US20140201416A1 - Offload processor modules for connection to system memory, and corresponding methods and systems - Google Patents

Abstract

A method can include receiving write data over a system memory bus via an in-line module connector, the write data including a metadata portion identifying a processing to be performed on at least a portion of the write data; performing the processing on at least a portion of the write data with at least one offload processor mounted on a module having the in-line module connector to generate processed data; and transmitting the processed data over the system memory bus; wherein the system memory bus is further connected to at least one processor connector configured to receive at least one host processor different from the at least one offload processor.

Description

PRIORITY CLAIMS
• This application claims the benefit of U.S. Provisional Patent Applications 61/753,892 filed on Jan. 17, 2013, 61/753,895 filed on Jan. 17, 2013, 61/753,899 filed on Jan. 17, 2013, 61/753,901 filed on Jan. 17, 2013, 61/753,903 filed on Jan. 17, 2013, 61/753,904 filed on Jan. 17, 2013, 61/753,906 filed on Jan. 17, 2013, 61/753,907 filed on Jan. 17, 2013, and 61/753,910 filed on Jan. 17, 2013, the contents all of which are incorporated by reference herein.
TECHNICAL FIELD
• The present invention relates generally to processing modules, and more particularly to offload processing modules that can be physically connected to a system memory bus to process data independent of a host processor of the system.
BACKGROUND
• Enterprises often rely on input/output (IO) intensive networked applications running on dedicated servers that support associated “state”, context or session-defined applications. Servers can run multiple applications, each associated with a specific state running on the server. Common server applications include an Apache web server, a MySQL database application, PHP hypertext preprocessing, video or audio processing with Kaltura supported software, packet filters, application cache, management and application switches, accounting, analytics, and logging.
• Unfortunately, servers can be limited by computational and memory storage costs associated with switching between applications. When multiple applications are constantly required to be available, the overhead associated with storing the session state of each application can result in poor performance due to constant switching between applications. Dividing applications between multiple processor cores can help alleviate the application switching problem, but does not eliminate it. Advanced processors often only have eight to sixteen cores, while hundreds of application or session states may be required.
• Enterprises also store and process their large amounts of data in a variety of ways. One manner in which enterprises store data is by using relational databases and corresponding relational database management systems (RDBMS). Such data, usually referred to as structured data, may be collected, normalized, formatted and stored in an RDBMS. Tools based on standardized data languages such as the Structured Query Language (SQL) may be used for accessing and processing structured data. However, it is estimated that such formatted structured data represents only a tiny fraction of an enterprise's stored data. All organizations are becoming increasingly aware that substantial information and knowledge resides in unstructured data (i.e. “Big Data”) repositories. Accordingly, simple and effective access to both structured and unstructured data are seen as necessary for maximizing the value of enterprise informational resources.
• However, the platforms that are currently being used to handle structured and unstructured data substantially differ in their architecture. In-memory processing and Storage Area Network (SAN)-like architecture are used for traditional SQL queries, while commodity or shared nothing architectures (each computing node, consisting of a processor, local memory, and disk resources, shares nothing with other nodes in the computing cluster) are usually used for processing unstructured data.
• A computing system architecture, hardware, and operational method that supports input-output (IO) intensive networked applications, as well as structured and unstructured data queries is needed. Such a system needs to readily handle high throughput data processing, be able to provide high parallelism for dividing tasks among multiple processors, and further provide efficient context switching to support multiple users or applications.
BRIEF DESCRIPTION OF THE DRAWINGS
• FIG. 1-0 shows a system according to an embodiment.
• FIG. 1-1 shows a system flow according to an embodiment.
• FIGS. 2-0 to2-3 show processor modules according to various embodiments.
• FIG. 2-4 shows a conventional dual-in-line memory module.
• FIG. 2-5 shows a system according to another embodiment.
• FIGS. 2-6 to2-11 show processor module operations according to an embodiment.
• FIG. 2-12 shows a method according to an embodiment.
• FIG. 2-13 shows a method according to another embodiment.
• FIG. 3 shows protocol stacks for processors according to embodiments.
• FIG. 4 shows a partitioning of system loads according to an embodiment.
• FIG. 5 shows a partitioning of system loads according to another embodiment.
• FIG. 6 shows a method according to another embodiment.
DETAILED DESCRIPTION
• Various embodiments of the present invention will now be described in detail with reference to a number of drawings. The embodiments show processing modules, systems, and methods in which offload processors are included on offload modules that connect to a system memory bus. Such offload processors are in addition to any host processors connected to the system memory bus, and can operate on data transferred over the system memory bus independent of any host processors. In particular embodiments, offload processors have access to a low latency memory, which can enable rapid storage and retrieval of context data for rapid context switching. In very particular embodiments, processing modules can populate physical slots for connecting in-line memory modules (e.g., DIMMs) to a system memory bus.
• In some embodiments, computing tasks can be automatically executed by offload processors according to data embedded within write data received over the system memory bus. In particular embodiments, such write data can include a “metadata” portion that identifies how the write data is to be processed.
• FIG. 2-0 is a block diagram of aprocessing module200 according to one embodiment. Aprocessing module200 can include a physical in-line module connector202, amemory interface204,arbiter logic206, offload processor(s)208,local memory210, andcontrol logic212. Aconnector202 can provide a physical connection to system memory bus. This is in contrast to a host processor which can access a system memory bus via a memory controller, or the like. In very particular embodiments, aconnector202 can be compatible with a dual in-line memory module (DIMM) slot of a computing system. Accordingly, a system including multiple DIMM slots can be populated with one ormore processing modules200, or a mix of processing modules and DIMM modules.
• Amemory interface204 can detect data transfers on a system memory bus, and in appropriate cases, enable write data to be stored in theprocessing module200 and/or read data to be read out from theprocessing module200. In some embodiments, amemory interface204 can be a slave interface, thus data transfers are controlled by a master device separate from the processing module. In very particular embodiments, amemory interface204 can be a direct memory access (DMA) slave, to accommodate DMA transfers over a system memory initiated by a DMA master. Such a DMA master can be a device different from a host processor. In such configurations,processing module200 can receive data for processing (e.g., DMA write), and transfer processed data out (e.g., DMA read) without consuming host processor resources.
• Arbiter logic206 can arbitrate between conflicting accesses of data withinprocessing module200. In some embodiments,arbiter logic206 can arbitrate between accesses byoffload processor208 and accesses external to theprocessor module200. It is understood that aprocessing module200 can include multiple locations that are operated on at the same time. It is understood that accesses arbitrated byarbiter logic206 can include accesses to physical system memory space occupied by theprocessor module200, as well as accesses to resources (e.g., processor resources). Accordingly, arbitration rules forarbiter logic206 can vary according to application. In some embodiments, such arbitration rules are fixed for a givenprocessor module200. In such cases, different applications can be accommodated by switching out different processing modules. However, in alternate embodiments, such arbitration rules can be configurable.
• Offloadprocessor208 can include one or more processors that can operate on data transferred over the system memory bus. In some embodiments, offload processors can run a general operating system which can run an application, such as Apache (as but one very particular example), enabling processor contexts to be saved and retrieved. Computing tasks executed byoffload processor208 can be handled by the hardware scheduler.Offload processors208 can operate on data buffered in theprocessor module200. In addition or alternatively, offloadprocessors208 can access data stored elsewhere in a system memory space. In some embodiments, offloadprocessors208 can include a cache memory configured to store context information. Anoffload processor208 can include multiple cores or one core.
