Neo4j cozies up to Microsoft as 'property sharding' promises to overcome scalability struggle
(2025/09/11)
- Reference: 1757593811
- News link: https://www.theregister.co.uk/2025/09/11/neo4j_cozies_up_to_microsoft/
- Source link:
Neo4j has introduced "property sharding" which, according to one analyst, will help overcome its earlier struggles with scalability, while also allowing transactional workloads on the same system.
Free Software Foundation rides to defend AGPLv3 against Neo4j license add-ons [1]READ MORE
The graph database stalwart — used by customers including BT Group, risk insight company Dun & Bradstreet, and chemicals giant BASF — launched Infinigraph earlier this month promising a new distributed graph architecture in its self-managed offering. The system will also soon be available in Microsoft's AuraDB DBaaS and as part of Fabrics, the Redmond software company's data lake and analytics platform.
Graph databases group data according to nodes and edges as opposed to the columns and rows of relational databases, in the hope that the structure is more apt for the analysis of networks of relationships, for example, between groups of companies or individuals using social media.
As well as newfound scalability, Infinigraph would allow the system to accommodate transactional workloads as well as the analytics use cases graph databases are known for.
Sudhir Hasbe, Technology President at Neo4j, said: "Infinigraph sets a new standard for enterprise graph databases: one system that runs real-time operations and deep analytics together, at full fidelity and massive scale. We're giving builders the power to create intelligent systems that transform data into knowledge, scale without limits, and solve their biggest data challenges - without added complexity or cost."
[2]
Neo4j said Infinigraph works using a feature it calls "property sharding," which stores the graph's structure (the nodes and relationships) in a single graph shard that preserves the structure of the graph as a cohesive unit. These property shards can then be distributed across different machines in a cluster to achieve horizontal scalability without sacrificing graph traversals — or searches — which happen completely within the graph shard.
[3]
[4]
Infinigraph also promises to run both transactional and analytics workloads on the same system, thereby avoiding extract-transform-load (ETL) pipelines, sync delays, or redundant infrastructure, Neo4j said.
But the system is unlikely to win over transactional workloads from RDBMS systems, said Robin Schumacher, senior research director and analyst at Gartner.
[5]
"While the new release won't necessarily take traditional operational use cases away from existing DBMS vendors, it will enable the company to tackle larger workloads that fit well into the standard graph use cases, which may also have mixed workloads of quickly traversed transactions and longer-running analytical queries," he told The Register .
He added that Neo4j had previously struggled with scalability, a view which has been borne out among users.
For example, in 2021, car manufacturer Jaguar Land Rover [6]rejected leading graph database Neo4j because of scalability concerns , opting for rival TigerGraph, which the head of data and analytics said performed better as a distributed system.
[7]
Schumacher said Infinigraph might go some way to allaying these concerns. "Neo4j has always been one of the first solutions thought of by those looking for a DBMS to address graph use cases, however its historical reputation has been one of struggling with scalability - a weakness exploited by its nearest competitors. The company's Infinigraph release is aimed at removing that shortcoming while also delivering more administrative simplicity and better support for graph-driven systems that contain both transactional and analytical traffic."
However, other concerns – over cost – persist. Earlier this year, The Register revealed that [8]Neo4j lost out to rival Memgraph due to cost at US space agency NASA , which had previously used the incumbent database to bring together data from the space agency's various enterprise applications to understand the relationship between knowledge, skills, abilities, tasks and technologies (KSATTs), and occupations, roles, and training.
[9]NASA jettisons Neo4j database for Memgraph citing costs
[10]FYI: An appeals court may kill a GNU GPL software license
[11]Manifest file destiny: Declare your funding needs via JSON
[12]Graph database shows Biden outspends Trump in social media ad war
What's more, [13]debate has raged about whether developers and data scientists need a separate database system for graph at all. Andy Pavlo, an associate professor of databaseology at Carnegie Mellon University, has argued these type of tasks can be just as well performed in an RDBMS.
For example, PostgreSQL has [14]a graph extension called Apache AGE which provides graph database functionality. Microsoft offers PostgreSQL as part of its Cosmos DB DBaaS and [15]supports the graph extension . ®
Get our [16]Tech Resources
[1] https://www.theregister.com/2025/03/04/free_software_foundation_agplv3/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://www.theregister.com/2021/05/10/jaguar_land_rover_tigergraph/
[7] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[8] https://www.theregister.com/2025/05/07/nasa_people_memgraph/
[9] https://www.theregister.com/2025/05/07/nasa_people_memgraph/
[10] https://www.theregister.com/2025/02/27/adverse_appeals_court_ruling_could/
[11] https://www.theregister.com/2024/10/17/zerodha_open_source_fund/
[12] https://www.theregister.com/2024/05/17/graph_database_presidential_ads/
[13] https://www.theregister.com/2023/03/06/great_graph_debate_monday/
[14] https://age.apache.org/
[15] https://learn.microsoft.com/en-us/azure/postgresql/flexible-server/generative-ai-age-overview
[16] https://whitepapers.theregister.com/
Free Software Foundation rides to defend AGPLv3 against Neo4j license add-ons [1]READ MORE
The graph database stalwart — used by customers including BT Group, risk insight company Dun & Bradstreet, and chemicals giant BASF — launched Infinigraph earlier this month promising a new distributed graph architecture in its self-managed offering. The system will also soon be available in Microsoft's AuraDB DBaaS and as part of Fabrics, the Redmond software company's data lake and analytics platform.
