Application-transparent near-memory processing architecture with memory channel network
The physical memory capacity of servers is expected to increase drastically with the deployment of the forthcoming non-volatile memory technologies. This is a welcomed improvement for the emerging data-intensive applications. For such servers to be cost-effective, nonetheless, we must cost-effective...
Gespeichert in:
| Veröffentlicht in: | 2018 51st Annual IEEE ACM International Symposium on Microarchitecture (MICRO) S. 802 - 814 |
|---|---|
| Hauptverfasser: | , , , , , , , , , , , , |
| Format: | Tagungsbericht |
| Sprache: | Englisch |
| Veröffentlicht: |
Piscataway, NJ, USA
IEEE Press
20.10.2018
IEEE |
| Schriftenreihe: | ACM Conferences |
| Schlagworte: | |
| ISBN: | 9781538662403, 153866240X |
| Online-Zugang: | Volltext |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Abstract | The physical memory capacity of servers is expected to increase drastically with the deployment of the forthcoming non-volatile memory technologies. This is a welcomed improvement for the emerging data-intensive applications. For such servers to be cost-effective, nonetheless, we must cost-effectively increase compute throughput and memory bandwidth commensurate with the increase in memory capacity without compromising the application readiness. Tackling this challenge, we present Memory Channel Network (MCN) architecture in this paper. Specifically, first, we propose an MCN DIMM, an extension of a buffered DIMM where a small but capable processor called MCN processor is integrated with a buffer device on the DIMM for near-memory processing. Second, we implement device drivers to give the host and MCN processors in a server an illusion that they are independent heterogeneous nodes connected through an Ethernet link. These allow the host and MCN processors in a server to run a given data-intensive application together based on popular distributed computing frameworks such as MPI and Spark without any change in the host processor hardware and its application software, while offering the benefits of high-bandwidth and low-latency communication between the host and MCN processors over the memory channels. As such, MCN can serve as an application-transparent framework which can seamlessly unify the near-memory processing within a server and the distributed computing across such servers for data-intensive applications. Our simulation running the full software stack shows that a server with 8 MCN DIMMs offers 4.56 x higher throughput and consume 47.5% less energy than a cluster with 9 conventional nodes connected through Ethernet links, as it facilitates up to 8.17 x higher aggregate DRAM bandwidth utilization. Lastly, we demonstrate the feasibility of MCN with an IBM POWER8 system and an experimental buffered DIMM. |
|---|---|
| AbstractList | The physical memory capacity of servers is expected to increase drastically with deployment of the forthcoming non-volatile memory technologies. This is a welcomed improvement for emerging data-intensive applications. For such servers to be cost-effective, nonetheless, we must cost-effectively increase compute throughput and memory bandwidth commensurate with the increase in memory capacity without compromising application readiness. Tackling this challenge, we present Memory Channel Network (MCN) architecture in this paper. Specifically, first, we propose an MCN DIMM, an extension of a buffered DIMM where a small but capable