In-Memory Computing Market (2026-2036)
The global In-Memory Computing Market was valued at USD 24.00 billion in 2025. This market is expected to reach USD 94.53 billion by 2036 from an estimated USD 27.36 billion in 2026, registering a CAGR of 13.2% during the forecast period (2026-2036).
- Published
- Oct 2026
- Pages
- 326
- Format
- PDF + Excel
- Report ID
- MR-2277
- Base year
- 2025
- 2025 · BASELINE
- $24.00B
- 2036
- $94.53B
- CAGR 2026–2036
- 13.2%
2025 baseline · 2026–2036 forecast at 13.2% CAGR · hover a bar for the value
Key highlights
The global In-Memory Computing Market is projected to reach USD 94.53 billion by 2036, driven by real-time data and AI workloads, ERP modernization, and processing-in-memory for AI.
North America is expected to account for the largest market share in 2026, while Asia-Pacific is projected to register the fastest growth during the forecast period.
Enterprise platforms built on in-memory databases are growing rapidly. SAP, whose ERP suite runs on its SAP HANA in-memory database, reported 2025 cloud revenue of EUR 21.02 billion, up 23%, Cloud ERP Suite revenue of EUR 18.12 billion, up 28%, and a total cloud backlog of EUR 77.29 billion, up 30% at constant currencies.
By component, Software is expected to account for the largest market share, whereas Hardware is projected to witness the fastest growth through 2036.
Memory costs are rising sharply. Samsung raised prices for some memory chips by up to 60% in December 2025, DRAM average prices were expected to rise 50% to 55% year over year in the fourth quarter of 2025, and data centers were expected to consume up to 70% of memory supply in 2026, increasing the cost of memory-intensive computing.
The open-source landscape has fragmented. After Redis changed its license in March 2024, the Linux Foundation launched the Valkey fork with backing from companies including Amazon Web Services, Google Cloud, and Oracle, and Redis added an open-source AGPL license option with Redis 8 in 2025.
Report summary
| Particulars | Details |
|---|---|
| Forecast Period | 2026-2036 |
| Base Year | 2025 |
| Estimated Year | 2026 |
| CAGR (Value) | 13.2% |
| Format | PDF, Excel & Cloud Portal · 326 pages |
| Market Size (Value) in 2026 | USD 27.36 Billion |
| Market Size (Value) in 2036 | USD 94.53 Billion |
| Segments Covered | By Component: Software (In-Memory Databases, In-Memory Data Grids & Caching, In-Memory Analytics, Stream Processing), Hardware (Processing-in-Memory, Compute-in-Memory Accelerators, In-Memory Appliances), Services. By Deployment: Cloud, On-Premises, Hybrid. By Hardware Technology: DRAM & HBM-Based Processing-in-Memory, Digital SRAM-Based Compute-in-Memory, Analog Non-Volatile Compute-in-Memory. By Application: Real-Time Transaction Processing, Real-Time Analytics & BI, Caching & Session Management, Fraud Detection & Risk Management, AI/ML Feature Stores & Vector Search, Edge AI Inference, ERP & Supply Chain. By Organization Size: Large Enterprises, SMEs. By Industry Vertical: BFSI, Retail & E-Commerce, Telecommunications & IT, Manufacturing, Healthcare & Life Sciences, Government & Defense, Media, Gaming & Entertainment, Transportation & Logistics. |
| Countries Covered | North America: U.S., Canada. Europe: Germany, U.K., France, Netherlands, Nordic Countries, Italy, Spain, Rest of Europe. Asia-Pacific: China, India, Japan, South Korea, Australia, Singapore, Rest of Asia-Pacific. Latin America: Brazil, Mexico, Colombia, Chile, Argentina, Rest of Latin America. Middle East & Africa: UAE, Saudi Arabia, Israel, South Africa, Kenya, Rest of Middle East & Africa. |
| Key Companies | SAP, Oracle, Microsoft, AWS, Google, IBM, Redis, Aerospike, SingleStore, GridGain, Hazelcast, GigaSpaces, Samsung Electronics, SK hynix, d-Matrix, EnCharge AI, Axelera AI, and Mythic. |
Report overview
Segments covered: component, deployment, hardware technology, application, organization size, industry vertical.