• Aprocessor module200 can be included in a system having a host processor (not shown). In some embodiments, offloadprocessors208 can be a different type of processor as compared to the host processor. In particular embodiments, offloadprocessors208 can consume less power and/or have less computing power than a host processor. In very particular embodiments, offloadprocessors208 can be “wimpy” core processors, while a host processor can be a “brawny” core processor. In alternate embodiments, offloadprocessors208 can have equivalent computing power to any host processor. In very particular embodiments, a host processor can be an x86 type processor, while anoffload processor208 can include an ARM, ARC, Tensilica, MIPS, Strong/ARM, or RISC type processor, as but a few examples.
• Local memory210 can be connected to offloadprocessor208 to enable the storing of context information. Accordingly, anoffload processor208 can store current context information, and then switch to a new computing task, then subsequently retrieve the context information to resume the prior task. In very particular embodiments,local memory210 can be a low latency memory with respect to other memories in a system. In some embodiments, storing of context information can include copying anoffload processor208 cache.
• In some embodiments, a same space withinlocal memory210 is accessible bymultiple offload processors208 of the same type. In this way, a context stored by one offload processor can be resumed by a different offload processor.
• Control logic212 can control processing tasks executed by offload processor(s). In some embodiments,control logic212 can be considered a hardware scheduler that can be conceptualized as including adata evaluator214,scheduler216 and aswitch controller218. Adata evaluator214 can extract “metadata” from write data transferred over a system memory bus. “Metadata”, as used herein, can be any information embedded at one or more predetermined locations of a block of write data that indicates processing to be performed on all or a portion of the block of write data and/or indicate a particular task/process to which the data belongs (e.g., classification data). In some embodiments, metadata can be data that indicates a higher level organization for the block of write data. As but one very particular embodiment, metadata can be header information of one or more network packets (which may or may not be encapsulated within a higher layer packet structure).
• Ascheduler216 can order computing tasks for offload processor(s)208. In some embodiments,scheduler216 can generate a schedule that is continually updated as write data for processing is received. In very particular embodiments, ascheduler216 can generate such a schedule based on the ability to switch contexts of offload processor(s)208. In this way, on-module computing priorities can be adjusted on the fly. In very particular embodiments, ascheduler216 can assign a portion of physical address space (e.g., memory locations within local memory210) to anoffload processor208, according to computing tasks. Theoffload processor208 can then switch between such different spaces, saving context information prior to each switch, and subsequently restoring context information when returning to the memory space.
• Switch controller218 can control computing operations of offload processor(s)208. In particular embodiments, according toscheduler216,switch controller218 can order offload processor(s)208 to switch contexts. It is understood that a context switch operation can be an “atomic” operation, executed in response to a single command fromswitch controller218. In addition or alternatively, aswitch controller218 can issue an instruction set that stores current context information, recalls context information, etc.
• In some embodiments,processor module200 can include a buffer memory (not shown). A buffer memory can store received write data on board the processor module. A buffer memory can be implemented on an entirely different set of memory devices, or can be a memory embedded with logic and/or the offload processor. In the latter case,arbiter logic206 can arbitrate access to the buffer memory. In some embodiments, a buffer memory can correspond to a portion of a system physical memory space. The remaining portion of the system memory space can correspond to other like processor modules and/or memory modules connected to the same system memory bus. In some embodiments buffer memory can be different thanlocal memory210. For example, buffer memory can have a slower access time thanlocal memory210. However, in other embodiments, buffer memory and local memory can be implemented with like memory devices.
• In very particular embodiments, write data for processing can have an expected maximum flow rate. Aprocessor module200 can be configured to operate on such data at, or faster than, such a flow rate. In this way, a master device (not shown) can write data to a processor module without danger of overwriting data “in process”.
• The various computing elements of aprocessor module200 can be implemented as one or more integrated circuit devices (ICs). It is understood that the various components shown inFIG. 2-0 can be formed in the same or different ICs. For example,control logic212,memory interface214, and/orarbiter logic206 can be implemented on one or more logic ICs, while offload processor(s)208 andlocal memory210 are separate ICs. Logic ICs can be fixed logic (e.g., application specific ICs), programmable logic (e.g., field programmable gate arrays, FPGAs), or combinations thereof.
• Advantageously, the foregoing hardware and systems can provide improved computational performance as compared to traditional computing systems. Conventional systems, including those based on x86 processors, are often ill-equipped to handle such high volume applications. Even idling, x86 processors use a significant amount of power, and near continuous operation for high bandwidth packet analysis or other high volume processing tasks make the processor energy costs one of the dominant price factors.
• In addition, conventional systems can have issues with the high cost of context switching wherein a host processor is required to execute instructions which can include switching from one thread to another. Such a switch can require storing and recalling the context for the thread. If such context data is resident in a host cache memory, such a context switch can occur relatively quickly. However, if such context data is no longer in cache memory (i.e., a cache miss), the data must be recalled from system memory, which can incur a multi-cycle latency. Continuous cache misses during context switching can adversely impact system performance.
• FIG. 2-1 shows a processor module200-1 according to one very particular embodiment which is capable of reducing issues associated with high volume processing or context switching associated with many conventional server systems. A processor module200-1 can include ICs220-0/1 mounted to a printed circuit board (PCB)type substrate222.PCB type substrate222 can include in-line module connector202, which in one very particular embodiment, can be a DIMM compatible connector. IC220-0 can be a system-on-chip (SoC) type device, integrating multiple functions. In the very particular embodiment shown, an IC220-0 can include embedded processor(s), logic and memory. Such embedded processor(s) can be offload processor(s)208 as described herein, or equivalents. Such logic can be any ofcontroller logic212,memory interface204 and/orarbiter logic206, as described herein, or equivalents. Such memory can be any oflocal memory210, cache memory for offload processor(s)208, or buffer memory, as described herein, or equivalents. Logic IC220-1 can provide logic functions not included IC220-0.
• FIG. 2-2 shows a processor module200-2 according to another very particular embodiment. A processor module200-2 can include ICs220-2, -3, -4, -5 mounted to aPCB type substrate222, like that ofFIG. 2-1. However, unlikeFIG. 2-1, processor module functions are distributed among single purpose type ICs. IC220-2 can be a processor IC, which can be anoffload processor208. IC220-3 can be a memory IC which can includelocal memory210, buffer memory, or combinations thereof. IC220-4 can be a logic IC which can includecontrol logic212, and in one very particular embodiment, can be an FPGA. IC220-5 can be another logic IC which can includememory interface204 andarbiter logic206, and in one very particular embodiment, can also be an FPGA.
• It is understood thatFIGS. 2-1 and2-2 represent but two of various implementations. The various functions of a processor module can be distributed over any suitable number of ICs, including a single SoC type IC.
• FIG. 2-3 shows an opposing side of a processor module200-1 or200-2 according to a very particular embodiment. Processor module200-3 can include a number of memory ICs, one shown as220-6, mounted to aPCB type substrate222, like that ofFIG. 2-1. It is understood that various processing and logic components can be mounted on an opposing side to that shown. A memory IC220-6 can be configured to represent a portion of the physical memory space of a system. Memory ICs220-6 can perform any or all of the following functions: operate independently of other processor module components, providing system memory accessed in a conventional fashion; serve as buffer memory, storing write data that can be processed with other processor module components, or serve as local memory for storing processor context information.
• FIG. 2-4 shows a conventional DIMM module (i.e., it serves only a memory function) that can populate a memory bus along with processor modules as described herein, or equivalents.
• FIG. 2-5 shows asystem230 according to one embodiment. Asystem230 can include asystem memory bus228 accessible via multiple in-line module slots (one shown as226). According to embodiments, any or all of theslots226 can be occupied by aprocessor module200 as described herein, or an equivalent. In the event allslots226 are not occupied by aprocessor module200, available slots can be occupied by conventional in-line memory modules224. In a very particular embodiment,slots226 can be DIMM slots.
