This week Samsung unveiled its 2nd generation SmartSSD, a computational storage device that handles data processing on-board. This means the SSD doesn’t have to communicate and share data with the CPU, GPU, or RAM, thereby improving performance across the board.
For example, Samsung’s new SmartSSD doesn’t lose speed or efficiency when a PC transfers app or program data to a CPU, nor does it need to send off graphics to render on a GPU. Instead, the SmartSSD does all that itself, freeing up all those CPU and GPU cores for other tasks.
The 2nd generation drive boasts massive gains, which could be a game changer. In the press release, Samsung reports a 50% increase in processing speeds, 70% efficiency savings, and reduced CPU utilization by nearly 97%.
“The SmartSSD is playing an increasingly important role. Especially with the growth of next-generation technologies such as AI, machine learning, and 5G/6G, which require large amounts of data processing.”
With regular solid-state drives, the storage device sends information to the CPU or GPU, which then has to get read, processed, and completed. Samsung’s SmartSSD does all of that on its own. This technology can increase performance, reduce bottlenecks, and deliver vastly better performance and efficiency across the board.
Samsung’s new SmartSSD uses an AMD Xilinx Versal FPGA system to handle all the processing tasks. For now, these next-gen smart SSDs are geared toward AI and machine learning business uses, but in the future, we can expect this type of technology to trickle down into mainstream products.
Unfortunately, Samsung didn’t share pricing details nor mention a release date.