In-memory computing (IMC) performs computations directly within memory, bypassing slow data movement between memory and processors. IMC means using a type of middleware software that allows one to store data in RAM, across a cluster of computers, and process it in parallel. Consider operational datasets typically stored in a centralized database which you can now store in “connected” RAM across multiple computers. RAM is roughly 5,000 times faster than traditional spinning disk. Add to the mix native support for parallel processing, and things get very fast. Really, really, fast. This approach significantly improves performance by reducing data access latency, making it ideal for demanding applications like real-time analytics, machine learning, and scientific computing. 

All you want to Know How It Works

RAM storage and parallel distributed processing are two fundamental pillars of in-memory computing. While in-memory data storage is expected of in-memory technology, the parallelization and distribution of data processing, which is an integral part of in-memory computing, calls for an explanation.

• Eliminating Data Movement: 

Traditional computers require data to be moved from storage (like hard drives or SSDs) to memory and then to the processor for computation. IMC eliminates this bottleneck by storing and processing data primarily in RAM. 

Parallel distributed processing capabilities of IMC are... a technical necessity. Consider this: a single modern computer can hardly have enough RAM to hold a significant dataset. In fact, a typical x86 server today would have somewhere between 32GB to 256GB of RAM. Although this could be a significant amount of memory for a single computer, that’s not enough to store many of today’s operational datasets that easily measure in terabytes.

• Von Neumann Architecture Alternative: 

IMC is a departure from the traditional Von Neumann architecture, where processing and memory are separate. In IMC, computation happens within the memory units themselves, or in close proximity to them. 

• In-Memory Processing Platforms: 

Software platforms can pool the RAM from multiple computers in a cluster to create a large, shared memory space for complex calculations. 

IMC software is designed from the ground up to store data in a distributed fashion, where the entire dataset is divided into individual computers’ memory, each storing only a portion of the overall dataset. Once data is partitioned - parallel distributed processing becomes a technical necessity simply because data is stored this way.

• Hardware Integration: 

At the hardware level, IMC can involve placing processing elements (like compute engines) directly within memory arrays or turning memory arrays into compute engines. 

Benefits

• High Speed and Low Latency: 

By keeping data in fast RAM, IMC drastically reduces the time it takes to access and process information. 

• Increased Throughput: 

Handling large volumes of data becomes more efficient, leading to higher throughput for complex operations. 

• Energy Efficiency: 

Reducing the physical movement of data between memory and processing units can lead to significant energy savings. 

IMC: What Is Good About It?

If one wants a 2-3x performance or scalability improvements - flash storage (SSD, Flash on PCI-E, Memory Channel Storage, etc.) can do the job. It is relatively cheap and can provide that kind of modest performance boost. however, what a difference in-memory computing can make, consider this real-live example...

In Ramco’s Payce we were given a job to process 1 lakh payslips in an hour. The problem involves loading the required data from DB into In-Memory, Processing it and pushing back to DB. Here using Ramco’s Inbuild IMC software we demonstrated and generated the required details within the time limits. We also containerised the servers to reduce the cost further. With 10 containers with mere 4-CPU, 16 GB machines we were able to achieve it. This had a tremendous impact on lowering the cost incurred.

One lakh real time payslips getting generated in an hour…, That is the in-memory computing difference -- not just 2-3x times faster; more than 10x faster than theoretically possible even with the most expensive flash-based storage available on today’s market (forget about spinning disks).

Importantly, that performance translates directly into the clear business value:

You can use less hardware to support the required performance and throughput SLAs, get better data center consolidation, and significantly reduce capital costs, as well as operational and infrastructure overhead, and

you can also significantly extend the lifetime of your existing hardware and software by getting increased performance and improve its ROI by using what you already have longer and making it go faster. You can extend the lifetime of the existing hardware by making it as a IMC cluster.

And that’s what makes IMC such a hot topic these days: the demand to process ever growing datasets in real-time can now be fulfilled with the extraordinary performance and scale of in-memory computing, with economics so compelling that the business case becomes clear and obvious.

Applications

• Real-Time Analytics / Use Cases: For applications that require immediate data analysis and decision-making. Eg Payroll Processing, Men & Material Soln, and Route Planning etc
• Machine Learning and Deep Learning: Crucial for accelerating AI algorithms by keeping large model parameters and intermediate data on-chip. 
• Scientific Computing: Handling large, complex calculations in fields like scientific research. 
• Database Operations: Improving the performance of database queries by keeping data in memory. 
• Investment banking: Where large transaction data are handled
• Insurance claim processing & modelling
• Real-time ad platforms
• Real-time sentiment analysis
• Merchant platform for online games
• Hyper-local advertising
• Geospatial/GIS processing
• Medical imaging processing
• Natural language processing & cognitive computing
• Real-time machine learning
• Complex event processing of streaming sensor data

In many of these real-life deployments in-memory computing was an enabling technology, the technology that made these particular systems possible to consider and ultimately possible to implement.

The bottom line is that IMC is beginning to unleash a wave of innovation that’s not built on Big Data per se, but on Big Ideas, ideas that are suddenly attainable. It’s blowing up the costly economics of traditional computing that frankly can’t keep up with either the growth of information or the scale of demand.
As the Internet expands from connecting people to connecting things, devices like refrigerators, thermostats, light bulbs, jet engines and even heart rate monitors are producing streams of information that will not just inform us, but also protect us, make us healthier and help us live richer lives. We’ll begin to enjoy conveniences and experiences that only existed in science fiction novels. The technology to support this transformation exists today - and it’s called in-memory computing.