processor called MCN processor is integrated with a buffer device on the DIMM for near-memory processing. Second, we implement device drivers to give the host and MCN processors in a server an illusion that they are independent heterogeneous nodes connected through an Ethernet link. These allow the host and MCN processors in a server to run a given data-intensive application together based on popular distributed computing frameworks such as MPI and Spark without any change in the host processor hardware and its application software, while offering the benefits of high-bandwidth and low-latency communications between the host and the MCN processors over memory channels. As such, MCN can serve as an application-transparent framework which can seamlessly unify near-memory processing within a server and distributed computing across such servers for data-intensive applications. Our simulation running the full software stack shows that a server with 8 MCN DIMMs offers 4.56X higher throughput and consume 47.5% less energy than a cluster with 9 conventional nodes connected through Ethernet links, as it facilitates up to 8.17X higher aggregate DRAM bandwidth utilization. Lastly, we demonstrate the feasibility of MCN with an IBM POWER8 system and an experimental buffered DIMM. The physical memory capacity of servers is expected to increase drastically with the deployment of the forthcoming non-volatile memory technologies. This is a welcomed improvement for the emerging data-intensive applications. For such servers to be cost-effective, nonetheless, we must cost-effectively increase compute throughput and memory bandwidth commensurate with the increase in memory capacity without compromising the application readiness. Tackling this challenge, we present Memory Channel Network (MCN) architecture in this paper. Specifically, first, we propose an MCN DIMM, an extension of a buffered DIMM where a small but capable processor called MCN processor is integrated with a buffer device on the DIMM for near-memory processing. Second, we implement device drivers to give the host and MCN processors in a server an illusion that they are independent heterogeneous nodes connected through an Ethernet link. These allow the host and MCN processors in a server to run a given data-intensive application together based on popular distributed computing frameworks such as MPI and Spark without any change in the host processor hardware and its application software, while offering the benefits of high-bandwidth and low-latency communication between the host and MCN processors over the memory channels. As such, MCN can serve as an application-transparent framework which can seamlessly unify the near-memory processing within a server and the distributed computing across such servers for data-intensive applications. Our simulation running the full software stack shows that a server with 8 MCN DIMMs offers 4.56 x higher throughput and consume 47.5% less energy than a cluster with 9 conventional nodes connected through Ethernet links, as it facilitates up to 8.17 x higher aggregate DRAM bandwidth utilization. Lastly, we demonstrate the feasibility of MCN with an IBM POWER8 system and an experimental buffered DIMM. |
| Author | McPadden, Adam O'Halloran, Oliver Roewer, Thomas Asgharimoghaddam, Hadi Dhar, Ashutosh Alian, Mohammad Kim, Nam Sung Hwu, Wen-mei Wang, Dong Kai Chen, Deming Min, Seung Won Xiong, Jinjun Kim, Daehoon |