• In some embodiments, aprocessor module200 can occupy one slot. However, in other embodiments, a processor module can occupy multiple slots.
• In some embodiments, asystem memory bus228 can be further interfaced with one or more host processors and/or input/output device (not shown).
• Having described processor modules according to various embodiments, operations of a processor module according to particular embodiments will now be described.
• FIGS. 2-6 to2-11 show processor module operations according to various embodiments.FIGS. 2-6 to2-11 show a processor module like that ofFIG. 2-0, along with asystem memory bus228, and abuffer memory232. It is understood that in some embodiments, abuffer memory232 can part ofprocessor module200. In such a case, arbitration between accesses viasystem memory bus228 and offload processors can be controlled byarbiter logic206.
• Referring toFIG. 2-6, write data234-0 can be received on system memory bus228 (circle “1”). In some embodiments, such an action can include the writing of data to a particular physical address space range of a system memory. In a very particular embodiment, such an action can be a DMA write independent of any host processor. Write data234-0 can include metadata (MD) as well as data to be processed (Data). In the embodiment shown, write data234-0 can correspond to a particular processing operation (Session0).
• Control logic212 can access metadata (MD) of the write data234-0 to determine a type of processing to be performed (circle “2”) and/or classification of such data. In some embodiments, such an action can include a direct read from a physical address (i.e., MD location is at a predetermined location). In addition or alternatively, such an action can be an indirect read (i.e., MD is accessed via pointer, or the like). The action shown by circle “2” can be performed by any of: a read bycontrol logic212 or read by anoffload processor208.
• From extracted metadata,scheduler216 can create a processing schedule, or modify an existing schedule to accommodate the new computing task (circle “3”).
• Referring toFIG. 2-7, in response to ascheduler216, theswitch controller218 can direct one ormore offload processors208 be begin processing data according to MD of the write data (circles “4”, “5”). Such processing of data can include any of the following and equivalents:offload processor208 can process write data stored in a buffer memory of theprocessor module200, with accesses being arbitrated byarbiter logic206, offloadprocessor208 can operate on data previously received,offload processor208 can receive and operate on data stored at a location different than theprocessor module200.
• Referring toFIG. 2-8, additional write data234-1 can be received on system memory bus228 (circle “6”). Write data234-1 can include metadata that indicates a different processing operation (Session1) than that for write data234-0.Control logic212 can access metadata (MD) of the new write data234-1 to determine a type of processing to be performed (circle “7”). From extracted metadata, thescheduler216 can modify the current schedule to accommodate the new computing task (circle “8”). In the particular example shown, the modified schedulere-tasks offload processor208. Thus,switch controller218 can direct theoffload processor208 to store its current context (ContextA) in local memory210 (circle “9”).
• Referring toFIG. 2-9, in response to switchcontroller218, offload processor(s)208 can begin the new processing task (circle “10”). Consequently, offload processor(s)208 can maintain a new context (ContextB) corresponding to the new processing task.
• Referring toFIG. 2-10, a processing task byoffload processor208 can be completed. In the very particular embodiment shown, such processing can modify write data234-1, and such data can be read out over system memory bus228 (circle “11”). In response to the completion of processing task,scheduler216 can update a schedule. In the example shown, in response to the updated schedule,switch controller218 can direct offload processor(s)208 to restore the previously saved context (ContextA) from local memory210 (circle “12”). As understood from above, a restored context (e.g., ContextA) may have been stored by an offload processor different from the one that saved the context in the first place.
• Referring toFIG. 2-11, with a previous context restored, offload processor(s)208 can return to processing data according to the previous task (Session0) (circle “13”).
• FIG. 2-12 shows amethod240 according an embodiment. Amethod240 can include detecting the write of session data to a system memory with a module interface242. Such an action can include determining if received write data has metadata (i.e., data identifying a particular processing). It is understood that “session data” is data corresponding to a particular processing task. Further, it is understood that MD accompanying (or embedded within) session data can identify sessions having priorities with respect to one another. In particular embodiments, a module interface can be a slave interface for an in-line module.
• Amethod240 can determine if current offload processing is sufficient for a new session or change ofsession244. Such an action takes into account a processing time required for any current sessions.
• If current processing resources can accommodate new session requirements (Y from244), a hardware schedule (schedule for controlling offload processor(s)) can be revised and the new session can be assigned to an offload processor. If current processing resources cannot accommodate new session requirements (N from244), one or more offload processors can be selected for re-tasking (e.g., a context switch)250 and the hardware schedule can be modified accordingly252. The selected offload processors can save their current context data254 and then switch to thenew session256. In some embodiments, revision of a hardware schedule (252) can include storing a context switch (e.g., in the case a new session has a lower priority than current sessions). In such a case,actions254 and256 would not occur until a later point in time.
• FIG. 2-13 shows amethod260 according another embodiment. Amethod260 can include determining if a computing session for an offload processor is complete262 or has been terminated264. In such cases (Y from262/264), it can be determined if the freed module offload processor (i.e., an offload processor whose session is complete/terminated) has a storedcontext266. That is, it can be determined if freed processor was previously operating on a session.
• If a free offload processor was operating according to another session (Y from266), the offload processor can restore theprevious context268. If a free offload processor has no stored context, it can be assigned to an existing session (if possible)270. An existing hardware schedule can be updated correspondingly272.
• Processor modules according to embodiments herein can be employed to accomplish various processing tasks. According to some embodiments, processor modules can be attached to a system memory bus to operate on network packet data. Such embodiments will now be described.
• FIG. 1-0 shows asystem101 that can transport packet data to one or more computational units (one shown as100) located on a module, which in particular embodiments, can include a connector compatible with an existing memory module. In some embodiments, acomputational unit100 can include a processor module as described in embodiments herein, or an equivalent. Acomputational unit100 can be capable of intercepting or otherwise accessing packets sent over amemory bus116 and carrying out processing on such packets, including but not limited to termination or metadata processing. Asystem memory bus116 can be a system memory bus like those described herein, or equivalents (e.g.,228).
• According to some embodiments, packets corresponding to a particular flow can be transported to a storage location accessible by, or included withincomputational unit100. Such transportation can occur without consuming resources of ahost processor module106c, connected tomemory bus116. In particular embodiments, such transport can occur without interrupting thehost processor module106c. In such an arrangement, ahost processor module106cdoes not have to handle incoming flows. Incoming flows can be directed tocomputational unit100, which in particular embodiments can include ageneral purpose processor108i. Suchgeneral purpose processors108ican be capable of running code for terminating incoming flows.
• In one very particular embodiment, ageneral purpose processor108ican run code for terminating particular network flow session types, such as Apache video sessions, as but one example.
• In addition or alternatively, ageneral purpose processor108ican process metadata of a packet. In such embodiments, such metadata can include one or more fields of a header for the packet, or a header encapsulated further within the packet.
• Referring still toFIG. 1-0, according to embodiments, asystem101 can carry out any of the following functions: 1) transport packets of a flow to a destination occupied by, or accessible by, acomputational unit100 without interrupting ahost processor module106c;2) transport packets to anoffload processor108icapable of terminating session flows (i.e., the offload processor is responsible for terminating session flows); 3) transport packets to midplane switch that can process the metadata associated with a packet and make a switching decision; 4) provide a novel high speed packet terminating system.
• Conventional packet processing systems can utilize host processors for packet termination. However, due to the context switching involved in handling multiple sessions, conventional approaches require significant processing overhead for such context switching, and can incur memory access and network stack delay.
