Overview and Research Directions of Storage Class Memory (SCM)
This article introduces Storage Class Memory (SCM), a non‑volatile memory technology faster than NAND but slower than DRAM, outlines its main types such as PCM, ReRAM, MRAM and NRAM, and discusses eight research topics ranging from system organization to security and software optimization.
Storage Class Memory (SCM) is a non‑volatile memory whose access speed is slightly slower than DRAM but far faster than NAND flash. The article briefly summarizes the characteristics and usage methods of SCM.
Current mainstream SCM technologies include PCM, ReRAM, PCRAM, RRAM, MRAM and NRAM, each with distinct operating principles and representative vendors such as Intel (3D XPoint), HPE, Crossbar, Toshiba, and Everspin.
Key research directions are presented:
System organization methods for SCM‑based storage, addressing the mismatch between traditional memory management and SCM’s unique properties.
Access methods for SCM, exploring byte‑addressable reads/writes, block‑level I/O, NVMe compatibility, and latency reduction.
Data reliability, focusing on error‑correction overhead, configurable checksum algorithms, wear‑leveling, and bad‑block reuse.
Data security, proposing OS‑level encryption and access‑control mechanisms to mitigate persistent‑data attacks.
Software optimization, including hot‑data identification, SCM‑aware file systems, memory allocation strategies, and scheduling algorithms.
Hardware prototype development, noting the scarcity of real SCM chips and the reliance on simulators like PCRAMsim, Simics, M5, DRAMsim, and GEM5.
Transactional storage systems built on SCM, addressing interface design, fault recovery, scalability, and durability.
Upper‑layer applications, such as in‑memory databases, real‑time analytics, and big‑data services that can benefit from SCM’s performance.
The article also contains promotional links for downloading related market analysis reports and a commercial offer for a bundled technical documentation package.
Architects' Tech Alliance
Sharing project experiences, insights into cutting-edge architectures, focusing on cloud computing, microservices, big data, hyper-convergence, storage, data protection, artificial intelligence, industry practices and solutions.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.