Persistent Memory (PMem) is non-volatile, byte addressable, low latency memory with densities greater than or equal to Dynamic Random Access Memory (DRAM). It is beneficial because it can dramatically increase system performance and enable a fundamental change in computing architecture. In this post, we will address the issue of Windows PC configured to use persistent memory slow on startup.

Persistent Memory is used today in database, storage, virtualization, big data, cloud computing/IoT, and artificial intelligence applications. Persistent Memory is supported by an industry-wide hardware, software, standards, and platform ecosystem.

PMem technologies allow development of products with the attributes of both storage and memory. The products are persistent, like storage, meaning they hold their content across power cycles, and they are byte-addressable, like memory, meaning programs can access data structures in place.

Unexpectedly slow startup on a Windows 10 computer configured to use PMem

You’ll notice that after you configure a Windows-based computer to use large amounts of memory, including persistent memory, the computer takes longer than expected to start up. Additionally, increased CPU usage occurs for a short time after startup – when an application frees and reallocates large ranges of memory in rapid succession.

PMem is used for both Windows clients and Windows servers. If you configure a system to use large amounts of memory, including persistent memory, additional startup time is required. During restarts and during freeing memory at runtime, the system must clear all physical memory before that memory can be used.

Startup performance is affected only if the persistent memory is used in memory mode. If the persistent memory is used as a storage device, system startup time is not affected.

According to Microsoft, this is a known issue which will be addressed in future Windows releases that will provide additional optimizations in the startup path.

Hope you find the information in this post clarifying!