A start-up has tested a new device designed to expand servers with additional memory via a CXL 2.0 connection, simply by plugging it into a laptop.
FADU, known to produce SSDs and other storage units, demonstrated its CXL 2.0 switch concept device at Flash Memory Summit (FMS) 2023, according Serve the House. This was a field programmable gate array (FPGA) switch, the company said, and it was designed to extend IT infrastructure with a monumental amount of memory to support next-gen workloads like AI.
A similar unit also on display at FMS 2023, the XConn Electronic design.
What’s new in CXL 2.0?
CXL 2.0 works over PCIe and is an interface between top CPUs, memory units, accelerators and other components. It is the latest standard in CXL technology.
The previous edition, CXL 1.1, focused on expanding systems with memory instead of relying on RAM that was traditionally built into the computer. Instead of installing RAM units, systems would get a memory boost by connecting it directly to the CPU or GPU, or to another component with a PCIe slot.
This is particularly useful for hyperscalers and large data center operators, as it means you can plan for when an increase in memory usage will occur during particularly intensive workloads.
The CXL 2.0 switch, known as Apollo, supports both CXL 2.0 and PCIe Gen 5 standards in a single chip, and is designed to reduce latency and power consumption during use. You also don’t need two separate chips for CXL and PCIe.
Additionally, processors and accelerators can package memory into one space, negating the limited capacity in data centers caused by the tight coupling between CPUs and RAM units.
More and more devices are coming to market to improve data center performance in the generative AI era, with FADU’s concept following the Xconn unit that also launched on FMS.