New Middleware Technology Quadruples SSD Speed?

This is just some PR wildly exaggerating results from a SSD garbage collection paper.<p>The linked graphs show the big improvement in just one of many workloads and only in the case where the SSD is near-full, and don&#x27;t show what disk&#x2F;ftl they&#x27;re comparing against. But even then these are their own hand picked benchmarks. (presumably the details will be published along with the paper at some point)<p>The trick itself is that they scramble the LBA addresses of blocks with some 1:1 function. But it raises the question why the FTL doesn&#x27;t already do this. Which is a reason to suspect they&#x27;re benchmarking against some el cheapo SSD with a dumb FTL.

Sounds basically like the fundamental research on log structured file systems done back in 1992. <a href="http://www.stanford.edu/~ouster/cgi-bin/papers/lfs.pdf" rel="nofollow">http:&#x2F;&#x2F;www.stanford.edu&#x2F;~ouster&#x2F;cgi-bin&#x2F;papers&#x2F;lfs.pdf</a>

i wonder if this is something similar to USB Supercharger [1]? i personally used this back in the USB 2.0 days to make running Firefox Portable bearable from what was already the fastest 4k random flash drive [2]. The write speed increase was truly astonishing, there was nothing like it on the market. If you look at their CrystalDiskMark screenshots, the 4k random goes up from 0.01 MB&#x2F;s to 10MB&#x2F;s - no joke. On my faster drive the increase was only a mere 100x.<p>[1] <a href="http://easyco.com/usb-sc.htm" rel="nofollow">http:&#x2F;&#x2F;easyco.com&#x2F;usb-sc.htm</a><p>[2] <a href="http://www.amazon.com/Patriot-Xporter-Boost-Flash-PEF8GUSB/dp/B000VE2RP6" rel="nofollow">http:&#x2F;&#x2F;www.amazon.com&#x2F;Patriot-Xporter-Boost-Flash-PEF8GUSB&#x2F;d...</a>

This post is quite vague on the details, but I came across a a more detailed paper recently that also promised big gains:<p><a href="http://www.ece.eng.wayne.edu/~sjiang/pubs/papers/ouyang14-SDF.pdf" rel="nofollow">http:&#x2F;&#x2F;www.ece.eng.wayne.edu&#x2F;~sjiang&#x2F;pubs&#x2F;papers&#x2F;ouyang14-SD...</a><p>SDF: Software-Defined Flash for Web-Scale Internet Storage Systems<p>Working with&#x2F;as Baidu, they are using an FPGA as a custom flash controller that exposes the underlying &quot;channels&quot; more directly.

It is stupendous how much effort we are using to make our nice flash disks look like the old rusty rotary ones instead of (re)writing our software to be aware of the true characteristics of the underlying hardware.

They just increase fragmentation and decrease amount of data written to disk. These are very traditional trade offs especially with database WAL and LSFS. They didn&#x27;t tell if it decreases read speed, because even if SSD &#x27;seek times&#x27; are low, most drives read random blocks from disk much slower than linear data. Which means that fragmenting data especially in very small blocks might cause reduce the drive read speed.

I wonder if this reduces durability as it inhibits the NAND block rewrite balancing or whatever it&#x27;s called

&quot;In a simulation, the research team confirmed that the new technology improves the writing speed of SSD by up to 300% and reduces power consumption by up to 60% and the number of write&#x2F;erase cycles by up to 55%, increasing product life. Because, with the new method, it is not necessary to make any changes to NAND flash memory, and the method is completed within the middleware, it can be applied to existing SSDs as it is.&quot;<p>Promising news! Essentially removing fragmentation on the fly instead of just accepting it as a done deal on SSD

Is there a way to actually access the NAND chips in a better way? Having the ssd firmware guessing stuff seams like quite a waste.

Didn&#x27;t realize NAND flash memory couldn&#x27;t write over itself...ie it is constantly having to use new storage areas

They only got FTL wrong, because it&#x27;s a Faster Than Light engine.