The growth of this market is mainly driven by demand for real-time data processing and AI, the modernization of enterprise resource planning systems onto in-memory platforms, and the memory wall in AI hardware, which is driving the development of processing-in-memory and compute-in-memory technologies. However, rising DRAM costs, licensing changes and fragmentation in open-source in-memory data platforms, the cost and complexity of migrating legacy systems, and the immaturity of in-memory computing hardware restrain the growth of this market.
Furthermore, vector search and AI feature serving, real-time payments and fraud detection, and energy-efficient AI inference at the edge and in data centers are expected to offer growth opportunities for the stakeholders in this market. However, ensuring data persistence and durability in volatile memory, securing in-memory data stores, integrating processing-in-memory into programming models and standards, and the shortage of skills for distributed in-memory architectures remain major challenges impacting the growth of this market. Additionally, the convergence of in-memory data platforms with vector search and AI, the rise of open-source forks and managed cloud services, memory expansion and pooling through Compute Express Link, and processing-in-memory in AI accelerators are prominent trends in this market.
The In-Memory Computing Market comprises software, hardware, and services that store and process data primarily in main memory, or perform computation within or adjacent to memory, rather than repeatedly moving data between memory, storage, and processors. The software segment covers in-memory databases such as SAP HANA and Oracle Database In-Memory, in-memory data grids and caches such as Redis, Valkey, Hazelcast, GridGain, and cloud-managed caching services, in-memory analytics, and in-memory stream processing. The hardware segment covers processing-in-memory (PIM) devices that integrate compute into DRAM and high-bandwidth memory, compute-in-memory (CIM) accelerators that perform AI calculations within SRAM or non-volatile memory arrays, and in-memory computing appliances. Services include consulting, implementation, migration, and managed services. General-purpose memory chips and servers sold without in-memory computing software or functionality are excluded. The ecosystem spans database and data platform vendors, cloud providers, memory manufacturers, AI chip startups, systems integrators, and enterprises.
Enterprise adoption of in-memory platforms is being driven by ERP modernization and real-time analytics. SAP, whose S/4HANA ERP suite runs on the SAP HANA in-memory database, reported 2025 cloud revenue of EUR 21.02 billion, up 23%, Cloud ERP Suite revenue of EUR 18.12 billion, up 28%, and a record total cloud backlog of EUR 77.29 billion, up 30% at constant currencies, while software license revenue fell 29% to EUR 990 million as customers moved to subscription. SAP reported that its Business AI was included in two-thirds of its fourth-quarter 2025 cloud order entry, reflecting the growing link between in-memory data platforms and AI. The approaching end of mainstream maintenance for SAP's legacy ERP Central Component in 2027 is prompting migrations across its large installed base.
The economics of in-memory computing are changing. AI demand has tightened memory supply: Samsung and SK hynix reported that 2026 orders exceeded their capacity, Samsung raised prices for some memory chips by up to 60% in December 2025, DRAM average prices were expected to rise 50% to 55% year over year in the fourth quarter of 2025, and data centers were expected to consume up to 70% of memory supply in 2026, raising the cost of memory-heavy deployments. The open-source ecosystem has also shifted: Redis moved to source-available licenses in March 2024, the Linux Foundation launched the Valkey fork later that month with support from Amazon Web Services, Google Cloud, Oracle, and others, and Redis added an AGPL open-source option with Redis 8 in 2025.
In hardware, the memory wall has become a central constraint in AI. Moving data between memory and processors consumes more energy and time than computation itself, and AI accelerators depend on scarce high-bandwidth memory. Memory makers have introduced processing-in-memory products, including Samsung's HBM-PIM in 2021 and SK hynix's GDDR6-based Accelerator-in-Memory in 2022, and startups such as d-Matrix, EnCharge AI, and Axelera AI are developing digital and analog in-memory computing accelerators for AI inference, with EnCharge AI raising USD 100 million in 2025. IBM research has demonstrated analog in-memory AI chips and its NorthPole architecture, which places memory and compute together. Commercial adoption of in-memory computing hardware remains early but is expected to grow as AI inference scales.
