What Is Storage-Class Memory?

What Is Storage-Class Memory?

Enterprise storage administrators have long had to balance the need for better data access speed with budgetary constraints. Ultrafast data access exists in forms such as the latest double data rate (DDR) synchronous dynamic random-access memory (SDRAM). But, DDR SDRAM is also ultraexpensive.

Another storage option is quickly gaining popularity as it offers higher performance than NAND Flash at lower costs than DDR SDRAM. It’s called storage-class memory (SCM).

SCM: A Bridge Between DDR and Today’s Enterprise Flash Drives

Consider the spectrum of enterprise storage options and the price vs. performance ratios they provide. Tape is cheap but slow, while DRAM is super-speedy but exceedingly pricey. SCM falls between the traditional NAND storage used in today’s enterprise flash drives and DRAM, as shown in the chart below. Not surprisingly, SCM shares some features with both DRAM and NAND flash storage. It delivers byte-addressable data access like you get with memory, at prices closer to traditional storage solutions.

SCM is also called persistent memory because it is non-volatile.. That means data won’t be lost if the storage system crashes or loses power. This is a big advantage over DRAM, which is volatile, or temporary, and is lost when the power is off.

SCM can be accessed similarly to DRAM, acting as another pool of server memory. Data accessed from this SCM memory pool is significantly faster than data accessed via more traditional interfaces like PCIe or SAS. In fact, SCM can read and write data up to 10 times faster than NAND flash drives, while being approximately half the price of DDR SDRAM.

The Advantages of SCM

With SCM technology, enterprises can take advantage of a new tier of high-performance storage that delivers a performance/price ratio previously unavailable. SCM is ideal for ultrafast, mission-critical databases and applications that require high durability, low latency, and a high level of data consistency.

Other advantages include:

• Lower latency than traditional NAND flash storage
• Faster processing of workloads that are write-intensive, such as online transaction processing (OLTP) workloads
• Lower cost than DRAM storage
• When used in Persistent Memory mode, eliminates potential for data loss due to power failures or system crashes

Put the Technology to Good Use

While SCM is significantly less expensive per GB than system RAM, it still costs more than other solid state storage options. SCM is ideal for computing environments that need access to complex data sets that can fit into a pool of SCM. It’s also a smart choice for environments that are more sensitive to power outages or that would incur more than the usual damage or expense from the loss of in-memory data.

Some use cases include:

• Analysis of cyber threats: Systems could more quickly parse incoming data sets to identify and mitigate anomalies.
• Fraud detection: Fast, accurate data analytics could help financial institutions and insurance companies pinpoint fraudulent transactions.
• Stock market trading: Data stored in SCM could be processed and trades executed in mere fractions of seconds.

How Pure Storage Leverages SCM

Pure Storage® offers SCM storage solutions that can help your organization boost efficiency without breaking your budget. DirectMemory™ Cache software is integrated directly into the core operating system of FlashArray™. It allows FlashArray to leverage SCM to deliver improved read performance for databases, analytics, and reporting. It’s designed specifically for read-intensive, high-performance workloads for data sets within the three to six TB range. DirectMemory Cache requires DirectMemory Modules, which can be installed in both FlashArray//X and FlashArray//XL for enhanced read performance.