| Author_xml | – sequence: 1 givenname: Mohammad surname: Alian fullname: Alian, Mohammad organization: University of Illinois, Urbana-Champaign – sequence: 2 givenname: Seung Won surname: Min fullname: Min, Seung Won organization: University of Illinois, Urbana-Champaign – sequence: 3 givenname: Hadi surname: Asgharimoghaddam fullname: Asgharimoghaddam, Hadi organization: University of Illinois, Urbana-Champaign – sequence: 4 givenname: Ashutosh surname: Dhar fullname: Dhar, Ashutosh organization: University of Illinois, Urbana-Champaign – sequence: 5 givenname: Dong Kai surname: Wang fullname: Wang, Dong Kai organization: University of Illinois, Urbana-Champaign – sequence: 6 givenname: Thomas surname: Roewer fullname: Roewer, Thomas organization: IBM Research – sequence: 7 givenname: Adam surname: McPadden fullname: McPadden, Adam organization: IBM Systems – sequence: 8 givenname: Oliver surname: O'Halloran fullname: O'Halloran, Oliver organization: IBM Systems – sequence: 9 givenname: Deming surname: Chen fullname: Chen, Deming organization: University of Illinois, Urbana-Champaign – sequence: 10 givenname: Jinjun surname: Xiong fullname: Xiong, Jinjun organization: IBM Research – sequence: 11 givenname: Daehoon surname: Kim fullname: Kim, Daehoon organization: DGIST – sequence: 12 givenname: Wen-mei surname: Hwu fullname: Hwu, Wen-mei organization: University of Illinois, Urbana-Champaign – sequence: 13 givenname: Nam Sung surname: Kim fullname: Kim, Nam Sung organization: University of Illinois, Urbana-Champaign |
| BookMark | eNqNkDtPwzAUhY0ACSjdkViyMqS9jhPbGauKR6WiSgjEaNnODTVtncgxqvrvSR8TE9PV0TnfHb4bcuEbj4TcURhRCuX4dTZ9W4wyoHIEAALOyLAUkhZMcp7lwM7_5Csy7LrvfppxySTwa_I5adu1szq6xqcxaN-1OqCPiUcd0g1umrBL2tBY7DrnvxId7NJFtPEnYLJ1cZmcNnapvcd1z8VtE1a35LLW6w6HpzsgH0-P79OXdL54nk0n81RTnsc0Bw2G8UxAQWvMpGFMVzVURgCnRlgjSioNSM1AFBLyrMx5LViJpqoqW3I2IPfHvw4RVRvcRoedkoXICyn6dnxstd0o0zSrTlFQe3XqoE7t1amDOmWCw7onHv5LsF9KznCj |
| CODEN | IEEPAD |
| ContentType | Conference Proceeding |
| DBID | 6IE 6IL CBEJK RIE RIL |
| DOI | 10.1109/MICRO.2018.00070 |
| DatabaseName | IEEE Electronic Library (IEL) Conference Proceedings IEEE Xplore POP ALL IEEE Xplore All Conference Proceedings IEEE/IET Electronic Library (IEL) (UW System Shared) IEEE Proceedings Order Plans (POP All) 1998-Present |
| DatabaseTitleList | |
| Database_xml | – sequence: 1 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Computer Science |
| EISBN | 9781538662403 153866240X |
| EndPage | 814 |
| ExternalDocumentID | 8574587 |
| Genre | orig-research |
| GroupedDBID | 6IE 6IF 6IL 6IN AAJGR ABLEC ALMA_UNASSIGNED_HOLDINGS BEFXN BFFAM BGNUA BKEBE BPEOZ CBEJK IEGSK OCL RIB RIC RIE RIL AAWTH |
| ID | FETCH-LOGICAL-a164t-40a0b3627051fe28b33adf0db7061b7cb7918b08a30758042946f739ebdddc963 |
| IEDL.DBID | RIE |
| ISBN | 9781538662403 153866240X |
| IngestDate | Wed Aug 27 02:46:37 EDT 2025 Sat Jun 15 16:36:41 EDT 2024 Wed Jan 31 06:41:02 EST 2024 |
| IsPeerReviewed | false |
| IsScholarly | true |
| Language | English |
| LinkModel | DirectLink |
| MeetingName | MICRO-51: The 51st Annual IEEE/ACM International Symposium on Microarchitecture |
| MergedId | FETCHMERGED-LOGICAL-a164t-40a0b3627051fe28b33adf0db7061b7cb7918b08a30758042946f739ebdddc963 |
| PageCount | 13 |
| ParticipantIDs | ieee_primary_8574587 acm_books_10_1109_MICRO_2018_00070_brief acm_books_10_1109_MICRO_2018_00070 |
| PublicationCentury | 2000 |
| PublicationDate | 20181020 2018-Oct |
| PublicationDateYYYYMMDD | 2018-10-20 2018-10-01 |
| PublicationDate_xml | – month: 10 year: 2018 text: 20181020 day: 20 |
