodiment of the present invention. (See reference...) Figure 1 as well as Figure 2 The data transmission system 100 can execute the UCIe-based data transmission method as described in steps S210 to S240. In step S210, the module end 111 of the first chip 110 communicates with the module partner end 121 of the second chip 120. In step S220, when the module partner end 121 receives a retraining trigger signal, it resets the circuitry of the module partner end 121 and sends a retraining request signal to the module partner end 121. In step S230, when the module partner end 121 receives the retraining request signal, it replies with a retraining response signal to the module end 111. In step S240, when the module end 111 receives the retraining response signal, it enters a retraining state.

[0039] Specifically, when the controller of the data transmission system 100 or at least one of the first chip 110 and the second chip 120 receives a retraining trigger signal, the circuitry of the module partner 121 can be reset first. Then, the module 111 sends a retraining request signal (e.g., including information with the information format "SB MSG{LinkMgmt.RDI.Req Retrain}") to the module partner 121, and the module partner 121 can reply with a retraining response signal (e.g., including information with the information format "SB MSG {LinkMgmt.RDI.Resp Retrain}") to the module 111, so that the module 111 enters the retraining state.

[0040] Outfit suggestions Figure 3 , Figure 3 This is a schematic diagram illustrating the data transmission process according to an embodiment of the present invention. In this embodiment, before resetting, module 111 can output data packets to module partner 121 via an auxiliary channel. Due to potential glitch on the auxiliary channel, module partner 121 may experience a sampling error, sampling a portion of the data packet as an incorrect data packet. A glitch can refer to a brief, undesirable voltage change or pulse that occurs during signal transmission.

[0041] Therefore, in this embodiment, before the first chip 110 and the second chip 120 enter the retraining state, the module partner 121 can automatically perform a reset operation to eliminate the impact of previously sampled glitches, and then it can normally sample the transmitted data packets. Furthermore, it should be noted that reset can refer to initializing the module or clearing error states through a reset signal. The reset signal of the module partner 121 can be configured through a register. The reset signal can reset the relevant receiving circuits of the module partner 121 to the initial state and clear all error states. Therefore, after the module partner 121 is reset, it can correctly sample the data packets output by the module 111. In addition, the reset operation of this embodiment can also be applied to the data transmission system 100, and the reset operation of this embodiment can be performed before the data transmission system 100 enters the first low-power state (L1) or the second low-power state (L2).

[0042] Figure 4 This is a flowchart illustrating a UCIe-based data transmission method according to another embodiment of the present invention. (See reference...) Figure 1 as well as Figure 4The data transmission system 100 can also execute the UCIe-based data transmission method as described in steps S401 to S412. In step S401, a counter starts counting when the system enters an active state. The counter can be set in the controller, first chip 110, or second chip 120 of the data transmission system 100, but the invention is not limited thereto. In step S402, when the counter count reaches a default value, the counter stops counting, and module 111 sends a recalibration tracking mode request (e.g., including information with the information format "SB MSG {RECAL.track pattern init req}") to module partner 121. In this embodiment, when the counter count reaches a default value, the counter stops counting, and in response to an auxiliary channel message parsing error, the system triggers an interrupt and causes the counter to start counting again.

[0043] In step S403, module partner 121 can determine whether the parsing of the recalibration tracking mode request signal is incorrect. If yes, step S411 is executed to trigger an interrupt, and step S412 is executed to clear the interrupt (state) and reset the counter. If no, in step S404, module partner 121 can send a recalibration tracking mode response signal (e.g., including information with the format "SB MSG {RECAL.track pattern init resp}") to module 111. Furthermore, module partner 121 can begin sending the tracking mode signal (track pattern) and the clock mode signal (clock pattern) to module 111.

