Slashdot Mirror

There are a dozen different memory technologies that "in 10 years time" will revolutionize everything.

It doesn't have to be in 10 years time. But did you expect the rise of the All Flash Array 10 years ago?

Legacy spinning disks will be as dead in 10 years as tape is today.

I test drove one of these for a couple of months: http://www.violin-memory.com/p...

It delivered way more than is advertised here and wasn't connected via PCIe. We're talking 2 GB/s BW and more than 250,000 IOPS with an average response time under 200 microseconds in my testing. It is kind of spendy and heavy as fuck.

Also I have a very large penis.

While they discuss individual SSDs, modern flash storage arrays ( http://www.violin-memory.com/products/6000-flash-memory-array/ ) can hide all the write latency and its effects on read latency. When you start talking about 16TB SSDs the same techniques can be used.

As far as bandwidth and IOPs, they use a 4K/8K write size for MLC/TLC, but MLC already exists with 8K pages, as well as having the ability to write more than one plane at once, which doubles the write bandwidth. Double the page size again and you double the BW.

Now bigger page sizes only help on the reads if you can use more than a single user read worth of data in the page, which might be possible depending on what the system knows about access patterns. But without making assumptions about the ability to store data together that's likely to be read together, garbage collection, which can wide up reading more bytes than the user does, can use most of the data in a page.

So there are factors of 2X, 4X maybe 8X in performance that the paper misses out on.

As far as density, it is not necessary to go to smaller features to get more bits per chip by using 3D techniques such as Toshiba's P-BiCS (Pipe-shaped Bit Cost Scalable) MLC NAND which allow vertical stacking which increases density without using smaller features with their worse performance and lifetime.

The group at UCSD that authored this has done some nice work so I don't mean to be too negative, but they are trying to predict too far from a limited and faulty set of assumptions which unfortunately negates much of the validity of this paper.

jon

p.s. in the interests of full disclosure, I make the arrays in the first link :)

Whee, 1GBps (10Gbps) per direction. How is this significantly better than COTS 8Gbps FC or 10Gb iSCSI/FCoE that don't need a proprietary card? Heck if you want LOTS of bandwidth use 40Gb IB. Plus according to the manufacturers site they DO support the hosts through COTS connections.

From summary:

One reason the flash SAN is so fast is that it doesn't use a SAS or PCIe backbone, but instead has a proprietary interface that offers up 5 to 6Gb/s throughput.

What are they talking about? The violin memory website says the appliances themselves support FC, 10 GbE, and Infiniband connections. Their performance page says that the appliance can be directly connected to a pcie bus, presumably using some sort of pass-through interface card, but what physical connector and media are used?

From summary:

One reason the flash SAN is so fast is that it doesn't use a SAS or PCIe backbone, but instead has a proprietary interface that offers up 5 to 6Gb/s throughput.

What are they talking about? The violin memory website says the appliances themselves support FC, 10 GbE, and Infiniband connections. Their performance page says that the appliance can be directly connected to a pcie bus, presumably using some sort of pass-through interface card, but what physical connector and media are used?

Yup, violin basically sells the same thing - in a yellow box, too.
The problems with these boxes, until recently, have been price and (obviously) persistence.

Since data is never really persisted you'd only buy them in addition to a traditional SAN (or SSD nowadays), not as a replacement.
When you have the dough you can do interesting things with them, though. I know a company that does most of their transaction processing on violins (financial sector, sick throughput) and uses spindles in a write-behind fashion. Technically very interesting but will imho always remain a niche market.

The Violin 1010 connects to a server and provides almost 10 times the capacity of memory per rack and per Watt. A single rack of Violin 1010 can support 10 TByte of DRAM at less than 1W per GByte! The 90% power savings per GByte is important for the green data center.

Read the article much? http://www.violin-memory.com/products/violin1010.html

What about the option of using an (albeit more expensive) (Volatile) DRAM-based SSD for your email servers?
If *someone* subpoenas it, kindly provide it (unplugged) with the any passwords and a full set of encryption keys...
(Assuming there are not already laws prohibiting a corporation from using a faster (700-1400MB/s @ 3s), more reliable (protected with both ECC and RAID), higher I/O preforming (3 million random IOPS), volatile DRAM SSD array for their email storage?)
"Here is my untouched email server storage device all boxed up and sealed as required per your subpoena order..."

504GB of DRAM would make a *nice* email storage device... (Violin 1010) http://www.violin-memory.com/products/violin1010.html

I think the folks at STEC might be a bit surprised by these claims. Especially considering that their product has been shipping for months already.

Of course, they might be quite a bit less expensive than the Zeus SSDs, which are quite pricey...

For some folks, high performance is the requirement and cost is no object. Those folks get their solid-state drives from folks like STEC, Texas Memory Systems, or (soon) Violin Scalable Memory. I'd love to be able to afford this stuff. Buy maybe Samsung will be able to provide a fraction of their performance at a smaller fraction of the cost.

saw something faster at linux world

Keep your eye on Violin http://www.violin-memory.com/products/violin1010.h tml