Physical vs effective write latency

By: Linus Torvalds (torvalds.delete@this.linux-foundation.org), September 11, 2008 7:16 pm
Room: Moderated Discussions
Doug Siebert (foo@bar.bar) on 9/11/08 wrote:
>
>Regular RAM and a small (pre-erased) reserved section of
>flash along with a capacitor would be a much better
>solution.

If you have pre-erased flash, why bother with the whole
charade to begin with?

Guys, this is basic queuing theory. You cannot get write
bandwidth higher than what your flash is able to absorb
(and that includes all the GC and erase cycles necessary)
in the steady state anyway.

No amount of buffering will ever change that.

Anybody who thinks that they can lower latency by just
adding buffers is a moron, because the fact is,
if you "lie" and tell the OS that the write has completed,
it will just write more. Until your buffer is full, and
you have to throttle the writes!

And at that point, you have to expose the real latency
(or worse - much bigger latencies because you then
end up waiting for all of the buffering you did!)

Buffering can only help temporary spikes.

And there is almost certainly no reason for those
temporary spikes, except for simplistic initial SSD
controller implementations.

Once you are clever and careful enough to try to do any
capacitor that writes back after power failure, you're
already spending more effort and engineering (on a hacky
workaround) than if you were just doing incremental GC and
compaction in the first place.

Because once you do incremental compaction and GC, you'll
not need that special reserved pre-erased part of
the flash - because you just make sure that your compaction
and GC stays just ahead of the curve.

That's what realtime GC means for chissake!

So you're all trying to solve the wrong problem. You're
looking at badly designed SSD's (in fairness, it always
takes a few generations for people to learn even the basic
"tricks", so "badly designed" is a bit harsh), and then
you are trying to make these insane workarounds for it,
when the right way to solve it is to just do it better
in the first place, and suddenly the workarounds aren't
needed any more.

See what I'm saying?

Let me put this another way, if it's unclear:

- you can never fundamentally lower the average latency
for one operation to below that of the average
throughput for that size operation of the device.

Think about it.

If you could make the write latency lower than the
throughput, the thing that feeds it would just feed it
more, and if you can absorb that, then your throughput
would be higher than you just started out claiming it
was.

- Ergo: assuming latencies are fairly stable (ie the
latencies aren't sometimes low and sometimes very high),
that means that you can never realistically win
by buffering more than the data you can write in one
such "latency time".

- Thinking that you can just make buffers bigger is silly.
The fundamental buffer size is going to be "throughput
times latency time". No more, no less. (Of course, for
implementation reasons you might want to do things like
switch buffers around so you'd do double buffering, but
we're talking about just small multipliers here).

- And people know how to do realtime GC and
compaction.

Another way of saying the same thing: in the absense of
things like mechanical arm movement etc that introduces
big latency jumps, it sure as h*ll should be entirely
possible to keep latencies stable. You don't need
to have events that suddenly rewrite big portions of the
disk in order to do GC and erase-cycles!

- once your buffers are in the ballpark of the above
"basic size", then trying to return success early is
not going to help anything anyway. If you lie and return
early (and have a complex battery back-up to protect you
from the downsides of lying), you're just going to be
stuck ont he next IO instead.

So having some kind of super-capacitor is fundamentally
stupid. The "claim the IO is done early" is just stupid.
Much better to not lie, and thus not need battery
backup.

So why do people do high-performance controllers that
really do do battery backup etc? I just told you
it was stupid.

A couple of reasons:

- you can absorb load spikes. The big buffers
won't help you under heavy load, but it will help
you under spiky load.

- for rotational media, you can absorb the much longer
latencies for actual seek events. This is especially
true for RAID controllers that have huge bandwidth
over multiple disks (so the "latency times throughput"
number is actually reasonably large).

- again, mainly for rotational media: you can sort the
final requests and actually improve throughput by doing
the IO in a different order. IOW, you can actually
decrease the latencies by seeking less, but you need
lots of data to do that well.

