2 ms·
If these drives are just fpgas sitting on the existing interface, so still hitting the PCIe limits, then this is un-impressive. If we start seeing multiples of
by sitkack 6y ago
If these drives are just fpgas sitting on the existing interface, so still hitting the PCIe limits, then this is un-impressive. If we start seeing multiples of device bandwidth available to the FPGA, then we can see huge cost savings.
- wtallis 6y agoI agree that on its own, it doesn't seem too interesting for the FPGA to be accessing the flash through the same PCIe x4 interface that the host system could use. But servers with 24+ NVMe SSDs don't always have the bandwidth to saturate all the SSDs simultaneously; they're often connected through a PCIe fanout switch that has just an x16 uplink (or an x16 per CPU socket). Having an accelerator to offload eg. search means the drives in aggregate don't have to send as much data up to the CPUs (or NICs). Even if these drives have what appears to be a sub-optimal design, they can still help alleviate bottlenecks elsewhere.