News for hackers. Stuff that matters.
Re:SSDs: A gift and a curse by justinsb on Tuesday June 02, 2015 @12:54PM (#9647152)
I am looking forward to the day when all SSDs ship minimal firmware, and offload all the complex work to (main-CPU) software.
Re:SSDs: A gift and a curse by wtallis on Tuesday June 02, 2015 @01:21PM (#9647381)
Having the flash translation layer on the drive itself is the only practical way for the drive to be usable by multiple operating systems. Almost everyone who uses a SSD wants it to be accessible to at least two operating systems (UEFI, and whatever OS resides on the SSD). Doing the FTL on the host system would basically relegate the SSD to being just a cache device.
Re:SSDs: A gift and a curse by kabdib on Wednesday June 03, 2015 @12:18AM (#9650558)
Yup. I looked at this while doing design work for a certain highly visible gaming product. You want the responsibility for block leveling and transactional goop close to the device, where the firmware can know stuff about device geometry and other nasty bits that affect reliability.

Putting that stuff in the OS puts you back to, oh, the same wicked stuff that people had to deal with for MFM hard drives (remember cylinder, head and sector counts? Only probably ten times more complicated).

A block level abstraction is fine, as long as the abstraction can make some transactional guarantees.

Re:SSDs: A gift and a curse by zokier on Tuesday June 02, 2015 @01:33PM (#9647487)
We have these things called standards which allow nice things like interoperability. Besides, you do not need the entire drive to be accessible from UEFI, it is enough that you can load a small bootloader which contains the drivers to handle the disk. Thirdly, UEFI has extensible right in the name, shouldn't be too hard for OSes to throw a disk driver blob into UEFI.
Re:SSDs: A gift and a curse by bduerst on Tuesday June 02, 2015 @03:48PM (#9648468)
Doesn't that require partitioning the disk? That doesn't seem like an OOB solution.
Re:SSDs: A gift and a curse by wtallis on Tuesday June 02, 2015 @04:10PM (#9648596)
It's not the requirement to partition the drive that kills the idea. Setting aside the first small chunk of the drive for the firmware to read is how it's always been done. The problem is that this would require the partitioning to be done below the wear leveling layer, which reduces the effectiveness of the wear leveling slightly and means you can only change your bootloader settings a thousand times before the drive is dead.
Re:SSDs: A gift and a curse by Dylan16807 on Wednesday June 03, 2015 @01:59AM (#9650808)
That's not really what kills it. You can use a small ring buffer to boost that to 100 thousand, and/or use the flash in a more reliable way there (like SLC or devoting 2/3 of the bits to ECC).
Re:SSDs: A gift and a curse by pgeorgi on Tuesday June 02, 2015 @04:52PM (#9648881)
sadly with NAND even reads contribute to wear.
Re:SSDs: A gift and a curse by VLM on Tuesday June 02, 2015 @01:37PM (#9647520)
The user experience of win-modems, win-printers, and win-scanners was absolutely horrific so you really need to click edit and add the sarcasm tag.
Re:SSDs: A gift and a curse by AYBABTME on Tuesday June 02, 2015 @01:50PM (#9647636)
Wouldn't that make performance a lot worse, for both CPU and disks? Offloading the job to something closer to the flash chips and fully dedicated to servicing them sounds like a better (and the current) path. Same for high speed network links.
Re:SSDs: A gift and a curse by bbulkow on Tuesday June 02, 2015 @06:29PM (#9649462)
FusionIO cards run with FTL maintenance on the main CPU. Run a workload and you'll see the driver spinning in a way that isn't accounted for by your operations. It's not pretty - at high workloads, the network interrupt traffic starts interfering with all the bus traffic to the cards. A controller can be designed to run exactly at the speed of the chips, no more no less, leaving the CPU doing what it does best.

That being said, if you build your system correctly, there is not any extra CPU load created by maintaining the FTL at app level. The main problem is the lack of an API for wear detection. I have discussed this with senior chip designers at the major manufacturers, and they are all loath to nail down to one api, as they believe how wear is specified will have to change over time. That's a broad brush - some manufacturers are more open to these discussions than others.

I'm still working on them.

Re:SSDs: A gift and a curse by rwmj on Tuesday June 02, 2015 @02:14PM (#9647812)
Re:SSDs: A gift and a curse by AYBABTME on Tuesday June 02, 2015 @02:34PM (#9647973)
Always have that in the back of my mind =]. However I don't think that's the current trend right now. Except where special hardware is integrated to the chip, but the work isn't left to the CPU itself.

I'm not so versed in the area so I'm just blind guessing here.