But apart from the first one, those reasons are largely not
valid for solid-state media. The first one is true, but
quite frankly, it's much better done somewhere else
than on the disk itself. IOW, you're better off doing it
on the host controller, where the fundamental latencies are
lower, so you get a bigger win (ie for bursty IO, you may
be getting a fake latency to the host controller, not the
one all the way to the disk).

Doing it on the host controller is better for another
reason too: it allows you basic building block (the disk)
to be generic and cheap. Which is what you want. Because
people are going to buy cheap, and the people who want
something extra and are willing to pay for it, are still
better off if they can use mass-produced basic building
blocks and then just adding some "secret sauce".

(That "standard building block" detail is true for the
other cases too, for that matter, and probably explains
why nobody I know of actually does a disk with battery
back-up).

To recap:

- doing large buffers is stupid, and indicates that you
have way more latency variation than you should have
in a solid-state setup in the first place.

- and doing them on the disk, rather than closer to the
host, is doubly stupid even if you wanted to go for
an extra kicker.

Hmm? I may be wrong, but I don't think I am.

So please stop thinking that large RAM buffers "speeds
things up". Because they do nothing of the kind. Yes, they
can smooth out any non-realtime issues you have in the
GC/compaction, but you really should see it as a
"smoothing" thing, not anything else.

And there really should be no reason why SSD's should act
particularly bursty in themselves over much bigger data
sets than the size of the erase block. So you should be
able to size the buffer by roughly doing the max throughput
over an average latency cycle, and adding one erase block
worth of data to that.

With a 250MB/s thing that has 0.1 microsecond average
latencies, you shouldn't need more than a 25kB buffer. Of
course, since the erase block size is probably bigger than
that, do say 64kB or 128kB.

And then, if you support multiple outstanding commands, and
since RAM is cheap, you might decide to have separate
buffers for each outstanding command. But at that point,
you're really just wasting buffer space in order to just
not have to manage it very carefully and worry about
completion order etc.

Of course, that all depends on how smooth you can make
your GC and compaction. Maybe you can't do it entirely
incrementally. Maybe you end up having batches of a few
erase blocks.. Who knows? The point is, those latencies on
the order of a second really are the fundamental problem,
and they shouldn't exist, and buffering is not the way
to solve it.