Market dynamics
18 factors across 5 forcesDemand for Real-Time Data Processing and AI
Demand for real-time data processing and AI is a major factor driving the In-Memory Computing Market. Businesses increasingly need to analyze transactions, detect fraud, personalize customer experiences, and serve AI models in milliseconds, which requires data to be held and processed in memory rather than retrieved from disk. SAP reported that its Business AI was included in two-thirds of its cloud order entry in the fourth quarter of 2025, and that its cloud revenue rose 23% to EUR 21.02 billion in 2025, reflecting enterprise demand for AI embedded in in-memory business platforms. In-memory databases and caches are also central to AI applications as vector stores, feature stores, and session caches for large language model applications, and real-time payment systems such as the U.S. FedNow service launched in 2023 and India's Unified Payments Interface require instant processing at very high volumes. As AI and real-time applications proliferate, in-memory computing is becoming a core layer of enterprise and cloud data architectures.
Modernization of Enterprise Resource Planning onto In-Memory Platforms
The modernization of enterprise resource planning systems onto in-memory platforms is significantly supporting the market. SAP's S/4HANA suite, which runs exclusively on the SAP HANA in-memory database, is replacing SAP's legacy ERP Central Component, for which mainstream maintenance is scheduled to end in 2027, and SAP reported Cloud ERP Suite revenue of EUR 18.12 billion in 2025, up 28%, with a record total cloud backlog of EUR 77.29 billion, up 30% at constant currencies. Customers including adidas, BioNTech, Toyota, Lockheed Martin, and Deutsche Bundesbank signed transformation agreements in the fourth quarter of 2025. SAP guided to 2026 cloud revenue of EUR 25.8 billion to EUR 26.2 billion, growth of 23% to 25%. These migrations drive demand for HANA database capacity, cloud infrastructure, and migration services across the installed base.
The Memory Wall in AI Hardware
The memory wall in AI hardware is driving the development of processing-in-memory and compute-in-memory technologies. AI workloads are increasingly limited by the bandwidth and energy required to move data between memory and processors rather than by computation, and high-bandwidth memory has become one of the scarcest components in AI systems, with Samsung and SK hynix reporting that 2026 orders exceeded their capacity. The International Energy Agency projects that data center electricity consumption will more than double from about 415 terawatt-hours in 2024 to around 945 terawatt-hours by 2030, increasing pressure to reduce the energy cost of AI. Memory makers have developed processing-in-memory products, including Samsung's HBM-PIM and SK hynix's Accelerator-in-Memory, and startups such as d-Matrix, EnCharge AI, and Axelera AI are commercializing in-memory computing accelerators that perform matrix operations where data is stored, promising large gains in energy efficiency for AI inference.
Table of contents
14 chapters · 172 sections · 326 pages · click to expandSegmental analysis
| Segment | Largest share (2026) | Fastest growth (2026–2036) |
|---|---|---|
| By Component | Software | Hardware |
| By Deployment | Cloud | Hybrid |
| By Hardware Technology | Digital SRAM-Based Compute-in-Memory | Analog Non-Volatile Compute-in-Memory |
| By Application | Real-Time Transaction Processing | AI/ML Feature Stores & Vector Search |
| By Industry Vertical | BFSI | Retail & E-Commerce |
By Component
- The Software segment is expected to account for the largest share of the market.
- The large share of this segment is mainly due to the widespread use of in-memory databases, caches, and analytics platforms across enterprises and cloud applications.
- However, the Hardware segment is projected to register the highest CAGR during the forecast period.
- The rapid growth of this segment is attributed to the memory wall in AI and the commercialization of processing-in-memory and compute-in-memory accelerators.
By Deployment
- The Cloud segment is expected to account for the largest share of the market.
- The large share of this segment is mainly due to the growth of managed in-memory database and caching services and cloud ERP, reflected in SAP's Cloud ERP Suite revenue of EUR 18.12 billion in 2025.
- However, the Hybrid segment is projected to register the highest CAGR during the forecast period.
- The rapid growth of this segment is attributed to data sovereignty requirements and the need to combine on-premises and cloud in-memory systems.
By Hardware Technology
- The Digital SRAM-Based Compute-in-Memory segment is expected to account for the largest share of the hardware market.
- The large share of this segment is mainly due to its compatibility with standard CMOS processes and more mature commercial products.
