Previewing Maxiotek's MK8115 SSD Controller: Can DRAM-less Drives Make The Cut?
by Billy Tallis on May 9, 2017 8:00 AM ESTAnandTech Storage Bench - The Destroyer
The Destroyer is an extremely long test replicating the access patterns of very IO-intensive desktop usage. A detailed breakdown can be found in this article. Like real-world usage and unlike our Iometer tests, the drives do get the occasional break that allows for some background garbage collection and flushing caches, but those idle times are limited to 25ms so that it doesn't take all week to run the test.
We quantify performance on this test by reporting the drive's average data throughput, a few data points about its latency, and the total energy used by the drive over the course of the test.
As expected, the MK8115 drives perform relatively poorly overall on The Destroyer. The TLC drive's average data rate is about 20% slower than the next slowest drive in this comparison. The MLC sample actually manages to slightly outperform the earlier JMicron drives, which weren't even handicapped by a lack of external DRAM. However, relative to current MLC drives, the MK8115 MLC drive is about 25% slower overall on this test.
The average service times of the MK8115 drives aren't particularly bad. The TLC drive's average service time is lower than either of the other two SATA drives using the same Micron 3D TLC. The MK8115 sample with MLC is slower than the Samsung 850 PRO and PNY CS2211, but slightly faster than the JMicron drives and the OCZ VX500.
The MK8115 drives do a decent job of keeping latency under control, with the MLC drive in particular having about the same number of outliers above 10ms as the best SATA SSDs. The MK8115 with TLC has some trouble but still has far fewer outliers above 10ms than the ADATA SU800.
However, when looking at the number of extreme latency outliers above 100ms, the MK8115 drives are the worst in the bunch and even the MLC drive scores worse than all the competing TLC solutions.
The power efficiency of the MK8115 drives is decent. Even though these drives take a bit longer to complete the test, the total energy used over the course of the test isn't any higher than most of the competition.
60 Comments
View All Comments
MajGenRelativity - Tuesday, May 9, 2017 - link
That might be a long way awayvladx - Tuesday, May 9, 2017 - link
You're gonna have to wait till mid-2019 for that.CheapSushi - Wednesday, May 10, 2017 - link
You'll get that most likely when QLC drives come out. But consider they'll be for bulk storage rather than general use.MajGenRelativity - Thursday, May 11, 2017 - link
YepHomeworldFound - Tuesday, May 9, 2017 - link
There's definitely some kind of manipulation occurring in the memory industry, it's happening with both DRAM and NAND.FH123 - Tuesday, May 9, 2017 - link
Let's see. JMicron SSD drives stalled (I had one). Two different USB drive enclosures, featuring JMicron chipsets, caused random data corruption for me and my colleague. The DVD / Blueray drives on my desktop randomly fail to show after boot. The chip they're hanging off of? JMicron. Will I consciously buy anything from this company or their offshoots? Nope. Their chips are ubiquitous and hard to avoid, but what trash they are.romrunning - Tuesday, May 9, 2017 - link
Don't forget the infamous JMicron "stutter" problem!vladx - Tuesday, May 9, 2017 - link
I have an ADATA SP600NS34 with JMicron JMF670H and it's a solid drive with 100% life remaining after 3TB written. No slowdowns or drive timeouts either.jabber - Tuesday, May 9, 2017 - link
I'm trying to hold off buying in any SSD drives at the moment. Prices have got silly compared to what I was paying a year ago for essentially a better product.romrunning - Tuesday, May 9, 2017 - link
So true on the pricing woes - I'm hoping there is more MLC capacity coming on soon.