Linus
< Previous Post in ThreadNext Post in Thread >
TopicPosted ByDate
First Dunnington benchmark resultsMichael S2008/08/19 10:54 AM
  First Dunnington benchmark resultsrwessel2008/08/19 01:42 PM
    First Dunnington benchmark resultsAaron Apink2008/08/19 05:49 PM
      First Dunnington benchmark resultsJoe Chang2008/08/19 06:28 PM
        First Dunnington benchmark resultsrwessel2008/08/21 09:49 AM
          First Dunnington benchmark resultsJoe Chang2008/08/21 03:10 PM
            First Dunnington benchmark resultsrwessel2008/08/21 06:42 PM
              First Dunnington benchmark resultsJoe Chang2008/08/21 07:12 PM
      First Dunnington benchmark resultsrwessel2008/08/21 09:45 AM
        First Dunnington benchmark resultsAaron Spink2008/08/21 01:12 PM
          First Dunnington benchmark resultsJoe Chang2008/08/21 03:15 PM
    First Dunnington benchmark resultsRichard Cownie2008/08/20 02:59 AM
      First Dunnington benchmark resultsAnders Jensen2008/08/20 03:26 AM
        +SSDAnders Jensen2008/08/20 03:30 AM
        First Dunnington benchmark resultsRichard Cownie2008/08/20 11:04 AM
          First Dunnington benchmark resultsslacker2008/08/20 12:35 PM
            First Dunnington benchmark resultsDoug Siebert2008/08/20 07:54 PM
              First Dunnington benchmark resultsRichard Cownie2008/08/20 08:58 PM
                SLC vs. MLCDavid Kanter2008/08/21 01:16 AM
                  SLC vs. MLCMatt Sayler2008/08/21 06:25 AM
                  SLC vs. MLCRichard Cownie2008/08/21 06:32 AM
                    SLC vs. MLCLinus Torvalds2008/08/21 08:39 AM
                      SLC vs. MLCMichael S2008/08/21 09:07 AM
                        SLC vs. MLCLinus Torvalds2008/08/21 09:52 AM
                          SLC vs. MLCMichael S2008/08/21 10:35 AM
                            OLTP appliance = mainframe? (NT)Potatoswatter2008/08/21 11:44 AM
                              OLTP appliance = HP NonStop?Michael S2008/08/21 12:03 PM
                                OLTP applianceJoe Chang2008/08/21 03:33 PM
                                  OLTP appliancePotatoswatter2008/08/21 03:59 PM
                            SLC vs. MLCAaron Spink2008/08/21 01:29 PM
                          SLC vs. MLCDan Downs2008/08/21 11:33 AM
                            SLC vs. MLCrwessel2008/08/21 12:45 PM
                              SLC vs. MLCDan Downs2008/08/22 08:21 AM
                            SLC vs. MLCAaron Spink2008/08/21 01:34 PM
                          SLC vs. MLC vs DRAMpgerassi2008/08/21 12:24 PM
                            SLC vs. MLC vs DRAMDavid Kanter2008/08/22 01:31 AM
                          SLC vs. MLCGroo2008/08/23 12:52 PM
                      SLC vs. MLCDoug Siebert2008/08/21 06:14 PM
                        SLC vs. MLCLinus Torvalds2008/08/22 08:05 AM
                          SLC vs. MLCDoug Siebert2008/08/22 02:27 PM
                            SLC vs. MLCEduardoS2008/08/22 06:26 PM
                            SSD Controller differentiationDavid Kanter2008/08/22 09:35 PM
                              SSD Controller differentiationDoug Siebert2008/08/22 10:34 PM
                                SSD Controller differentiation (supercaps, cost...)anon2008/08/23 10:18 AM
                                  SSD Controller differentiation (supercaps, cost...)Doug Siebert2008/08/23 10:40 AM
                            SLC vs. MLCLinus Torvalds2008/08/23 10:50 AM
                              SLC vs. MLCLinus Torvalds2008/09/08 12:03 PM
                                SLC vs. MLCMax2008/09/08 01:51 PM
                                  SLC vs. MLCHoward Chu2008/09/08 09:04 PM
                                    SLC vs. MLCMax2008/09/08 10:29 PM
                                      SLC vs. MLCHoward Chu2008/09/09 12:12 AM
                                        RAM vs SSD?Jouni Osmala2008/09/09 01:06 AM
                                          RAM vs SSD?Max2008/09/12 12:51 PM
                                            RAM vs SSD?EduardoS2008/09/12 04:27 PM
                                            Disk cache snapshottingMax2008/09/13 08:34 AM
                                              Disk cache snapshottingHoward Chu2008/09/14 09:58 PM
                                                Disk cache snapshottingMax2008/09/15 12:50 PM
                                        SLC vs. MLCLinus Torvalds2008/09/09 07:43 AM
                                          SLC vs. MLCHoward Chu2008/09/09 09:42 AM
                                            SLC vs. MLCLinus Torvalds2008/09/09 10:39 AM
                                              SLC vs. MLCMichael S2008/09/10 12:29 AM
                                                SLC vs. MLCanon2008/09/10 02:51 AM
                                                  SLC vs. MLCMichael S2008/09/10 03:09 AM
                                                    SLC vs. MLCMax2008/09/10 04:48 AM
                                                      SLC vs. MLCMichael S2008/09/10 05:52 AM
                                                        SLC vs. MLCMax2008/09/10 06:28 AM
                                                    SLC vs. MLCMatt Sayler2008/09/10 06:21 AM
                                                      SLC vs. MLCMichael S2008/09/10 09:17 AM
                                                    SLC vs. MLCanon2008/09/10 06:29 AM