Re:SSDs: A gift and a curse by wtallis on Tuesday June 02, 2015 @02:21PM (#9647878)
Flash translation layers don't really need to be offloaded for the sake of performance. You need some kind of controller to multiplex access to a dozen or more flash chips from a narrow PCIe link, so you might as well throw in an ARM core to do the translation and wear leveling and garbage collection, and then you can stick a few supercaps in the drive too and have a robust hardware platform.
Re:SSDs: A gift and a curse by wmf on Tuesday June 02, 2015 @02:44PM (#9648047)
The FTL runs much faster on a 3GHz Xeon core than on a 150MHz ARM core. Of course, the cost and power consumption is also much higher.
Re:SSDs: A gift and a curse by rwmj on Tuesday June 02, 2015 @12:57PM (#9647176)
Unfortunately that day isn't going to happen. In fact the opposite is happening -- interesting spinning rust technology like Shingled Magnetic Recording (SMR) drives are getting more complex translation layers.
Re:SSDs: A gift and a curse by Turing_Machine on Tuesday June 02, 2015 @02:01PM (#9647710)
I wouldn't say that it's never going to happen based on current trends.

This kind of thing tends to be quasi-cyclic.

At one time, desktop computers had separate FPUs. Over time, those got integrated onto the main CPU, but then the trend switched back, with heavy calculations moving off-CPU to the GPU.

The mainframe->desktop->cloud story is similar.

I say "quasi"-cyclic because (e.g.) cloud isn't really the same as mainframe, but the transition back and forth between centralized and decentralized is still striking nonetheless.

Re:SSDs: A gift and a curse by marssaxman on Tuesday June 02, 2015 @04:32PM (#9648744)
I'm eagerly awaiting the part of the cycle where the GPU gets folded back into the processor. I'm really not interested in writing code against mystery-meat architectures with no documented ABI...
Re:SSDs: A gift and a curse by gcr on Tuesday June 02, 2015 @08:28PM (#9649914)
We're already seeing that with integrated, "soldered-on" graphics cards, and with frameworks like OpenCL, and higher-level abstractions like Theano, Torch7, etc.
Re:SSDs: A gift and a curse by marssaxman on Wednesday June 03, 2015 @12:02PM (#9653512)
All we need now is documentation for the low-level abstractions so we can scrap all the vendor code and compile our own GPU machine language.
Re:SSDs: A gift and a curse by greggyb on Wednesday June 03, 2015 @02:23PM (#9654727)
I am not sure based on reading your post whether or not I am providing new information to you:

Keep an eye on the AMD APUs moving forward and their HSA progress.

Re:SSDs: A gift and a curse by marssaxman on Wednesday June 03, 2015 @04:07PM (#9655484)
Thanks, that sounds like an interesting development.
Re:SSDs: A gift and a curse by nine_k on Tuesday June 02, 2015 @01:03PM (#9647229)
SSDs that use SATA aren't going to allow any finer control, I suspect, due to the interface's limitations.

PCIe SSD devices, though, already feature an advanced controller (that is, a CPU core). This core could be programmed from the "computer" side somehow — but building any public API is always a major commitment, and takes much more effort than a proprietary API which you can change at will.

Re:SSDs: A gift and a curse by ebbv on Tuesday June 02, 2015 @01:40PM (#9647548)
This is a bad idea for a lot of reasons. A well designed system has components that just work and have standardized interfaces. What you're suggesting is that each device should be totally unusable unless the machine has the specific software running that knows how to use it. That's poor design.
Re:SSDs: A gift and a curse by zanny on Tuesday June 02, 2015 @05:17PM (#9649073)
Thats how... pretty much every peripheral in existance works. You need webcam drivers. Video card drivers. Sound card drivers. USB drivers, at the least for generic HID devices. Nothing stops there from being common standard generic SSD drivers, where vendors could publish their own if they had an optimization.

Right now the SATA interface prevents anything like that from happening, and the market has already moved to these proprietary broken controllers with braindead GL-class protocols to access them. (SATA, NVMe)

Re:SSDs: A gift and a curse by wtallis on Tuesday June 02, 2015 @09:55PM (#9650148)
Any such interface for SSDs would need to be able to work with a dozen different kinds of flash, a few sizes for each, a wide range of quantities of flash chips, all while exposing the performance and endurance characteristics of the flash and the topology of its connection to the host bus so that the software FTL can properly tune itself.

You're asking for a hardware interface of unprecedented complexity and cross-platform vendor neutral drivers of comparable complexity to 3d graphics drivers, all for the sake of letting SSDs have "dumb" controllers for a generation or two before the technology changes enough that the controllers need to start translating things again.

Re:SSDs: A gift and a curse by nqzero on Tuesday June 02, 2015 @04:29PM (#9648723)
this would be a huge win for database design - FTL is incredibly wasteful in terms of unnecessary block erases for cache
Re:SSDs: A gift and a curse by sliverstorm on Tuesday June 02, 2015 @05:20PM (#9649092)
Welcome back to driver hell. We thought you weren't ever coming back, but here you are!
Re:SSDs: A gift and a curse by moonbug on Tuesday June 02, 2015 @05:55PM (#9649289)
If you think that, I've a crate of ST-506s to sell you.
Re:SSDs: A gift and a curse by Retric on Tuesday June 02, 2015 @01:19PM (#9647358)
In the long term latency makes that a non starter.