- However, the Analog Non-Volatile Compute-in-Memory segment is projected to register the highest CAGR during the forecast period.
- The rapid growth of this segment is attributed to its potential for very high energy efficiency in AI inference.
By Application
- The Real-Time Transaction Processing segment is expected to account for the largest share of the market.
- The large share of this segment is mainly due to the use of in-memory databases for high-volume transactional systems in finance, retail, and telecommunications.
- However, the AI/ML Feature Stores & Vector Search segment is projected to register the highest CAGR during the forecast period.
- The rapid growth of this segment is attributed to the rapid deployment of generative AI and retrieval-augmented applications.
By Industry Vertical
- The BFSI segment is expected to account for the largest share of the market.
- The large share of this segment is mainly due to demand for real-time payments, trading, risk management, and fraud detection.
- However, the Retail & E-Commerce segment is projected to register the highest CAGR during the forecast period.
- The rapid growth of this segment is attributed to real-time personalization, inventory visibility, and AI-driven customer experiences.
Geographic analysis
North America
Largest shareIn 2026, North America is expected to account for the largest share of the global In-Memory Computing Market. The region's dominance is supported by the headquarters of leading cloud providers and in-memory platform vendors, large financial services and technology sectors, and early adoption of AI. The U.S. is home to Oracle, Microsoft, Amazon Web Services, Google, Redis, Aerospike, SingleStore, and GridGain, and to compute-in-memory startups such as d-Matrix and EnCharge AI, which raised USD 100 million in 2025. The U.S. Federal Reserve's FedNow instant payment service, launched in July 2023, and demand for real-time fraud detection and AI applications support investment in in-memory platforms. Canada's financial services and telecommunications sectors are also significant users. Mexico, Colombia, Chile, and Argentina have fast-growing digital banks, fintechs, and e-commerce companies that rely on in-memory databases and caches for real-time transactions, fraud detection, and customer personalization, increasingly consumed as managed cloud services. North America
Europe
Europe is expected to account for a significant share of the market and is home to SAP, the largest provider of in-memory enterprise software through SAP HANA, which reported 2025 cloud revenue of EUR 21.02 billion and a total cloud backlog of EUR 77.29 billion. Migration from SAP's legacy ERP ahead of the end of mainstream maintenance in 2027 is a major driver in Germany, France, the U.K., the Nordic countries, and across Europe. The EU's Digital Operational Resilience Act, applicable from January 2025, is raising requirements for financial system resilience, and Netherlands-based Axelera AI is developing in-memory computing chips for edge AI. Data sovereignty requirements under the GDPR are driving demand for hybrid and sovereign cloud deployments. Europe
Asia-Pacific
Fastest growthHowever, Asia-Pacific is projected to register the highest CAGR during the forecast period. The rapid growth of this region is attributed to rapid digitalization, very large real-time payment volumes, and the region's leadership in memory manufacturing. India's Unified Payments Interface processes billions of transactions each month, requiring in-memory processing for authorization and fraud detection, and SAP reported strong cloud performance in its Asia Pacific Japan region in 2025. South Korea's Samsung and SK hynix, which reported that 2026 memory orders exceeded capacity, are developing processing-in-memory technologies, including HBM-PIM and Accelerator-in-Memory. China, Japan, Singapore, and Australia have large financial, e-commerce, and telecommunications sectors adopting in-memory platforms. Asia-Pacific
Latin America
Latin America is expected to account for a moderate share of the market, with growth driven by real-time payments and digital banking. Brazil's Pix instant payment system, launched in 2020, processes billions of transactions per month, and Brazil was among the countries where SAP reported notable cloud performance in the fourth quarter of 2025. Latin America
Middle East & Africa
The Middle East & Africa is expected to register strong growth. Gulf countries, including the UAE and Saudi Arabia, are investing heavily in digital government, banking modernization, and AI infrastructure, including the Stargate UAE cluster announced in May 2025, creating demand for in-memory platforms and AI inference hardware. Israel is home to in-memory computing companies such as GigaSpaces and to development centers of global technology firms, and South Africa and Kenya have large mobile money and digital banking markets that require real-time transaction processing. The initial 1-gigawatt Stargate UAE cluster is part of a planned 5-gigawatt AI campus in Abu Dhabi, and Saudi Arabia's Humain agreed in May 2025 to purchase 18,000 NVIDIA GB300 chips, investments that will require in-memory data and caching layers for AI applications as well as memory-efficient inference hardware. Middle East & Africa
Competitive landscape
The global In-Memory Computing Market comprises large enterprise software and database vendors, cloud providers offering managed in-memory services, specialized in-memory database and data grid companies, open-source communities, memory manufacturers developing processing-in-memory, and startups developing compute-in-memory AI accelerators. Competition centers on performance and latency, scalability and resilience, AI and vector search capabilities, licensing and openness, cloud integration, total cost of ownership, and, in hardware, energy efficiency and software support.
