Western Digital is formally coming into a brand new period, rebranding itself merely as WD and dropping an enormous roadmap that targets the guts of the AI revolution. At its Innovation Day 2026, the storage large showcased the way it plans to maintain up with the insatiable knowledge calls for of synthetic intelligence.
The headline act is a transparent path to 100TB arduous drives, but it surely isn’t nearly uncooked capability anymore.
WD is introducing new efficiency and power-optimised applied sciences that look to problem the dominance of flash storage within the knowledge centre.
The highway to 100TB and past
For these of us who keep in mind when a 1TB drive was an enormous deal, the numbers being mentioned immediately are staggering. WD has confirmed that its 40TB UltraSMR ePMR HDD is already within the fingers of hyperscale prospects for qualification.
This isn’t only a lab experiment; quantity manufacturing is slated for the second half of 2026, proper right here in our yard and globally. The corporate is taking a dual-track strategy to know-how, utilizing each ePMR (energy-assisted perpendicular magnetic recording) and HAMR (heat-assisted magnetic recording).
By 2027, we’ll see HAMR drives ramping up, with a objective to hit 60TB on ePMR and an enormous 100TB on HAMR by 2029.
This technique permits prospects to decide on the know-how that matches their present infrastructure with out forcing a painful transition.
Rethinking arduous drive efficiency
One of many greatest bottlenecks for AI has at all times been the pace of conventional spinning rust in comparison with solid-state drives. WD is seeking to slender that hole with two new architectural breakthroughs: Excessive Bandwidth Drive and Twin Pivot know-how.
Excessive Bandwidth Drive know-how permits the drive to learn and write from a number of heads concurrently, successfully doubling the bandwidth.
WD says this could scale as much as 8x bandwidth positive factors sooner or later, making HDDs viable for workloads that used to require costly flash. The Twin Pivot design provides a second set of independently working actuators, giving a 2x increase to sequential IO efficiency.
In contrast to earlier makes an attempt at dual-actuator drives, this design doesn’t sacrifice capability or require prospects to rewrite their whole software program stack.
When these applied sciences work collectively, WD expects to keep up the IO-per-TB efficiency whilst drive sizes balloon to 100TB.
Energy effectivity for the AI age
AI coaching generates a mountain of “chilly” knowledge that must be saved on-line however isn’t accessed each second.
Conventional drives might be power-hungry, and tape is just too gradual for contemporary AI inference wants. WD’s new power-optimised HDDs use 20% much less energy than normal fashions, particularly designed for these huge knowledge tiers.
By buying and selling off a small quantity of random IO for higher effectivity, these drives assist decrease the whole value of possession for large knowledge centres.
These power-sipping drives are anticipated to succeed in buyer qualification in 2027.
A brand new model for a brand new period
The shift to the “WD” branding isn’t only a beauty change; it displays an organization that’s now laser-focused on storage infrastructure.
With 90% of its income now pushed by AI and cloud sectors, the corporate is positioning itself as a significant associate for the information financial system.
“For the previous 12 months, WD has remained constantly targeted on execution and accelerating innovation, which has enabled us to actually reimagine the arduous drive to fulfill the necessities of AI,”
Irving Tan, Chief Govt Officer, WD.
The corporate’s latest monetary success has seen it land a spot within the Nasdaq 100, proving that there’s nonetheless loads of life within the HDD market.
As AI continues to scale, the necessity for cost-effective, high-capacity storage is barely going to develop.
“WD Innovation Day is the place our customer-centric enterprise transformation meets our breakthrough know-how for the AI period,”
Ahmed Shihab, Chief Product Officer, WD.
Scaling for everybody
WD isn’t simply wanting on the giants like Google and Microsoft; it’s additionally increasing its Platforms enterprise.
A brand new clever software program layer, utilizing an open API, is ready to launch in 2027 to assist mid-scale corporations handle petabyte-scale knowledge. This can deliver hyperscale-level economics to a broader vary of companies, making it simpler to deploy UltraSMR and HAMR tech.
It’s about decreasing the “time-to-value,” making certain that corporations can get their AI tasks up and working with out getting slowed down in {hardware} complexity.
“The market validation is already evident – prospects are deploying these options as a result of WD is fixing what issues most for AI infrastructure: dependable capability at scale, efficiency that meets demanding workloads, and economics that allow profitability,”
Ed Burns, HDD Analysis Director, IDC.
Whereas we don’t have native Australian retail pricing for 40TB drives simply but, these enterprise-grade items sometimes begin within the 1000’s.
For a degree of reference, present high-capacity enterprise drives in Australia retail for roughly A$800 to A$1,200 relying on the particular mannequin and provider.
Anticipate the 40TB items to hold a premium once they land in native distribution later subsequent 12 months.
For extra data, head to https://www.westerndigital.com
Elevate your perspective with NextTech Information, the place innovation meets perception.
Uncover the newest breakthroughs, get unique updates, and join with a worldwide community of future-focused thinkers.
Unlock tomorrow’s tendencies immediately: learn extra, subscribe to our e-newsletter, and grow to be a part of the NextTech group at NextTech-news.com
