Seagate promises to HAMR us all with spinning rust next year
HAE: HDD Acronym Excess. We have SMR, TDMR, BPM...
Comment Seagate has revealed fresh data on how it will increase areal density as well as a disk's IO rate from next year onwards.
Analyst haus Stifel Nicolaus' MD, Aaron Rakers, attended an analysts' day at which Seagate said it is learning about potential caching techniques to ameliorate the write performance disadvantage of shingled magnetic recording (SMR) disk drives. These have blocks of overlapping write tracks, necessitating the re-writing of an entire block if a section of data in the block of tracks needs re-writing.
Layers of caching could potentially be used to effectively shorten the overall write time.
On the overall IO performance side, disks have a single read/write head per platter and this single channel has to be a pipe for more and more data as the disk's capacity increases. According to Rakers, Seagate says two-dimensional magnetic recording (TDMR), which involves two or more readers per device, will be a feature of drives shipped in the future.
It has previously stated TDMR drives would become generally available next year.
We envisage data being organised on the platters so that parts of a large file could read in parallel, thus speeding sequential access. Alternatively, the separate heads could operate independently and read different files or records simultaneously, making queued random reads faster as they could operate in parallel.
Rakers said Seagate expects HAMR (heat-assisted magnetic recording) drives to become available towards the "the latter end of the prior 2016-2018 expectation." This could be taken to mean the second half of 2018. We have learnt from a Seagate statement that the first HAMR drives will be of comparable areal density to the then-contemperaneous existing PMR drives.
The prospect, though, is of increasing areal density; otherwise there is no point in HAMR developments continuing.
We're told Bit-Patterned Media (BPM) technology testing is ongoing but BPM drives shouldn't be expected until after 2018. In the near term, Seagate says we should expect volume production of 1.2Tbit/in2, meaning 2.5-inch drives with 1TB platters, over coming quarters. Beyond that Seagate says its research and development labs have increased areal density by 30 per cent this year, implying 1.56Tbit/in2, 1.3TB platters, and 2.6TB 2.5-inch drives. We should expect new products at this density in 2016.
We note Seagate isn't saying that its 2.5-inch drives are shingled. We have also learnt that Seagate has helium-filled drive technology in development and this, with its consequential addition of an extra platter to drives, if HGST's experience is replicated, should mean an increase in drive capacity above and beyond areal density improvements. However, Seagate has not said when it expects helium-filled drives to ship, so we have no timescale on that.
Our expectation is that this could be a 2018 or later development but we might have a nice surprise and see them earlier. ®