Industry challenge: Developing storage for the incoming data deluge
While we weren't looking, WD assembled a formidable array of storage technologies, beyond their leadership in disks and SSDs. That includes all-flash arrays, scale-out object storage, and industry-leading enclosures, with patented cooling and anti-vibration technology. Today's announcements[1] take them to another level.
Big data, fast data
CPU performance hit a wall years ago. Multi-core chips have helped, but GPUs - with hundreds or thousands of execution units - have kept the industry pushing the performance envelope, especially for machine learning. So there's been a renewed focus on moving processing closer to data, as the growth of big data has made it costly to move the data to processors.
Fast SSDs, including the NVMe/PCIe ones pioneered by WD acquisition Fusion-io, are still improving in performance, capacity, and cost. But the cloud - and most enterprises - are still relying on hard drives for capacity, where WD's HGST unit has led the way with their helium filled drives[2].
At the same time, the growth in realtime systems, such as computer vision for self-driving cars, and streaming data analysis, have put a premium on managing data that is both fast and big. In what may be their most consequential announcement this morning, WD is now offering flexible and efficient drive enclosures with powerful built-in servers that support the Docker container framework. Moving apps to your data has never been easier or more cost-effective.
Who is WD?
WD, founded in 1970, has a long and varied history. It began as a chipmaker - for a time it was the world's largest calculator chip company - but it found lasting success creating hard drive controllers, especially the ATA drives popularized