The Intel Optane Memory H10 Review: QLC and Optane In One SSD
by Billy Tallis on April 22, 2019 11:50 AM ESTRandom Read Performance
Our first test of random read performance uses very short bursts of operations issued one at a time with no queuing. The drives are given enough idle time between bursts to yield an overall duty cycle of 20%, so thermal throttling is impossible. Each burst consists of a total of 32MB of 4kB random reads, from a 16GB span of the disk. The total data read is 1GB.
The burst random read test easily fits within the Optane cache on the Optane Memory H10, so it outperforms all of the flash-based SSDs, but is substantially slower than the pure Optane storage devices.
Our sustained random read performance is similar to the random read test from our 2015 test suite: queue depths from 1 to 32 are tested, and the average performance and power efficiency across QD1, QD2 and QD4 are reported as the primary scores. Each queue depth is tested for one minute or 32GB of data transferred, whichever is shorter. After each queue depth is tested, the drive is given up to one minute to cool off so that the higher queue depths are unlikely to be affected by accumulated heat build-up. The individual read operations are again 4kB, and cover a 64GB span of the drive.
On the longer random read test that covers a wider span of the disk than the Optane cache can manage, the H10's performance is on par with the TLC-based SSDs.
 |
|||||||||
The Optane cache provides little benefit over pure QLC storage at lower queue depths, but at the higher queue depths the H10 with caching enabled starts to develop a real lead over the QLC portion on its own. Unfortunately, but the time queue depths are this high, the flash-based SSDs have all surpassed the H10's random read throughput.
Random Write Performance
Our test of random write burst performance is structured similarly to the random read burst test, but each burst is only 4MB and the total test length is 128MB. The 4kB random write operations are distributed over a 16GB span of the drive, and the operations are issued one at a time with no queuing.
The burst random write performance of the H10 with caching enabled is better than either half of the drive can manage on its own, but far less than the sum of its parts. A good SLC write cache on a TLC drive is still better than the Optane caching on top of QLC.
As with the sustained random read test, our sustained 4kB random write test runs for up to one minute or 32GB per queue depth, covering a 64GB span of the drive and giving the drive up to 1 minute of idle time between queue depths to allow for write caches to be flushed and for the drive to cool down.
On the longer random write test that covers a much wider span than the Optane cache can handle, the Optane Memory H10 falls behind all of the flash-based competition. The caching software ends up creating more work that drags performance down far below what the QLC portion can manage with just its SLC cache.
 |
|||||||||
Random write performance on the Optane Memory H10 is unsteady but generally trending downward as the test progresses. Two layers of caching getting in each others way is not a good recipe for consistent sustained performance.
Facebook
Twitter
RSS
60 Comments
View All Comments
The_Assimilator - Tuesday, April 23, 2019 - link
> I don't understand the purpose of this product.It's Intel still trying, and still failing, to make Optane relevant in the consumer space.
tacitust - Tuesday, April 23, 2019 - link
It works in the sense that the OEMs who use this drive will be able to use the fact that customers will be getting cutting edge Optane storage. As the review says, this is a low effort solution, so it likely didn't cost much to develop, so they won't need too many design wins to recoup their costs. It also gets Optane into many more consumer devices, which helps in the long run in terms of perception, if nothing else.Note: most users won't know or even care that the drive itself doesn't provide faster performance than other solutions, so it doesn't really matter to Intel either. If they get the design win, Optane does gain relevance in the consumer space, just not with the small segment of power users who read AnandTech for the reviews.
ironargonaut - Monday, April 29, 2019 - link
Seems it does provide faster performance in some usage cases.https://www.pcworld.com/article/3389742/intel-opta...
CheapSushi - Wednesday, April 24, 2019 - link
I can't stand these dumb posts where people shut down the usage for consumers. I use it all the time for OS and other programs/files. I use it as cache. I use it for different reasons. Even the cheap early x2 laned variants. I'm not in IT or anything enterprise.name99 - Thursday, April 25, 2019 - link
It's worse than that.The OPTANE team clearly want to sell as many Optanes as they can.
But INTC management has decided that they can extract maximal money from enterprise by limiting
name99 - Thursday, April 25, 2019 - link
It's worse than that.The OPTANE team clearly want to sell as many Optanes as they can.
But INTC management has decided that they can extract maximal money from enterprise by limiting the actually sensible Optane uses (in the memory system, either as persistent memory ---for enterprise, or as a good place to swap to, for consumers).
And so we have this ridiculous situation where the Optane team keeps trying to sell Optane in ways that make ZERO sense because the way that makes by far the most sense (sell a 16 or 32 GB or 64GB DIMM that acts as the swap space) is prevented by Intel high management (who presumably are scared that if cheap CPUs can talk to Optane DIMMs, then someone somewhere will figure out how to use them in bulk rather than super expensive special Xeons).
Corporate dysfunction at its finest...
Billy Tallis - Friday, April 26, 2019 - link
I think it's too soon to say that Intel's artificially holding back Optane DIMMs from market segments where they might have a chance. They had initially planned to have Optane DIMM support in Skylake-SP but couldn't get it working until Cascade Lake, which has only been shipping in volume for a few months. Now that they have got one working Optane-compatible memory controller out the door, they can consider bringing those memory controller features down to other product segments. But we've seen that they have given up on updating the memory controllers on their 14nm consumer parts even to provide LPDDR4 support, which certainly is a more compelling and widely-demanded feature than Optane support. I wouldn't expect Intel to be able to introduce Optane support to their consumer CPUs until their second generation of 10nm (not counting CNL) processors at the earliest. Trying to squeeze it into their first mass-market 10nm would be unreasonable since they should be trying at all costs to avoid feature creep on those parts and just ship something that works and isn't still Skylake.ironargonaut - Monday, April 29, 2019 - link
Read here for an actual real world usage test. Two system with only memory difference and same input sometimes significantly different results.https://www.pcworld.com/article/3389742/intel-opta...
3X speed up for some tasks. I don't know about ya'll but I multitask a lot at work so I will let background stuff go while I do something else that is in front of me.
weevilone - Monday, April 22, 2019 - link
That's too bad. I tried to tinker with the Optane caching when it launched and it was a software disaster. I wrote it off to early days stuff and put it in my kids' PC when they began to allow non-boot drives to be cached. It was another disaster and Intel's techs couldn't figure it out.I wound up re-installing Windows the first time and I had to redo the kids' game drive the second time. No thanks.
CheapSushi - Wednesday, April 24, 2019 - link
The problem is you were using the proprietary HDD caching they marketed. There are so many ways to do drive caching on Windows that doesn't involve that Intel software. It's way better and smoother. even if still software. Software RAID and cache is superior to hardware cache unless you're using $1K+ add-on cards.