• In contrast to conventional approaches, embodiments as disclosed herein can enable high speed packet termination by reducing context switch overhead of a host processor. Embodiments can provide any of the following functions: 1) offload computation tasks to one or more processors via a system memory bus, without the knowledge of the host processor, or significant host processor involvement; 2) interconnect servers in a rack or amongst racks by employing offload processors as switches; or 3) use I/O virtualization to redirect incoming packets to different offload processors.
• Referring still toFIG. 1-0, asystem101 can include an I/O device102 which can receive packet or other I/O data from an external source. In some embodiments I/O device102 can include physical or virtual functions generated by the physical device to receive a packet or other I/O data from the network or another computer or virtual machine. In the very particular embodiment shown, an I/O device102 can include a network interface card (NIC) having input buffer102a(e.g., DMA ring buffer) and an I/O virtualization function102b.
• According to embodiments, an I/O device102 can write a descriptor including details of the necessary memory operation for the packet (i.e. read/write, source/destination). Such a descriptor can be assigned a virtual memory location (e.g., by an operating system of the system101). I/O device102 then communicates with an input output memory management unit (IOMMU)104 which can translate virtual addresses to corresponding physical addresses with anIOMMU function104b. In the particular embodiment shown, a translation look-aside buffer (TLB)104acan be used for such translation. Virtual function reads or writes data between I/O device and system memory locations can then be executed with a direct memory transfer (e.g., DMA) via amemory controller106bof thesystem101. An I/O device102 can be connected toIOMMU104 by ahost bus112. In one very particular embodiment, ahost bus112 can be a peripheral interconnect (PCI) type bus.IOMMU104 can be connected to ahost processing section106 at a central processing unit I/O (CPUIO)106a. In the embodiment shown, such aconnection114 can support a HyperTransport (HT) protocol.
• In the embodiment shown, ahost processing section106 can include the CPUIO106a,memory controller106b, processingcore106candcorresponding provisioning agent106d.
• In particular embodiments, acomputational unit100 can interface with thesystem bus116 via standard in-line module connection, which in very particular embodiments can include a DIMM type slot. In the embodiment shown, amemory bus116 can be a DDR3 type memory bus. Alternate embodiments can include any suitable system memory bus. Packet data can be sent bymemory controller106bviamemory bus116 to aDMA slave interface110a.DMA slave interface110acan be adapted to receive encapsulated read/write instructions from a DMA write over thememory bus116.
• A hardware scheduler (108b/c/d/e/h) can perform traffic management on incoming packets by categorizing them according to flow using session metadata. Packets can be queued for output in an onboard memory (110b/108a/108m) based on session priority. When the hardware scheduler determines that a packet for a particular session is ready to be processed by theoffload processor108i, the onboard memory is signaled for a context switch to that session. Utilizing this method of prioritization, context switching overhead can be reduced, as compared to conventional approaches. That is, a hardware scheduler can handle context switching decisions and thus optimize the performance of the downstream resource (e.g., offloadprocessor108i).
• As noted above, in very particular embodiments, anoffload processor108ican be a “wimpy” core” type processor. According to some embodiments, ahost processor106ccan be a “brawny core” type processor (e.g., an x86 or any other processor capable of handling “heavy touch” computational operations). While an I/O device102 can be configured to trigger host processor interrupts in response to incoming packets, according to embodiments, such interrupts can be disabled, thereby reducing processing overhead for thehost processor106c. In some very particular embodiments, anoffload processor108ican include an ARM, ARC, Tensilica, MIPS, Strong/ARM or any other processor capable of handling “light touch” operations. Preferably, an offload processor can run a general purpose operating system for executing a plurality of sessions, which can be optimized to work in conjunction with the hardware scheduler in order to reduce context switching overhead.
• Referring still toFIG. 1-0, in operation, asystem101 can receive packets from an external network over a network interface. The packets are destined for either ahost processor106cor anoffload processor108ibased on the classification logic and schematics employed by I/O device102. In particular embodiments, I/O device102 can operate as a virtualized NIC, with packets for a particular logical network or to a certain virtual MAC (VMAC) address can be directed into separate queues and sent over to the destination logical entity. Such an arrangement can transfer packets to different entities. In some embodiments, each such entity can have a virtual driver, a virtual device model that it uses to communicate with virtual network interfaces it is connected to.
• According to embodiments, multiple devices can be used to redirect traffic to specific memory addresses. So, each of the network devices operates as if it is transferring the packets to the memory location of a logical entity. However, in reality, such packets are transferred to memory addresses where they can be handled by one or more offload processors (e.g.,108i). In particular embodiments such transfers are to physical memory addresses, thus logical entities can be removed from the processing, and a host processor can be free from such packet handling.
• Accordingly, embodiments can be conceptualized as providing a memory “black box” to which specific network data can be fed. Such a memory black box can handle the data (e.g., process it) and respond back when such data is requested.
• Referring still toFIG. 1-0, according to some embodiments, I/O device102 can receive data packets from a network or from a computing device. The data packets can have certain characteristics, including transport protocol number, source and destination port numbers, source and destination IP addresses, for example. The data packets can further have metadata that is processed (108d) that helps in their classification and management.
• I/O device102 can include, but is not limited to, peripheral component interconnect (PCI) and/or PCI express (PCIe) devices connecting with a host motherboard via PCI or PCIe bus9 (e.g.,112). Examples of I/O devices include a network interface controller (NIC), a host bus adapter, a converged network adapter, an ATM network interface, etc.
• In order to provide for an abstraction scheme that allows multiple logical entities to access the same I/O device102, the I/O device may be virtualized to provide for multiple virtual devices each of which can perform some of the functions of the physical I/O device. The IO virtualization program (e.g.,102b) according to an embodiment, can redirect traffic to different memory locations (and thus to different offload processors attached to modules on a memory bus). To achieve this, an I/O device102 (e.g., a network card) may be partitioned into several function parts; including controlling function (CF) supporting input/output virtualization (IOV) architecture (e.g., single-root IOV) and multiple virtual function (VF) interfaces. Each virtual function interface may be provided with resources during runtime for dedicated usage. Examples of the CF and VF may include the physical function and virtual functions under schemes such as Single Root I/O Virtualization or Multi-Root I/O Virtualization architecture. The CF acts as the physical resources that sets up and manages virtual resources. The CF is also capable of acting as a full-fledged IO device. The VF is responsible for providing an abstraction of a virtual device for communication with multiple logical entities/multiple memory regions.
• The operating system/the hypervisor/any of the virtual machines/user code running on ahost processor106cmay be loaded with a device model, a VF driver and a driver for a CF. The device model may be used to create an emulation of a physical device for thehost processor106cto recognize each of the multiple VFs that are created. The device model may be replicated multiple times to give the impression to a VF driver (a driver that interacts with a virtual IO device) that it is interacting with a physical device of a particular type.
• For example, a certain device module may be used to emulate a network adapter such as the Intel® Ethernet Converged Network Adapter (CNA) X540-T2, so that the I/O device102 believes it is interacting with such an adapter. In such a case, each of the virtual functions may have the capability to support the functions of the above said CNA, i.e., each of the Physical Functions should be able to support such functionality. The device model and the VF driver can be run in either privileged or non-privileged mode. In some embodiments, there is no restriction with regard to who hosts/runs the code corresponding to the device model and the VF driver. The code, however, has the capability to create multiple copies of device model and VF driver so as to enable multiple copies of said I/O interface to be created.
• An application orprovisioning agent106d, as part of an application/user level code running in a kernel, may create a virtual I/O address space for each VF, during runtime and allocate part of the physical address space to it. For example, if an application handling the VF driver instructs it to read or write packets from or to memory addresses 0xaaaa to 0xffff, the device driver may write I/O descriptors into a descriptor queue with a head and tail pointer that are changed dynamically as queue entries are filled. The data structure may be of another type as well, including but not limited to aring structure102aor hash table.