[0044] In step S405, module 111 can determine whether the parsing of the recalibrated tracking mode response signal is incorrect. If yes, step S411 is executed to trigger an interrupt, and step S412 is executed to clear the interrupt (state) and reset the counter. If no, in step S406, module 111 can execute the tracking recalibration algorithm. During long-term operation, the signal quality of data transmission deteriorates due to environmental factors, chip temperature, etc., primarily due to phase differences between the clock and data signals. The tracking signal has the same circuit structure and phase as the data signal; therefore, recalibrating the tracking signal can improve the phase relationship between the clock and data signals. The tracking recalibration algorithm is a method to correct the phase relationship between the clock and data signals. It involves recalibrating the tracking signal to retrain the phase relationship between the clock and data signals, thereby adjusting the phase relationship between them. In step S407, when the tracking recalibration algorithm finishes execution, module 111 may send a recalibration tracking mode completion signal (e.g., including information with the format "SB MSG {RECAL.track pattern done req}") to module partner 121. In step S408, module partner 121 may determine whether the parsing of the recalibration tracking mode completion signal is incorrect. If yes, step S411 is executed to trigger an interrupt, and step S412 is executed to clear the interrupt (state) and reset the counter. If no, in step S409, module partner 121 may send a recalibration tracking mode completion confirmation signal (e.g., including information with the format "SB MSG {RECAL.track pattern done req}") to module 111.

[0045] In step S410, module 111 can determine whether the parsing of the recalibration tracking mode completion confirmation signal is incorrect. If yes, step S411 is executed to trigger an interrupt, and step S412 is executed to clear the interrupt (state) and reset the counter. If no, the counter is reset to start the next round of counting.

[0046] To address this, since glitches in the clock channel between module 111 and module partner 121 can cause auxiliary channel message parsing errors, the UCIe-based data transmission method in this embodiment continuously monitors the auxiliary channel signal exchange between module 111 and module partner 121. Furthermore, when an interrupt is detected, the system directly clears the interrupt and jumps to the next counter cycle. This effectively ensures that both module 111 and module partner 121 receive the correct auxiliary channel messages.

[0047] Figure 5 This is a flowchart illustrating a UCIe-based data transmission method according to another embodiment of the present invention. The data transmission system 100 can also execute the UCIe-based data transmission method as described in steps S510 to S540. In step S510, when module 111 enters an active state, module partner 121 can directly begin sending tracking mode signals and clock mode signals to module 111. In step S520, a counter starts counting. In step S530, when the counter reaches a default value, the counter stops counting, and module 111 can execute a tracking recalibration algorithm. In step S540, when the tracking recalibration algorithm finishes execution, the counter restarts counting. In other words, the UCIe-based data transmission method of this embodiment can ensure the normal triggering of the recalibration algorithm without using auxiliary channel signal interaction.

[0048] In one embodiment, the present invention also provides a computer-readable storage medium. The computer-readable storage medium can be used to store a computer program, and the computer program can be executed by a processor to implement the UCIe-based data transmission methods of the above embodiments.

[0049] Figure 6 This is a schematic diagram of an artificial intelligence chip according to an embodiment of the present invention. (See reference) Figure 6 In one embodiment, the artificial intelligence chip 600 may include a first chip 110 and a second chip 120. The first chip 110 includes a module end 111, and the second chip 120 includes a module partner end 121. The first chip 110 and the second chip 120 are used to realize inter-chip communication based on universal chip interconnect technology. The first chip 110 and the second chip 120 can form a data transmission system. For related implementation methods and technical details regarding the first chip 110 and the second chip 120, please refer to the descriptions of the above embodiments. Furthermore, the UCIe-based data transmission method between the first chip 110 and the second chip 120 can also refer to the processes of the above embodiments, thus providing sufficient guidance, suggestions, and implementation instructions. In addition, in another embodiment, the number of chips in the artificial intelligence chip 600 is not limited to... Figure 6The first core 110 and the second core 120 are shown.

[0050] In one embodiment, the artificial intelligence chip 600 may be any one of a central processing unit, a graphics processing unit (GPU), a tensor processing unit (TPU), a neural network processing unit (NPU), a deep learning processing unit (DPU), an accelerated processing unit (APU), and a general-purpose graphics processing unit (GPGPU).

[0051] In summary, the UCIe-based data transmission method, artificial intelligence chip, medium, and electronic device of this invention can effectively prevent glitches from occurring on the clock signal lines during signal transmission due to power supply noise and crosstalk, thus avoiding errors in parsing auxiliary channel information. The UCIe-based data transmission method, artificial intelligence chip, medium, and electronic device of this invention can effectively ensure that all chips receive the correct auxiliary channel information.

