Microsoft Research has announced Project HSD, detailing its work on a new holographic storage device (HSD) built for cloud-scale data that has a potential to be a successor to today’s NAND flash and spinning hard disk drives.
Project HSD, a collaboration between Microsoft Research Cambridge and Microsoft Azure, aims to exploit the commoditization of high-resolution smartphone cameras to reimagine holographic storage, which was first thought of in the 1960s after laser was invented.
Mark Russinovich, CTO of Azure, announced Project HSD atthis week.
“In Project HSD we are exploring the use of holographic storage in rewritable electro-optic materials for warm data storage to see if this technology makes sense in the cloud era,” the group says.
It’s got a different goal to Microsoft’s Project Silica for archival ‘write once, read many times’ data that’s stored on quartz glass.
Microsoft Research says it’s achieved 1.8x higher density than the state of the art for volumetric holographic storage, and it’s aiming to increase density and access rates further.
The team note that flash storage offers high access rates but is relatively expensive, leaving many cloud applications’ data stored and accessed from cheaper but slower HDDs.
Hence, they’re hoping to exploit the ability to write and read multiple bits in parallel available with holographic storage to provide high access rates and high data throughput.
Taking a fresh look at storage hardware for the cloud also has the potential to free it from 2.5-inch and 3.5-inch hard disk form factors and design new ‘rack-scale’ hardware.
The group explains that a key ingredient to better holographic storage is the arrival of high-resolution camera technology from the smartphone industry, which has allowed it to move complex tasks, like pixel matching, from the optical hardware to software.
“In the previous state of the art, it was necessary to use complex optics to achieve one-to-one pixel matching from the display device to the camera to maximize the density,” the group explains.
“Today, we can leverage commodity high-resolution cameras and modern deep-learning techniques to shift the complexity into the digital domain. This lets us use simpler, cheaper optics without pixel matching and compensate for the resulting optical distortions with commodity hardware and software.”
Microsoft says this approach also reduces manufacturing tolerances, because the system can be compensated and calibrated at runtime in software.
The group still needs to overcome a few key obstacles to achieve cloud-scale holographic storage with high access speeds. As the researchers explain, holographic data storage relies on light to record data pages that are stored as a tiny hologram inside a crystal.
“The hologram occupies a small volume inside the crystal, which we think of as a zone, and multiple pages can be recorded in the same physical volume or zone.”
The challenge for the group is to find a way to achieve high access rates across multiple zones.
“While we have seen compelling performance in terms of write/read time and storage density that we can obtain in a single zone, the challenge to make holographic storage practical for the cloud is to develop approaches that allow for scaling of the storage capacity by increasing the number of zones while maintaining the same access rates across multiple zones,” they explain.
Microsoft Research’s video shows how holographic storage works. Source: Microsoft/YouTube