• The VF can read from or write data to the address location pointed to by the driver. Further, on completing the transfer of data to the address space allocated to the driver, interrupts, which are usually triggered to the host processor to handle said network packets, can be disabled. Allocating a specific I/O space to a device can include allocating said IO space a specific physical memory space occupied.
• In another embodiment, the descriptor may comprise only a write operation, if the descriptor is associated with a specific data structure for handling incoming packets. Further, the descriptor for each of the entries in the incoming data structure may be constant so as to redirect all data write to a specific memory location. In an alternate embodiment, the descriptor for consecutive entries may point to consecutive entries in memory so as to direct incoming packets to consecutive memory locations.
• Alternatively, said operating system may create a defined physical address space for an application supporting the VF drivers and allocate a virtual memory address space to the application orprovisioning agent106d, thereby creating a mapping for each virtual function between said virtual address and a physical address space. Said mapping between virtual memory address space and physical memory space may be stored in IOMMU tables (e.g., aTLB104a). The application performing memory reads or writes may supply virtual addresses to say virtual function, and the host processor OS may allocate a specific part of the physical memory location to such an application.
• Alternatively, VF may be configured to generate requests such as read and write which may be part of a direct memory access (DMA) read or write operation, for example. The virtual addresses is be translated by theIOMMU104 to their corresponding physical addresses and the physical addresses may be provided to the memory controller for access. That is, theIOMMU104 may modify the memory requests sourced by the I/O devices to change the virtual address in the request to a physical address, and the memory request may be forwarded to the memory controller for memory access. The memory request may be forwarded over abus114 that supports a protocol such asHyperTransport114. The VF may in such cases carry out a direct memory access by supplying the virtual memory address to theIOMMU104.
• Alternatively, said application may directly code the physical address into the VF descriptors if the VF allows for it. If the VF cannot support physical addresses of the form used by thehost processor106c, an aperture with a hardware size supported by the VF device may be coded into the descriptor so that the VF is informed of the target hardware address of the device. Data that is transferred to an aperture may be mapped by a translation table to a defined physical address space in the system memory. The DMA operations may be initiated by software executed by the processors, programming the I/O devices directly or indirectly to perform the DMA operations.
• Referring still toFIG. 1-0, in particular embodiments, parts ofcomputational unit100 can be implemented with one or more FPGAs. In the system ofFIG. 1-0,computational unit100 can include FPGA110 in which can be formed a DMAslave device module110aandarbiter110f. ADMA slave module110acan be any device suitable for attachment to amemory bus116 that can respond to DMA read/write requests. In alternate embodiments, aDMA slave module110acan be another interface capable of block data transfers overmemory bus116. TheDMA slave module110acan be capable of receiving data from a DMA controller (when it performs a read from a ‘memory’ or from a peripheral) or transferring data to a DMA controller (when it performs a write instruction on theDMA slave module110a). TheDMA slave module110amay be adapted to receive DMA read and write instructions encapsulated over a memory bus, (e.g., in the form of a DDR data transmission, such as a packet or data burst), or any other format that can be sent over the corresponding memory bus.
• ADMA slave module110acan reconstruct the DMA read/write instruction from the memory R/W packet. TheDMA slave module110amay be adapted to respond to these instructions in the form of data reads/data writes to the DMA master, which could either be housed in a peripheral device, in the case of a PCIe bus, or a system DMA controller in the case of an ISA bus.
• I/O data that is received by theDMA device110acan then queued for arbitration. Arbitration can include the process of scheduling packets of different flows, such that they are provided access to available bandwidth based on a number of parameters. In general, anarbiter110fprovides resource access to one or more requestors. If multiple requestors request access, anarbiter110fcan determine which requestor becomes the accessor and then passes data from the accessor to the resource interface, and the downstream resource can begin execution on the data. After the data has been completely transferred to a resource, and the resource has competed execution, thearbiter110fcan transfer control to a different requestor and this cycle repeats for all available requestors. In the embodiment ofFIG. 1-10,arbiter110fcan notify other portions of computational unit100 (e.g.,108) of incoming data.
• Alternatively, acomputation unit100 can utilize an arbitration scheme shown in U.S. Pat. No. 7,813,283, issued to Dalal on Oct. 12, 2010, the contents of which are incorporated herein by reference. Other suitable arbitration schemes known in art could be implemented in embodiments herein. Alternatively, the arbitration scheme of the current invention might be implemented using an OpenFlow switch and an OpenFlow controller.
• In the very particular embodiment ofFIG. 1-0,computational unit100 can further include notify/prefetch circuits110cwhich can prefetch data stored in abuffer memory110bin response toDMA slave module110a, and as arbitrated byarbiter110f. Further,arbiter110fcan access other portions of thecomputational unit100 via a memory mapped I/O ingress path110eand egress path110g.
• Referring toFIG. 1-0, a hardware scheduler can include ascheduling circuit108b/nto implement traffic management of incoming packets. Packets from a certain source, relating to a certain traffic class, pertaining to a specific application or flowing to a certain socket are referred to as part of a session flow and are classified using session metadata. Such classification can be performed byclassifier108e.
• In some embodiments,session metadata108dcan serve as the criterion by which packets are prioritized and scheduled and as such, incoming packets can be reordered based on their session metadata. This reordering of packets can occur in one or more buffers and can modify the traffic shape of these flows. The scheduling discipline chosen for this prioritization, or traffic management (TM), can affect the traffic shape of flows and micro-flows through delay (buffering), bursting of traffic (buffering and bursting), smoothing of traffic (buffering and rate-limiting flows), dropping traffic (choosing data to discard so as to avoid exhausting the buffer), delay jitter (temporally shifting cells of a flow by different amounts) and by not admitting a connection (e.g., cannot simultaneously guarantee existing service (SLAs) with an additional flow's SLA).
• According to embodiments,computational unit100 can serve as part of a switch fabric, and provide traffic management with depth-limited output queues, the access to which is arbitrated by ascheduling circuit108b/n. Such output queues are managed using a scheduling discipline to provide traffic management for incoming flows. The session flows queued in each of these queues can be sent out through an output port to a downstream network element.
• It is noted that a conventional traffic management circuit doesn't take into account the handling and management of data by downstream elements except for meeting the SLA agreements it already has with said downstream elements.
• In contrast, according to embodiments ascheduler circuit108b/ncan allocate a priority to each of the output queues and carry out reordering of incoming packets to maintain persistence of session flows in these queues. Ascheduler circuit108b/ncan be used to control the scheduling of each of these persistent sessions into a general purpose operating system (OS)108j, executed on anoffload processor108i. Packets of a particular session flow, as defined above, can belong to a particular queue. Thescheduler circuit108b/nmay control the prioritization of these queues such that they are arbitrated for handling by a general purpose (GP) processing resource (e.g., offloadprocessor108i) located downstream. AnOS108jrunning on adownstream processor108ican allocate execution resources such as processor cycles and memory to a particular queue it is currently handling. TheOS108jmay further allocate a thread or a group of threads for that particular queue, so that it is handled distinctly by the generalpurpose processing element108ias a separate entity. The fact that there can be multiple sessions running on a GP processing resource, each handling data from a particular session flow resident in a queue established by the scheduler circuit, tightly integrates the scheduler and the downstream resource (e.g.,108i). This can bring about persistence of session information across the traffic management and scheduling circuit and the generalpurpose processing resource108j.
• Dedicated computing resources (e.g.,108i), memory space and session context information for each of the sessions can provide a way of handling, processing and/or terminating each of the session flows at thegeneral purpose processor108i. Thescheduler circuit108b/ncan exploit this functionality of the execution resource to queue session flows for scheduling downstream. Thescheduler circuit108b/ncan be informed of the state of the execution resource(s) (e.g.,108i), the current session that is run on the execution resource; the memory space allocated to it, the location of the session context in the processor cache.