                                                      SLC vs. MLCMichael S2008/09/10 09:23 AM
                                                        SLC vs. MLCMatt Sayler2008/09/10 10:45 AM
                                                SLC vs. MLCLinus Torvalds2008/09/10 07:25 AM
                                                  SLC vs. MLCMichael S2008/09/10 09:54 AM
                                                    SLC vs. MLCLinus Torvalds2008/09/10 10:31 AM
                                                      Physical vs effective write latencyMax2008/09/11 07:35 AM
                                                        Physical vs effective write latencyLinus Torvalds2008/09/11 09:06 AM
                                                          Physical vs effective write latencyLinus Torvalds2008/09/11 09:48 AM
                                                            Physical vs effective write latencyLinus Torvalds2008/09/11 11:39 AM
                                                              Physical vs effective write latencyMark Roulo2008/09/11 12:18 PM
                                                                Physical vs effective write latencyDoug Siebert2008/09/11 05:59 PM
                                                                  Physical vs effective write latencyLinus Torvalds2008/09/11 07:16 PM
                                                                    Physical vs effective write latencyDoug Siebert2008/09/11 10:28 PM
                                                        Physical vs effective write latencyMS2009/02/03 03:06 PM
                                                  SLC vs. MLC - the trick to latencyAnonymous2008/09/11 12:39 PM
                                                    SLC vs. MLC - the trick to latencyanon2008/09/11 01:17 PM
                                                      SLC vs. MLC - the trick to latencyAnonymous2008/09/11 05:25 PM
                                                        SLC vs. MLC - the trick to latencyDoug Siebert2008/09/11 05:47 PM
                                                          SLC vs. MLC - the trick to latencyrwessel2008/09/11 06:01 PM
                                                        SLC vs. MLC - the trick to latencyanon2008/09/12 12:00 AM
                                                          SLC vs. MLC - the trick to latencyAnonymous2008/09/12 08:52 PM
                                                            SLC vs. MLC - the trick to latencyanon2008/09/13 10:06 AM
                                                            SLC vs. MLC - the trick to latencyUngo2008/09/15 12:18 PM
                                                              To SSD or not? One lady's perspectiveDavid Kanter2008/09/22 01:12 AM
                                                                To SSD or not? One lady's perspectiveHoward Chu2008/09/22 04:02 AM
                                                                  To SSD or not? Real data..Linus Torvalds2008/09/22 07:33 AM
                                                                    To SSD or not? Real data..Ungo2008/09/22 12:27 PM
                                                                      4K sectorsWes Felter2008/09/22 06:03 PM
                                                                        4K sectorsDaniel2008/09/22 10:31 PM
                                                                      Reasons for >512 byte sectorsDoug Siebert2008/09/22 09:38 PM
                                                                        Reasons for >512 byte sectorsrwessel2008/09/22 10:09 PM
                                                                          Reasons for >512 byte sectorsHoward Chu2008/09/23 02:50 AM
                                                                        Reasons for >512 byte sectorsDaniel2008/09/22 10:40 PM
                                                                          Reasons for >512 byte sectorsrwessel2008/09/23 09:11 AM
                                                                            Reasons for >512 byte sectorsDaniel2008/09/23 12:10 PM
                                                                      HDD long sector size availabilityEtienne Lehnart2008/09/23 05:32 AM
                                                                        HDD long sector size availabilityrwessel2008/09/23 09:19 AM
                                                                          HDD long sector size availabilityEtienne Lehnart2008/09/23 02:17 PM
                                                                    To SSD or not? Real data..Jouni Osmala2008/09/22 11:16 PM
                                                                  To SSD or not? One lady's perspectiveWes Felter2008/09/22 11:25 AM
                                                                How should SSDs be engineered into systems?Rob Thorpe2008/09/22 02:01 PM
                                                                  How should SSDs be engineered into systems?Matt Craighead2008/09/23 06:59 PM
                                                                    How should SSDs be engineered into systems?Matt Sayler2008/09/24 04:17 AM
                                                            ATA/SCSIS, Write Flushes and Asych FilesystemsTruePath2009/01/25 04:44 AM
