Tuesday, February 3rd 2026
Western Digital Designs High-Bandwidth HDDs That Quadruple I/O Speeds
Western Digital has today presented its latest effort to catch up with traditional QLC NAND Flash SSDs by improving its HDD offerings. With the latest High-Bandwidth HDDs, Western Digital has implemented two new technologies in a classical multi-platter HDD design. The first innovation comes in a form of High Bandwidth Drive Technology, which enables double the I/O bandwidth with a path to 8x the current bandwidth in the future. It relies on simultaneous reading and writing from multiple heads on multiple tracks, which is already in customer hands for validation. The second one is Dual Pivot Technology that introduces a second set of independent actuators on a separate pivot, which will not scarify drive capacity unlike older dual actuator designs.
Using Dual Pivot Technology, HDDs can pack more drive platters in a standard 3.5-inch body for higher capacities, and the performance grows by an additional 2x, which is 4x I/O bandwidth compared to today's drives. This technology will pave the way for 100 TB HDDs, that offer speeds comparable to QLC-based SATA III SSDs, at much better price/performance ratio and better data retention, prompting the massive boom of HDD development. Western Digital's drives with High Bandwidth Drive Technology are already shipping to customers, while drives with Dual Pivot Technology are in development in Western Digital's labs and are scheduled to become available in 2028, with early customer sampling probably much sooner.However, this is not where Western Digital is stopping. The company is also developing new Power-Optimized HDDs, which are designed to help customers access cold data in seconds, sitting right between traditional tape and regular HDDs. These Power-Optimized HDDs will trade some of the random I/O speeds for higher capacity, while operating at 20% lower power. For example, AI data centers hold massive amounts of data in a cold storage, but when the data is needed, the systems need it in seconds, not hours as seen with the traditional tape. So for this purpose, Western Digital made these HDDs that are trading off some performance for longevity, speed, and capacity, all while maintaining sub-second access time that the traditional cold storage tiers like tape can't execute. These drives will be in customer's hands for qualification in 2027.
Using Dual Pivot Technology, HDDs can pack more drive platters in a standard 3.5-inch body for higher capacities, and the performance grows by an additional 2x, which is 4x I/O bandwidth compared to today's drives. This technology will pave the way for 100 TB HDDs, that offer speeds comparable to QLC-based SATA III SSDs, at much better price/performance ratio and better data retention, prompting the massive boom of HDD development. Western Digital's drives with High Bandwidth Drive Technology are already shipping to customers, while drives with Dual Pivot Technology are in development in Western Digital's labs and are scheduled to become available in 2028, with early customer sampling probably much sooner.However, this is not where Western Digital is stopping. The company is also developing new Power-Optimized HDDs, which are designed to help customers access cold data in seconds, sitting right between traditional tape and regular HDDs. These Power-Optimized HDDs will trade some of the random I/O speeds for higher capacity, while operating at 20% lower power. For example, AI data centers hold massive amounts of data in a cold storage, but when the data is needed, the systems need it in seconds, not hours as seen with the traditional tape. So for this purpose, Western Digital made these HDDs that are trading off some performance for longevity, speed, and capacity, all while maintaining sub-second access time that the traditional cold storage tiers like tape can't execute. These drives will be in customer's hands for qualification in 2027.



27 Comments on Western Digital Designs High-Bandwidth HDDs That Quadruple I/O Speeds
Regardless, it would be really sweet if HDDs could finally reach so high sequential read/write speeds. I mainly use them for cold backups, imagine being able to write them out in 1-2 hours instead of nearly a day.
I wonder how well those drives would do today with all the advances made with NAND since 2015. Of course the mechanical side of it as well as the sata interface will still remain a bottleneck. But with the newer motor here to bump up mechanical reads and writes? Maybe SSHDs might come back again.
Since you mentioned 2015 in one context. I'll also suggest the other, WD Ae cold storage.
Before or after the QLC drives exhaust their cache? 'cause as far as I know QLC drives are abysmal when they run out of cache, like sub 100MB/s.
I still think they'd be good for those but the SSD cache would have to be increased in size and speed to make them worth it.
Twice the speed would be very nice. 4 times even better.
In my eyes it was an experiment at best. Not a mass produced product to be iterated on.
And i was right. There have been no successors with higher capacities and the initial models were EOL'd and are nigh impossible to find now.Is there a problem with reliability with existing drives?
I cant think of anyone asking for mystical "improvements in reliability", but i see plenty of people complaining about the slowness of HDD's.
Aside from certain Seagate models having higher AFR i cant think of any high profile cases on mass reliability issues.
Even first gen Helium filled drives have held up far longer than anyone predicted given the technology was uncertain at the time.
The reason we need higher transfer speeds is precisely because speeds haven't kept up with capacity increases. Capacity has more than doubled, but speeds have remained the same. This affects several aspects of using the newest models. Many of them mentioned here.
It's like getting a bigger water tank, but still using the same small dinky water hose with low throughput.Indeed. We already have a faster standard that could be used without a problem on consumer boards.
I suppose thus far there hasn't been great incentive to use it since drives were not saturating SATA-III anyway.