• According to embodiments, ascheduler circuit108b/ncan further include switching circuits to change execution resources from one state to another. Thescheduler circuit108b/ncan use such a capability to arbitrate between the queues that are ready to be switched into the downstream execution resource. Further, the downstream execution resource can be optimized to reduce the penalty and overhead associated with context switch between resources. This is further exploited by thescheduler circuit108b/nto carry out seamless switching between queues, and consequently their execution as different sessions by the execution resource.
• Ascheduler circuit108b/naccording to embodiments can schedule different sessions on a downstream processing resource, wherein the two are operated in coordination to reduce the overhead during context switches. An important factor in decreasing the latency of services and engineering computational availability can be hardware context switching synchronized with network queuing. In embodiments, when a queue is selected by a traffic manager, a pipeline coordinates swapping in of the cache (e.g., L2 cache) of the corresponding resource (e.g.,108i) and transfers the reassembled I/O data into the memory space of the executing process. In certain cases, no packets are pending in the queue, but computation is still pending to service previous packets. Once this process makes a memory reference outside of the data swapped, the scheduler circuit (108b/n) can enable queued data from an I/O device102 to continue scheduling the thread.
• In some embodiments, to provide fair queuing to a process not having data, a maximum context size can be assumed as data processed. In this way, a queue can be provisioned as the greater of computational resource and network bandwidth resource. As but one very particular example, a computation resource can be an ARM A9 processor running at 800 MHz, while a network bandwidth can be 3 Gbps of bandwidth. Given the lopsided nature of this ratio, embodiments can utilize computation having many parallel sessions (such that the hardware's prefetching of session-specific data offloads a large portion of the host processor load) and having minimal general purpose processing of data.
• Accordingly, in some embodiments, ascheduler circuit108b/ncan be conceptualized as arbitrating, not between outgoing queues at line rate speeds, but arbitrating between terminated sessions at very high speeds. The stickiness of sessions across a pipeline of stages, including a general purpose OS, can be a scheduler circuit optimizing any or all such stages of such a pipeline.
• Alternatively, a scheduling scheme can be used as shown in U.S. Pat. No. 7,760,715 issued to Dalal on Jul. 20, 2010, incorporated herein by reference. This scheme can be useful when it is desirable to rate limit the flows for preventing the downstream congestion of another resource specific to the over-selected flow, or for enforcing service contracts for particular flows. Embodiments can include arbitration scheme that allows for service contracts of downstream resources, such as general purpose OS that can be enforced seamlessly.
• Referring still toFIG. 1-0, a hardware scheduler according to embodiments herein, or equivalents, can provide for the classification of incoming packet data into session flows based on session metadata. It can further provide for traffic management of these flows before they are arbitrated and queued as distinct processing entities on the offload processors.
• In some embodiments, offload processors (e.g.,108i) can be general purpose processing units capable of handling packets of different application or transport sessions. Such offload processors can be low power processors capable of executing general purpose instructions. The offload processors could be any suitable processor, including but not limited to: ARM, ARC, Tensilica, MIPS, StrongARM or any other processor that serves the functions described herein. Such offload processors have a general purpose OS running on them, wherein the general purpose OS is optimized to reduce the penalty associated with context switching between different threads or group of threads.
• In contrast, context switches on host processors can be computationally intensive processes that require the register save area, process context in the cache and TLB entries to be restored if they are invalidated or overwritten. Instruction Cache misses in host processing systems can lead to pipeline stalls and data cache misses lead to operation stall and such cache misses reduce processor efficiency and increase processor overhead.
• In contrast, anOS108jrunning on theoffload processors108iin association with ascheduler circuit108b/n, can operate together to reduce the context switch overhead incurred between different processing entities running on it. Embodiments can include a cooperative mechanism between a scheduler circuit and the OS on theoffload processor108i, wherein the OS sets up session context to be physically contiguous (physically colored allocator for session heap and stack) in the cache; then communicates the session color, size, and starting physical address to the scheduler circuit upon session initialization. During an actual context switch, a scheduler circuit can identify the session context in the cache by using these parameters and initiate a bulk transfer of these contents to an external low latency memory. In addition, the scheduler circuit can manage the prefetch of the old session if its context was saved to alocal memory108g. In particular embodiments, alocal memory108gcan be low latency memory, such as a reduced latency dynamic random access memory (RLDRAM), as but one very particular embodiment. Thus, in embodiments, session context can be identified distinctly in the cache.
• In some embodiments, context size can be limited to ensure fast switching speeds. In addition or alternatively, embodiments can include a bulk transfer mechanism to transfer out session context to alocal memory108g. The cache contents stored therein can then be retrieved and prefetched during context switch back to a previous session. Different context session data can be tagged and/or identified within thelocal memory108gfor fast retrieval. As noted above, context stored by one offload processor may be recalled by a different offload processor.
• In the very particular embodiment ofFIG. 1-0, multiple offload processing cores can be integrated into a computation FPGA108. Multiple computational FPGAs can be arbitrated by arbitrator circuits in another FPGA110. The combination of computational FPGAs (e.g.,108) and arbiter FPGAs (e.g.,110) are referred to as “XIMM” modules or “Xockets DIMM modules” (e.g., computation unit100). In particular applications, these XIMM modules can provide integrated traffic and thread management circuits that broker execution of multiple sessions on the offload processors.
• FIG. 1-0 also shows an offload processor tunnel connection608k, as well as amemory interface108mand port1081 (which can be an accelerator coherency port (ACP)).Memory interface108mcan accessbuffer memory108a.
• FIG. 1-1 shows a system flow according to an embodiment. Packet or other I/O data can be received at an I/O device120. An I/O device can be physical device, virtual device or combination thereof. Interrupts generated from the I/O data intended for ahost processor124 can be disabled, allowing such I/O data to be processed without resources of thehost processor124.
• AnIOMMU121 can map received data to physical addresses of a system address space.DMA master125 can transmit such data to such memory addresses by operation of amemory controller122.Memory controller122 can execute DRAM transfers over a memory bus with a DMA Slave. Upon receiving transferred I/O data, ahardware scheduler123 can schedule processing of such data with anoffload processor126. In some embodiments, a type of processing can be indicated by metadata within the I/O data. Further, in some embodiments such data can be stored in anonboard memory129. According to instructions fromhardware scheduler123, one ormore offload processors126 can executing computing functions in response to the I/O data. In some embodiments, such computing functions can operate on the I/O data, and such data can be subsequently read out on memory bus via a read request processed byDMA Slave127.
• Parallelization of tasks into multiple thread contexts is well known in art to provide for increased throughput. Processors architectures such as MIPS may include deep instructions pipelines to improve the number of instructions per cycle. Further, the ability to run a multi-threaded programming environment results in enhanced usage of existing processor resources. To further increase parallel execution on the hardware, processor architecture may include multiple processor cores. Multi-core architectures comprising of the same type of cores, referred to as homogeneous core architectures, provide higher instruction throughput by parallelizing threads or processes across multiple cores. However, in such homogeneous core architectures, the shared resources, such as memory, are amortized over a small number of processors.
• Memory and I/O accesses can incur a high amount of processor overhead. Further, context switches in conventional general purpose processing units can be computationally intensive. It is therefore desirable to reducing context switch overhead in a networked computing resource handling a plurality of networked applications in order to increase processor throughput. Conventional server loads can require complex transport, high memory bandwidth, extreme amounts of data bandwidth (randomly accessed, parallelized, and highly available), but often with light touch processing: HTML, video, packet-level services, security, and analytics. Further, idle processors still consume more than 50% of their peak power consumption.