                                                        SLC vs. MLC - the trick to latencyMichael S2008/09/12 04:58 AM
                                                        overlapped erase and read Michael S2008/09/12 04:59 AM
                                                          overlapped erase and read David W. Hess2008/09/12 09:56 AM
                                                          overlapped erase and read Anonymous2008/09/12 08:45 PM
                                                            overlapped erase and read Jouni Osmala2008/09/12 11:56 PM
                                                            overlapped erase and read Michael S2008/09/13 11:29 AM
                                                            overlapped erase and read Michael S2008/09/13 12:09 PM
                                                              overlapped erase and read Linus Torvalds2008/09/13 02:05 PM
                                                    SLC vs. MLC - the trick to latencyDoug Siebert2008/09/11 05:31 PM
                                SLC vs. MLCEduardoS2008/09/08 02:07 PM
                                  SLC vs. MLCLinus Torvalds2008/09/08 02:30 PM
                                    SLC vs. MLCEduardoS2008/09/08 04:01 PM
                                      SSD and RAIDJoe Chang2008/09/08 07:42 PM
                                        SSD and RAIDDoug Siebert2008/09/08 09:46 PM
                                          SSD and RAIDAaron Spink2008/09/09 04:27 PM
                                            SSD and RAIDGroo2008/09/10 01:02 PM
                              SLC vs. MLCJoern Engel2009/01/06 10:22 AM
                                SLC vs. MLCLinus Torvalds2009/01/06 02:04 PM
                                  SLC vs. MLCJoern Engel2009/01/06 03:24 PM
                                    SLC vs. MLCrwessel2009/01/06 04:47 PM
                                      SLC vs. MLCanonymous2009/01/06 05:17 PM
                                        SLC vs. MLCrwessel2009/01/06 05:58 PM
                                      SLC vs. MLCJoern Engel2009/01/07 12:35 AM
                                    SLC vs. MLCLinus Torvalds2009/01/06 05:45 PM
                                      SLC vs. MLCrwessel2009/01/06 06:09 PM
                                        SLC vs. MLCLinus Torvalds2009/01/06 07:47 PM
                                      SLC vs. MLCJoern Engel2009/01/07 12:26 AM
                                    SLC vs. MLCanon2009/01/06 08:23 PM
                                      SLC vs. MLCJoern Engel2009/01/07 12:52 AM
                                        SLC vs. MLCanon2009/01/07 02:34 AM
                                  SLC vs. MLCIntelUser20002009/01/07 07:43 AM
                                    SLC vs. MLCLinus Torvalds2009/01/07 10:28 AM
                                      drop data filesystem semanticDoug Siebert2009/01/09 12:21 PM
                                      FTL and FSiz2009/01/09 07:49 PM
                                        FTL and FSLinus Torvalds2009/01/09 09:53 PM
                                          FTL and FSiz2009/01/10 02:09 AM
                                            FTL and FSMichael S2009/01/10 03:19 PM
                                              compiling large programsiz2009/01/10 05:51 PM
                                                compiling large programsLinus Torvalds2009/01/10 07:58 PM
                                                  compiling large programspeter2009/01/11 05:30 AM
                                                  compiling large programsAndi Kleen2009/01/11 01:03 PM
                                                  The File AbstractionTruePath2009/01/25 06:45 AM
                                                    The File AbstractionHoward Chu2009/01/25 01:49 PM
                                                      The File AbstractionLinus Torvalds2009/01/26 09:23 AM
                                                        The File AbstractionMichael S2009/01/26 01:39 PM
                                                          The File AbstractionLinus Torvalds2009/01/26 02:31 PM
                                                            The File AbstractionDean Kent2009/01/26 03:06 PM
                                                              The File AbstractionLinus Torvalds2009/01/26 04:29 PM
                                                                The File AbstractionMark Christiansen2009/01/27 09:24 AM
                                                                  The File AbstractionMark Christiansen2009/01/27 10:14 AM
                                                                  The File AbstractionLinus Torvalds2009/01/27 10:15 AM
                                                                    The File Abstractionslacker2009/01/27 11:20 AM
                                                                      The File AbstractionLinus Torvalds2009/01/27 01:16 PM
                                                                        Attributes All The Way DownMark Christiansen2009/01/27 02:17 PM
                                                                        The File Abstractionslacker2009/01/27 05:25 PM