| PublicationDecade | 2010 |
| PublicationPlace | Piscataway, NJ, USA |
| PublicationPlace_xml | – name: Piscataway, NJ, USA |
| PublicationSeriesTitle | ACM Conferences |
| PublicationTitle | 2018 51st Annual IEEE ACM International Symposium on Microarchitecture (MICRO) |
| PublicationTitleAbbrev | MICRO |
| PublicationYear | 2018 |
| Publisher | IEEE Press IEEE |
| Publisher_xml | – name: IEEE Press – name: IEEE |
| SSID | ssj0002683806 |
| Score | 2.2941432 |
| Snippet | The physical memory capacity of servers is expected to increase drastically with the deployment of the forthcoming non-volatile memory technologies. This is a... The physical memory capacity of servers is expected to increase drastically with deployment of the forthcoming non-volatile memory technologies. This is a... |
| SourceID | ieee acm |
| SourceType | Publisher |
| StartPage | 802 |
| SubjectTerms | Application Transparent Bandwidth Buffer Device Computer architecture Distributed Systems DRAM Ethernet General and reference General and reference -- Cross-computing tools and techniques General and reference -- Cross-computing tools and techniques -- Performance Hardware Memory Channel Mobile Processors Near Memory Processing Processing In Memory Random access memory Servers TCP IP |
| Title | Application-transparent near-memory processing architecture with memory channel network |
| URI | https://ieeexplore.ieee.org/document/8574587 |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3fS8MwED624YNPUzdx_iKIDz5Yl60_kjwOcfjiHKK4t9CmKQy0k60T_O-9S2unIIhvpSQp_cjx5e6-ywGcR3jq8C03XjDgvod8bTwVpBTIVxk6XEhCceyaTYjJRM5matqAy7oWxlrrxGf2ih5dLj9dmDWFyvoyFEEoRROaQkRlrVYdTxlG0pe8zkRy1cffeLgn8RapJTl1I27G5vVHFxVHIuP2_z6_A91NNR6b1jyzCw2b70H7qx0Dq6yzA8-jTTLaKy8tp0qvgk1wN3t3JKn9YFVhAC7ERt9yCIzisawaQxUHuX3BeU4i3oWn8c3j9a1X9U3wYnR-CnQJY54gMQk0uMwOZeL7cZrxNBFI3okwiVADmXAZo32HkhgpiDLhK5ukaWrQIvehlS9yewAsSq3xfRXawKqA2pIToQfcRFlo8MAb9uAMMdXkEKy08ye40g54TcBrB3wPLv4epJPl3GY96BDu-q28akNXkB_-_voItmmBUlt3DK1iubYnsGXei_lqeer2xyc4SLer |
| linkProvider | IEEE |
| linkToHtml | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3fS8MwED62KejT1E2cP4v44IN12Zq2yeMQx8RtDpm4t9CmKQx0k60T_O-9y-qmIIhvpSSh_cjx5e6-ywFcBHjq8AzTLm8wz0W-1q7kCQXyZYoOF5JQFNlmE2G_L0YjOSjA1aoWxhhjxWfmmh5tLj-Z6gWFyurCD7kvwiJs-Jw32bJaaxVRaQbCE2yVi2Syjj_y-EDyLdJLMupHXIz0648-KpZG2uX_fcAOVNf1eM5gxTS7UDCTPSh_NWRwcvuswHNrnY52l9eWU61X5vRxP7s9EtV-OHlpAC7ktL5lERyKyDr5GKo5mJgXnGdF4lV4at8Obzpu3jnBjdD9ydApjFiM1BSiyaWmKWLPi5KUJXGI9B2HOg5lQ8RMRGjhviBO4kEaetLESZJotMl9KE2mE3MATpAY7XnSN9xITo3JidI500Hqazzy-jU4R0wVuQRzZT0KJpUFXhHwygJfg8u_B6l4NjZpDSqEu3pbXrahcsgPf399BludYa-runf9-yPYpsWWSrtjKGWzhTmBTf2ejeezU7tXPgHy-7ry |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=Proceedings+of+the+51st+Annual+IEEE%2FACM+International+Symposium+on+Microarchitecture&rft.atitle=Application-transparent+near-memory+processing+architecture+with+memory+channel+network&rft.au=Alian%2C+Mohammad&rft.au=Min%2C+Seung+Won&rft.au=Asgharimoghaddam%2C+Hadi&rft.au=Dhar%2C+Ashutosh&rft.series=ACM+Conferences&rft.date=2018-10-20&rft.pub=IEEE+Press&rft.isbn=9781538662403&rft.spage=802&rft.epage=814&rft_id=info:doi/10.1109%2FMICRO.2018.00070 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=9781538662403/lc.gif&client=summon&freeimage=true |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=9781538662403/mc.gif&client=summon&freeimage=true |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=9781538662403/sc.gif&client=summon&freeimage=true |