Leading companies are adding vector search and AI capabilities to in-memory platforms, shifting to cloud subscription models, responding to open-source forks with new licensing and managed services, adopting CXL and tiered memory, and investing in processing-in-memory and compute-in-memory hardware.
The report provides a comprehensive competitive assessment of the leading companies operating in the global In-Memory Computing Market. The key players profiled in the report include SAP SE (Germany), Oracle Corporation (U.S.), Microsoft Corporation (U.S.), Amazon Web Services, Inc. (U.S.), Google LLC (U.S.), IBM Corporation (U.S.), Redis Ltd. (U.S.), Aerospike, Inc. (U.S.), SingleStore, Inc. (U.S.), GridGain Systems, Inc. (U.S.), Hazelcast, Inc. (U.S.), GigaSpaces Technologies Ltd. (Israel), Samsung Electronics Co., Ltd. (South Korea), SK hynix Inc. (South Korea), d-Matrix Corporation (U.S.), EnCharge AI, Inc. (U.S.), Axelera AI B.V. (Netherlands), and Mythic, Inc. (U.S.).
- SAP
- Oracle
- Microsoft
- AWS
- IBM
- Redis
- Aerospike
- SingleStore
- GridGain
- Hazelcast
- GigaSpaces
- Samsung Electronics
- SK hynix
- EnCharge AI
- Axelera AI
- Mythic
Expert perspectives
In-memory computing is expanding on two fronts. In software, in-memory databases and caches are becoming the real-time and AI layer of enterprise and cloud architectures, as reflected in SAP's EUR 21.02 billion in 2025 cloud revenue and its report that Business AI featured in two-thirds of fourth-quarter cloud orders. In hardware, the memory wall in AI is driving processing-in-memory and compute-in-memory accelerators from research toward commercial products.
Three structural changes are expected to shape the market through 2036. First, in-memory platforms will converge with AI, serving as vector stores, feature stores, and semantic caches, and will increasingly be consumed as managed cloud services amid the realignment triggered by the 2024 Redis license change. Second, memory cost and capacity, affected by AI-driven shortages and price increases of up to 60%, will push adoption of CXL memory pooling and tiered memory. Third, in-memory computing hardware will gain a role in energy-efficient AI inference as power constraints tighten.
For companies planning entry or expansion, the most attractive positions over the forecast period are likely to be found in in-memory vector and feature stores, managed in-memory database services, ERP migration services, real-time payment and fraud platforms, CXL-based memory solutions, and compute-in-memory accelerators for inference. The principal risks are rising DRAM costs, open-source licensing fragmentation, migration complexity, and the immaturity of in-memory hardware.
Customer perspectives
Insights gathered during primary interviews with chief data officers, database architects, SAP program leaders, and AI hardware engineers highlight where purchasing priorities are shifting. The following perspectives reflect recurring themes raised across these discussions.
“This reflects the central role of in-memory computing in payments and the pressure of rising memory costs.”
“This indicates ERP modernization as a driver and skills and complexity as constraints.”
“This points to licensing realignment and AI use cases reshaping platform choices.”
Frequently asked questions
The global In-Memory Computing Market is estimated at USD 27.36 billion in 2026.
Cite this report
Meticulous Research. (2026). In-Memory Computing Market - Opportunity Analysis and Industry Forecast (2026-2036) (Report No. MR-2277). Meticulous Market Research Pvt. Ltd. https://www.meticulousresearch.com/product/in-memory-computing-market-6960