                                                                          The File AbstractionLinus Torvalds2009/01/28 08:17 AM
                                                                            The File Abstraction: API thoughtsCarlie Coats2009/01/28 09:35 AM
                                                                            The File Abstractionslacker2009/01/28 10:09 AM
                                                                              The File AbstractionLinus Torvalds2009/01/28 01:44 PM
                                                                                Programs already 'hide' their metadata in the bytestream, unbeknownst to usersanon2009/01/28 09:28 PM
                                                                                The File Abstractionslacker2009/01/29 10:39 AM
                                                                                  The File AbstractionLinus Torvalds2009/01/29 11:08 AM
                                                                                    The File AbstractionDean Kent2009/01/29 11:49 AM
                                                                                      The File AbstractionHoward Chu2009/01/29 02:58 PM
                                                                                        The File Abstractionrwessel2009/01/29 04:23 PM
                                                                                    Extended Attributes in Actionslacker2009/01/29 03:05 PM
                                                                                      Extended Attributes in Actionstubar2009/01/29 04:49 PM
                                                                                      Extended Attributes in ActionLinus Torvalds2009/01/29 05:15 PM
                                                                                        Like Duhanon2009/01/29 07:42 PM
                                                                                          Like Duhanon2009/01/29 09:15 PM
                                                                                            Like Duhanon2009/02/01 07:18 PM
                                                                                              Double Duh.Anonymous2009/02/01 10:58 PM
                                                                                                Double Duh.anon2009/02/02 02:08 AM
                                                                                                  Double Duh.Anonymous2009/02/02 05:11 PM
                                                                                                    Double Duh.anon2009/02/02 07:33 PM
                                                                                              Like DuhDavid Kanter2009/02/01 11:05 PM
                                                                                                Like Duhpeter2009/02/01 11:55 PM
                                                                                                Like Duhanon2009/02/02 01:55 AM
                                                                                                Xattrs, Solar power, regulation and politicsRob Thorpe2009/02/02 04:36 AM
                                                                                                  Terminology seems too fuzzy to mehobold2009/02/02 06:14 AM
                                                                                                    Terminology seems too fuzzy to merwessel2009/02/02 12:33 PM
                                                                                                      good summaryMichael S2009/02/03 02:41 AM
                                                                                                        good summaryMark Christiansen2009/02/03 09:57 AM
                                                                                                          good summaryHoward Chu2009/02/03 10:21 AM
                                                                                                            good summaryMark Christiansen2009/02/03 11:18 AM
                                                                                                              good summaryHoward Chu2009/02/03 12:00 PM
                                                                                                                good summaryMark Christiansen2009/02/03 12:36 PM
                                                                                                          good summaryRagingDragon2009/02/03 10:39 PM
                                                                                                            good summaryrwessel2009/02/03 11:03 PM
                                                                                                              good summaryRagingDragon2009/02/03 11:46 PM
                                                                                                      Terminology seems too fuzzy to meslacker2009/02/04 05:06 PM
                                                                                                        Terminology seems too fuzzy to meMichael S2009/02/05 01:05 AM
                                                                                                          Terminology seems too fuzzy to meUngo2009/02/05 01:15 PM