• In contrast, according to embodiments herein, complex transport, data bandwidth intensive, frequent random access oriented, ‘light’ touch processing loads can be handled behind a socket abstraction created on the offload processor cores. At the same time, “heavy” touch, computing intensive loads can be handled by a socket abstraction on a host processor core (e.g., x86 processor cores). Such software sockets can allow for a natural partitioning of these loads between ARM and x86 processor cores. By usage of new application level sockets, according to embodiments, server loads can be broken up across the offload processing cores and the host processing cores.
• FIG. 3 shows protocol stacks that can be included in a system according to embodiments. A host processor (e.g.,106c) can include a brawnycore protocol stack317. Such aprotocol stack317 can include one or more applications (application)318 sitting on an operating system (OS)319. Unlike conventional host processing stacks, an OS can include aXockets Socket330 andXockets Tunneling Driver320 which can access XIMMs as described herein or equivalents. In particular embodiments,Socket330 can facilitates communication between a host processor and an offload processor module (e.g.,100) andDriver320 can enable transmission of packet data to an offload processor module (e.g.,100) (e.g., Ethernet-over-DDR tunneling). A brawnycore protocol stack317 can further include aHypervisor312 to supervise sessions and/or enablevirtual switches310 as described herein, or equivalents.
• An offload processor can include wimpycore protocol stack300. In the embodiment shown, such aprotocol stack300 can include asingle session OS302 which can run anapplication303. Additional software functions304 can control context switching, prefetching of data, and memory mapped scheduling. As packets belonging to different sessions ingress, the offload processors can rapidly switch contexts and read from different parts of memory in order to service them. Queuing and reassembly functions (320) can take decrypted incoming fragments of data, assemble them into their original form and queue them for processing on one of multiple offload processors onboard a module (e.g.,100). Input-output memory management software (e.g., IOMMU of308) can be provided in order to facilitate safe sharing and usage of physical memory when the arbitration processor switches between virtual sessions for incoming packets. Direct memory access (e.g., R/DMA of308) can allow for direct read/write to an internal memory of a module (e.g.,100). An offload processor can also utilize a virtual switching software (312) to provide switching as described herein. Header services (310) can process header data of received packets.
• Example embodiments of offload processors can include, but are not limited to, ARM A9 Cortex processors, which have a clock speed of 800 MHz and a data handling capacity of 3 Ghz. The queue depth for the traffic management circuit can be configured to be the smaller of the processing power and the network bandwidth. Given the lopsided nature of this ratio, in order to handle complete network bandwidth, sessions can be of a lightweight processing nature. Further, sessions to can be switched with minimum context switch overhead to allow the offload processor to process the high bandwidth network traffic. Further, the offload processors can provide session handling capacity greater than conventional approaches due to the ability to terminate sessions with little or no overhead. The offload processors according to embodiments are favorably disposed to handle complete offload of Apache video routing, as but one very particular embodiment.
• Alternatively in another embodiment, when equipped with many XIMMs, each containing multiple “wimpy” cores, systems may be placed near the top of rack, where they can be used as a cache for data and a processing resource for rack hot content or hot code, a means for interconnecting between racks and TOR switches, a mid tier between TOR switches and second-level switches, rack-level packet filtering, logging, and analytics, or various types of rack-level control plane agents. Simple passive optical mux/demux-ing can separate high bandwidth ports on the x86 systems into many lower bandwidth ports as needed.
• Embodiments can be favorably disposed to handle Apache, HTML, and application cache and rack level mid plane functions. In other embodiments, a network of XIMMs and a host x86 processors may be used to provide routing overlays.
• FIG. 4 is a flow schematic whereinnetwork packets450 are transferred to offload processors430 mounted withmemory device410, which can be wimpy core devices, completely without intervention of thehost CPU420, which can be a brawny core device. The offload processors430 can act as full-fledged processors hosting server applications. In one embodiment,packets450 can be received via an I/O device440 (e.g., NIC). Ahost processor420 can be free to execute other operations660 while offload processors430service network packets450.
• FIG. 5 is an alternate embodiment, in which server loads are partitioned such that packet meta data processing, routing overlays, filtering, packet logging and other hygiene functions are offloaded to the offload processors (530) mounted withmemory devices510, while and the host processor (520) implements server sessions.Offload processors530 can be wimpy core devices, whilehost processor520 can be a brawny core device. As in the case ofFIG. 4, packets can flow through an I/O device540 and offloadprocessors530 can be mounted with amemory device510.
• In another embodiment, a network of offload processor modules (i.e., XIMMs), each comprising a plurality of said offload processors may be employed to provide video overlays by associating said offload processors with local memory elements, including closely located DIMMs or solid state storage devices (SSDs). The network of XIMM modules may be used to perform memory read or writes for prefetching the data contents before they are serviced. In this case, real-time transport protocol (RTP) can be processed before packets enter traffic management, and their corresponding video data can be pre-fetched to match the streaming. Prefetches can be physically issued as (R)DMAs to other (remote) local DIMMs/SSDs. For enterprise applications, the number of the videos is limited and can be kept in local Xockets DIMMs. For public cloud/content delivery network (CDN) applications, this allows a rack to provide a shared memory space for the corpus of videos. The prefetching may be setup from any memory DIMM on any machine.
• It is anticipated that prefetching can be balanced against peer-to-peer distribution protocols (e.g. P4P) so that blocks of data can be efficiently sourced from all relevant servers. The bandwidth metric indicates how many streams can be sustained when using 10 Mbps (1 Mbps) streams. As the stream bandwidth goes down the number of streams goes up and the same session limitation becomes manifest in the RTP processing of server. Architecture according to such embodiments can allow over 10,000 high definition streams to be sustained in a 1U (one rack unit) form factor.
• In an alternate embodiment shown inFIG. 6, a complex publish and subscribership model, for handling pipelined computational tasks partitioned across a network of x86 cores and offload processor cores, can be implemented. Each task in the pipeline may be handled by the type of cores that are most favorably disposed for it. For example, offload processor modules (e.g., Xockets DIMMs) may be employed to carry out acceleration of Map-Reduce algorithms by an order of magnitude. The mid-plane defined by Xockets DIMMs can drive and receive the large PCI-e 3.0 bandwidth connecting Map steps with Reduce steps within a rack and outside of the rack. Because the shuffle step is often the bottleneck, the number of reducers is kept to a minimum so that CPUs are not overwhelmed with having to filter keys. With traffic-managed according to embodiments, the number of Reducers can rival the number of Mappers.
• FIG. 6 shows a method which can start602 and fetch input data from a file system (e.g., Hadoop Distributed File System)604. Such data can be fetched to wimpy core devices (e.g., offload processors as described herein or equivalents). Such input data can be partitioned into splits by thewimpy core devices606. Input pairs (which can include a key and value) for such splits can be obtained608. A map function can be performed on all input pairs with brawny core devices (e.g., host processor(s))610. All the mapping operation results can be reduced and sorted based on key values withbrawny core devices612. Results can be written back to thefile system614. Such an operation can be accomplished via wimpy core devices. A method may then end616.
• Alternatively, embodiments can employ the Xocket DIMMs to implement rack level disks using memory mapped file paradigm. Such embodiments can effectively unify all of the contents on the Xockets DIMMs on the rack to every x86 processor socket.
• It should be appreciated that in the foregoing description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure aiding in the understanding of one or more of the various inventive aspects. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.
• It is also understood that the embodiments of the invention may be practiced in the absence of an element and/or step not specifically disclosed. That is, an inventive feature of the invention may be elimination of an element.
• Accordingly, while the various aspects of the particular embodiments set forth herein have been described in detail, the present invention could be subject to various changes, substitutions, and alterations without departing from the spirit and scope of the invention.

