Graid’s SupremeRAID expertise combines a GPU with a software-defined storage (SDS) stack to ship frighteningly excessive storage speeds that go above what most servers can hit right now with standard RAID configurations.
By pairing Graid’s SupremeRAID card with the Nvidia RTX A2000 GPU, Storage Overview managed to hit 5.6M IOPS at 1.1ms utilizing the perfect RAID efficiency at 4K chunk measurement, whereas Graid landed a results of 28.5M IOPS in the identical take a look at.
The system they used fitted a Supermicro AS-2125HS-TNR server with two AMD EPYC 9654 CPUs, 384GB DRAM, and 24 3.84TB KIOXIA’s CM7-R Gen5 SSDs. They configured the drives right into a RAID5 configuration with a 4K stripe measurement for Graid.
The perfect RAID configuration hit 205k IOPS at 15.01ms with 64K chunk, in comparison with Graid which hit 2.02M IOPS at 1.52ms.
RAID vs SupremeRAID
RAID playing cards are restricted to 28GB/s speeds with Gen4 SSDs, with 4 Gen4 SSDs saturating one RAID card, and a number of other required to maximise utilization of a 24-bay server.
SupremeRAID, then again, can assist 32 drives in a single system and is not restricted by most PCIe slot bandwidth. Utilizing the Graid system, due to this fact, may go to resolve the bottlenecks in a traditional server confiugration.
It additionally provides the GPU license to deal with the drive administration and information safety – assuaging this from the CPU and releasing it to run purposes.
Storage Overview used Nvidida A2000 GPUs in testing, that are comparatively cheap, to hit 28M IOPS alongside speeds of as much as 260GB/s. That is regardless of the cardboard utilizing a Gen4 interface, however this does not stop it from reaching efficiency ranges of Gen5 SSDs.
The quickest SSDs in isolation can hit speeds approaching 15GB/s, in contrast, with Phison’s PS5026-E26 Max14um Gen5 SSDfor example, reaching sequential learn speeds of 14,175MB/s and writes of 12,741MB/s
Though the publication examined this on Nvidia A2000 GPUs, in the meantime, Graid’s software program is appropriate with nearly any Nvidia GPU. These are simple to put in and do not require extra battery energy, it added.