                                                                                                          Terminology seems too fuzzy to meslacker2009/02/05 02:19 PM
                                                                                                            Terminology seems too fuzzy to meHoward Chu2009/02/05 04:44 PM
                                                                                          Like Duhiz2009/01/30 02:03 AM
                                                                                          EAs (security labels) hosed me badlyanon2009/01/30 09:48 PM
                                                                                        Extended Attributes in ActionRagingDragon2009/01/29 09:31 PM
                                                                                      Extended Attributes in Actionanonymous2009/01/29 08:13 PM
                                                                                      Extended Attributes in ActionHoward Chu2009/01/29 09:38 PM
                                                                                        Extended Attributes in Actionslacker2009/01/30 11:24 AM
                                                                                          Extended Attributes in Actionanon2009/01/30 05:50 PM
                                                                                      Extended Attributes in ActionEtienne Lehnart2009/01/30 12:22 AM
                                                                                      Extended Attributes in ActionRob Thorpe2009/01/30 12:39 PM
                                                                                        Extended Attributes in Actionslacker2009/01/30 01:16 PM
                                                                                          Extended Attributes in Actionanon2009/01/30 06:03 PM
                                                                                          Extended Attributes in ActionHoward Chu2009/01/30 11:22 PM
                                                                                            Extended Attributes in Actionrwessel2009/01/31 12:08 AM
                                                                                              Extended Attributes in Actionanonymous2009/01/31 12:22 AM
                                                                                                Extended Attributes in Actionrwessel2009/01/31 12:56 AM
                                                                                                  ScalingDean Kent2009/01/31 09:04 AM
                                                                                                    ScalingRob Thorpe2009/02/02 02:39 AM
                                                                                                      Scalingrwessel2009/02/02 11:41 AM
                                                                                                        ScalingHoward Chu2009/02/02 12:30 PM
                                                                                                          ScalingDean Kent2009/02/02 02:27 PM
                                                                                                            ScalingRob Thorpe2009/02/03 05:08 AM
                                                                                                              ScalingDean Kent2009/02/03 07:38 AM
                                                                                                              Scalingrwessel2009/02/03 02:34 PM
                                                                                                                ScalingRagingDragon2009/02/03 10:46 PM
                                                                                                    in defense of software that does not scaleMatt Sayler2009/02/03 11:27 AM
                                                                                                      in defense of software that does not scaleHoward Chu2009/02/03 12:03 PM
                                                                                                        in defense of software that does not scaleMatt Sayler2009/02/03 12:17 PM
                                                                                                        in defense of software that does not scaleRagingDragon2009/02/03 11:00 PM
                                                                                                          in defense of software that does not scaleMichael S2009/02/04 06:46 AM
                                                                                                            in defense of software that does not scaleRagingDragon2009/02/04 09:33 PM
                                                                                                      in defense of software that does not scaleDean Kent2009/02/03 12:17 PM
                                                                                                        in defense of software that does not scaleMatt Sayler2009/02/03 12:24 PM
                                                                                                        in defense of software that does not scaleVincent Diepeveen2009/02/04 10:43 AM