Claims (19)

What is claimed is:

1. A method, comprising:

receiving write data over a system memory bus via an in-line module connector, the write data including a metadata portion identifying a processing to be performed on at least a portion of the write data;

performing the processing on at least a portion of the write data with at least one offload processor mounted on a module having the in-line module connector to generate processed data; and

transmitting the processed data over the system memory bus; wherein

the system memory bus is further connected to at least one processor connector configured to receive at least one host processor different from the at least one offload processor.

2. The method ofclaim 1, wherein:

receiving the write data includes processing a direct memory access (DMA) write request with slave interface of the module.

3. The method ofclaim 2, wherein:

the DMA write request is not issued by a host processor.

4. The method ofclaim 1, wherein:

receiving the write data includes storing the write data in a buffer memory of the module.

5. The method ofclaim 1, further including:

in response to predetermined conditions, storing a processing context of the at least one offload processor within the module, and redirecting the offload processor to process other data.

6. The method ofclaim 5, further including:

after processing or terminating processing of the other data, restoring the stored processing context to the offload processor.

7. The method ofclaim 1, wherein:

transmitting the processed data over the system memory bus includes transmitting the processed data to an input/output device that is different than a host processor.

8. A method, comprising:

writing data to predetermined physical addresses of a system memory, the data including metadata that identifies a processing type;

configuring an processor module to include the predetermined physical addresses, the processor module being physically connected to a system memory bus by a memory module connection; and

processing the write data according to the processing type with an offload processor mounted on the processor module.

9. The method ofclaim 8, wherein:

writing data includes servicing a direct memory access (DMA) request from a DMA controller.

10. The method ofclaim 9, wherein:

at least one host processor is coupled to the system memory bus by a memory controller; and

the DMA request is initiated by device other than the host processor.

12. The method ofclaim 9, further including:

scheduling multiple processing types for multiples write data according to the metadata of the write data.

13. The method ofclaim 12, wherein:

the scheduling of multiple processing types includes

storing a current context of the offload processor, and

processing different write data with the offload processor.

14. The method ofclaim 8, further including:

the processing of the write data generates processed data; and

reading the processed data from at least a portion of the predetermined physical addresses.

15. A method, comprising:

mapping first type data received from an input device to a physical address space of a system memory;

transferring the first data over a system memory bus coupled to at least one host processor via a memory controller; and

processing the first data transferred over the system memory bus with at least one offload processor, different than a host processor, mounted on a processor module physically connected to the system memory bus, and independent of any host processor.

16. The method ofclaim 15, wherein:

mapping first type data received from an input device to a physical address space of a system memory includes

mapping data packets received over a same connection to different virtual data paths, and

assigning each virtual data path to a different region of the physical address space.

17. The method ofclaim 15, wherein:

transferring the first data over the system memory bus includes writing the first memory to memory locations on the processor module.

18. The method ofclaim 15, further including:

transferring second data over the system memory bus via the memory controller;

storing the second data in memory locations on a memory module physically connected to the system memory bus; and

processing the second data with the host processor.

19. The method ofclaim 15, wherein:

the first data includes metadata and payload data; and

processing the first data includes

selecting a processing type according to the metadata, and

processing at least a portion of the payload data according to the selected processing type.

20. The method ofclaim 15, further including:

processing third data transferred over the system bus with the offload processor with a first processing type; and

processing the third data with the host processor with a second processing type.

Images