Graph databases group data according to nodes and edges as opposed to the columns and rows of relational databases, in the hope that the structure is more apt for the analysis of networks of relationships, for example, between groups of companies or individuals using social media.
As well as newfound scalability, Infinigraph would allow the system to accommodate transactional workloads as well as the analytics use cases graph databases are known for.
Sudhir Hasbe, Technology President at Neo4j, said: "Infinigraph sets a new standard for enterprise graph databases: one system that runs real-time operations and deep analytics together, at full fidelity and massive scale. We're giving builders the power to create intelligent systems that transform data into knowledge, scale without limits, and solve their biggest data challenges - without added complexity or cost."
[2]
Neo4j said Infinigraph works using a feature it calls "property sharding," which stores the graph's structure (the nodes and relationships) in a single graph shard that preserves the structure of the graph as a cohesive unit. These property shards can then be distributed across different machines in a cluster to achieve horizontal scalability without sacrificing graph traversals — or searches — which happen completely within the graph shard.
[3]
[4]
Infinigraph also promises to run both transactional and analytics workloads on the same system, thereby avoiding extract-transform-load (ETL) pipelines, sync delays, or redundant infrastructure, Neo4j said.
But the system is unlikely to win over transactional workloads from RDBMS systems, said Robin Schumacher, senior research director and analyst at Gartner.
[5]
"While the new release won't necessarily take traditional operational use cases away from existing DBMS vendors, it will enable the company to tackle larger workloads that fit well into the standard graph use cases, which may also have mixed workloads of quickly traversed transactions and longer-running analytical queries," he told The Register .
He added that Neo4j had previously struggled with scalability, a view which has been borne out among users.
For example, in 2021, car manufacturer Jaguar Land Rover [6]rejected leading graph database Neo4j because of scalability concerns , opting for rival TigerGraph, which the head of data and analytics said performed better as a distributed system.
[7]
Schumacher said Infinigraph might go some way to allaying these concerns. "Neo4j has always been one of the first solutions thought of by those looking for a DBMS to address graph use cases, however its historical reputation has been one of struggling with scalability - a weakness exploited by its nearest competitors. The company's Infinigraph release is aimed at removing that shortcoming while also delivering more administrative simplicity and better support for graph-driven systems that contain both transactional and analytical traffic."
However, other concerns – over cost – persist. Earlier this year, The Register revealed that [8]Neo4j lost out to rival Memgraph due to cost at US space agency NASA , which had previously used the incumbent database to bring together data from the space agency's various enterprise applications to understand the relationship between knowledge, skills, abilities, tasks and technologies (KSATTs), and occupations, roles, and training.
[9]NASA jettisons Neo4j database for Memgraph citing costs
[10]FYI: An appeals court may kill a GNU GPL software license
[11]Manifest file destiny: Declare your funding needs via JSON
[12]Graph database shows Biden outspends Trump in social media ad war
What's more, [13]debate has raged about whether developers and data scientists need a separate database system for graph at all. Andy Pavlo, an associate professor of databaseology at Carnegie Mellon University, has argued these type of tasks can be just as well performed in an RDBMS.
For example, PostgreSQL has [14]a graph extension called Apache AGE which provides graph database functionality. Microsoft offers PostgreSQL as part of its Cosmos DB DBaaS and [15]supports the graph extension . ®
Get our [16]Tech Resources
[1] https://www.theregister.com/2025/03/04/free_software_foundation_agplv3/
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[5] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[6] https://www.theregister.com/2021/05/10/jaguar_land_rover_tigergraph/
[7] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/databases&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aMLyFfacdAFAzUHqYwU2QQAAAQc&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[8] https://www.theregister.com/2025/05/07/nasa_people_memgraph/
[9] https://www.theregister.com/2025/05/07/nasa_people_memgraph/
[10] https://www.theregister.com/2025/02/27/adverse_appeals_court_ruling_could/
[11] https://www.theregister.com/2024/10/17/zerodha_open_source_fund/
[12] https://www.theregister.com/2024/05/17/graph_database_presidential_ads/
[13] https://www.theregister.com/2023/03/06/great_graph_debate_monday/
[14] https://age.apache.org/
[15] https://learn.microsoft.com/en-us/azure/postgresql/flexible-server/generative-ai-age-overview
[16] https://whitepapers.theregister.com/
"property sharding"
I first read that as 'sharting'...
Which is no surprise as MS is involved...