                                                                                                      in defense of software that does not scalerwessel2009/02/03 02:44 PM
                                                                                                        in defense of software that does not scaleanon2009/02/04 02:35 AM
                                                                                                          in defense of software that does not scaleCarlie Coats2009/02/04 05:24 AM
                                                                                                      Scaling with time vs. scaling from the beginning.mpx2009/02/05 01:57 AM
                                                                                                  Extended Attributes in ActionMichael S2009/01/31 10:33 AM
                                                                                                  Extended Attributes in Actionanon2009/01/31 10:37 PM
                                                                                          Extended Attributes in ActionJasonB2009/01/31 08:11 AM
                                                                                            Extended Attributes in ActionHoward Chu2009/01/31 11:43 AM
                                                                                              Extended Attributes in ActionJasonB2009/01/31 04:37 PM
                                                                                                Extended Attributes in ActionHoward Chu2009/02/02 02:42 PM
                                                                                                  Extended Attributes in ActionHoward Chu2009/02/02 02:44 PM
                                                                    The File AbstractionRob Thorpe2009/01/27 11:20 AM
                                                              The File AbstractionHoward Chu2009/01/27 12:28 AM
                                                                The File AbstractionMichael S2009/01/27 03:00 AM
                                                                The File AbstractionDean Kent2009/01/27 08:30 AM
                                                        The File AbstractionAndi Kleen2009/01/27 02:05 AM
                                  SLC vs. MLCMichel2009/01/12 06:54 PM
                                    SLC vs. MLCLinus Torvalds2009/01/12 07:38 PM
                                      SLC vs. MLCrwessel2009/01/13 12:52 AM
                                      SLC vs. MLCUngo2009/01/13 03:04 PM
                                        SLC vs. MLCWes Felter2009/01/13 05:42 PM
                                      SLC vs. MLCTruePath2009/01/25 05:05 AM
                    SLC vs. MLCUngo2008/08/21 12:54 PM
                    SLC vs. MLCAaron Spink2008/08/21 01:20 PM
                  MLC vs. SLCMichael S2008/08/21 08:57 AM
                First Dunnington benchmark resultsrwessel2008/08/21 10:40 AM
              First Dunnington benchmark resultsAaron Spink2008/08/21 03:18 AM
      First Dunnington benchmark resultsEtienne Lehnart2008/08/20 04:38 AM
  Will x86 dominate big iron?Tom W2008/08/19 10:10 PM
    Will x86 dominate big iron?Jesper Frimann2008/08/20 12:28 AM
      Will x86 dominate big iron?Tom W2008/08/20 03:42 PM
        Will x86 dominate big iron?David Kanter2008/08/21 01:13 AM
          Will x86 dominate big iron?Joe Chang2008/08/21 06:54 AM
            Will x86 dominate big iron?asdf2008/08/22 01:18 PM
              Will x86 dominate big iron?Dean Kent2008/08/22 07:54 PM
        Will x86 dominate big iron?Jesper Frimann2008/08/22 09:48 AM
          Will x86 dominate big iron?Tom W2008/08/24 01:06 AM
            Will x86 dominate big iron?Michael S2008/08/24 04:19 AM
              Will x86 dominate big iron?Dean Kent2008/08/24 09:30 AM
                Will x86 dominate big iron?Paul2008/08/24 11:16 AM
                  Will x86 dominate big iron?Dean Kent2008/08/24 12:37 PM
                Will x86 dominate big iron?Michael S2008/08/25 12:53 AM
        Will x86 dominate big iron?someone2008/08/22 10:19 AM
          Will x86 dominate big iron?aaron spink2008/08/23 02:56 AM
            Will x86 dominate big iron?Michael S2008/08/23 09:58 AM
              Will x86 dominate big iron?someone2008/08/23 01:51 PM
            Will x86 dominate big iron?someone2008/08/23 01:55 PM
              Will x86 dominate big iron?Aaron Spink2008/08/23 04:52 PM
                Will x86 dominate big iron?anonymous2008/08/23 05:28 PM
                  Will x86 dominate big iron?Dean Kent2008/08/23 06:12 PM
                    Off road and topicEduardoS2008/08/23 06:28 PM
                  Will x86 dominate big iron?someone2008/08/23 06:26 PM
                    Will x86 dominate big iron?Dean Kent2008/08/23 09:40 PM
                    Will x86 dominate big iron?anonymous2008/08/24 01:46 AM
                  Off road and topicDavid W. Hess2008/08/24 03:24 AM
                    Off road and topicAaron Spink2008/08/24 04:14 AM
  Beckton vs. DunningtonMr. Camel2008/08/22 06:30 AM
    Beckton vs. Dunningtonjokerman2008/08/22 12:12 PM
    Beckton vs. DunningtonMr. Camel2009/05/29 10:16 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell green?