[0052] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data transmission method based on UCIe, characterized in that, include: Communication is achieved between the module end of the first core and the module partner end of the second core. When the module partner receives a retraining trigger signal, it resets the circuit of the module partner and sends a retraining request signal to the module partner through the module. When the module partner receives the retraining request signal, it replies with a retraining response signal to the module through the module partner. as well as When the module receives the retraining response signal, the module enters the retraining state. as well as When the module enters an active state, the counter starts counting; When the counter reaches its default value, in response to an auxiliary channel message parsing error, an interrupt is triggered, and the counter restarts its count; specifically including: When the counter count reaches the default value, the module sends a recalibration tracking mode request signal to the module partner. And when the module partner fails to resolve the recalibration tracking mode request signal, an interrupt is triggered through the module partner, and the counter is reset.

2. The data transmission method based on UCIe according to claim 1, characterized in that, Also includes: When the module partner successfully parses the recalibration tracking mode request signal, it sends a recalibration tracking mode response signal to the module through the module partner. as well as When the module fails to resolve the recalibration tracking mode response signal, an interrupt is triggered at the module end, and the counter is reset.

3. The data transmission method based on UCIe according to claim 2, characterized in that, Also includes: When the module partner successfully parses the recalibration tracking mode request signal, it begins to send the tracking mode signal and the clock mode signal through the module partner.

4. The UCIe-based data transmission method according to claim 2, characterized in that, Also includes: When the module successfully resolves the recalibrated tracking mode response signal, the tracking recalibration algorithm is executed through the module. as well as When the tracking recalibration algorithm finishes execution, the module sends a recalibration tracking mode completion signal to the module partner.

5. The UCIe-based data transmission method according to claim 4, characterized in that, Also includes: When the module partner fails to resolve a signal error in the recalibration tracking mode, an interrupt is triggered through the module partner, and the counter is reset.

6. The UCIe-based data transmission method according to claim 5, characterized in that, Also includes: When the module partner successfully completes signal parsing for the recalibration tracking mode, it sends a recalibration tracking mode completion confirmation signal to the module through the module partner. as well as When the module fails to resolve the confirmation signal error for the recalibration tracking mode, an interrupt is triggered on the module side, and the counter is restarted.

7. The UCIe-based data transmission method according to claim 1, characterized in that, Also includes: When the module enters the active state, the tracking mode signal and clock mode signal are directly sent to the module through the module partner, and the counter starts counting. When the counter count reaches the default value, the tracking recalibration algorithm is executed through the module. as well as When the tracking recalibration algorithm finishes execution, the counter is reset.

8. The data transmission method based on UCIe according to claim 1, characterized in that, When a glitch occurs in the clock channel between the module and its partner, an auxiliary channel message parsing error occurs.

9. An artificial intelligence chip, characterized in that, include: The first core chip has a module end; as well as The second core is coupled to the first core and has a module partner end. When the module partner receives a retraining trigger signal, the circuit of the module partner is reset, and the module sends a retraining request signal to the module partner. When the module partner receives the retraining request signal, the module partner replies with a retraining response signal to the module. When the module receives the retraining response signal, the module enters the retraining state. And when the module enters an active state, the counter starts counting; When the counter reaches its default value, in response to an auxiliary channel message parsing error, an interrupt is triggered, and the counter restarts its count; specifically including: When the counter count reaches the default value, the module sends a recalibration tracking mode request signal to the module partner. And when the module partner fails to resolve the recalibration tracking mode request signal, the module partner triggers an interrupt and causes the counter to start counting again.

10. A computer-readable storage medium for storing a computer program, characterized in that, The computer program is executed by a processor to implement the steps of the UCIe-based data transmission method according to any one of claims 1 to 8.

11. An electronic device, characterized in that, include: Storage unit, used to store computer programs; as well as A processor, coupled to the storage unit, is configured to execute the computer program stored in the storage unit to cause the electronic device to perform the steps of the UCIe-based data transmission method as described in any one of claims 1 to 8.

Images