Physical vs effective write latency

By: Linus Torvalds (torvalds.delete@this.linux-foundation.org), September 11, 2008 6:16 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 S08/19/08 09:54 AM
  First Dunnington benchmark resultsrwessel08/19/08 12:42 PM
    First Dunnington benchmark resultsAaron Apink08/19/08 04:49 PM
      First Dunnington benchmark resultsJoe Chang08/19/08 05:28 PM
        First Dunnington benchmark resultsrwessel08/21/08 08:49 AM
          First Dunnington benchmark resultsJoe Chang08/21/08 02:10 PM
            First Dunnington benchmark resultsrwessel08/21/08 05:42 PM
              First Dunnington benchmark resultsJoe Chang08/21/08 06:12 PM
      First Dunnington benchmark resultsrwessel08/21/08 08:45 AM
        First Dunnington benchmark resultsAaron Spink08/21/08 12:12 PM
          First Dunnington benchmark resultsJoe Chang08/21/08 02:15 PM
    First Dunnington benchmark resultsRichard Cownie08/20/08 01:59 AM
      First Dunnington benchmark resultsAnders Jensen08/20/08 02:26 AM
        +SSDAnders Jensen08/20/08 02:30 AM
        First Dunnington benchmark resultsRichard Cownie08/20/08 10:04 AM
          First Dunnington benchmark resultsslacker08/20/08 11:35 AM
            First Dunnington benchmark resultsDoug Siebert08/20/08 06:54 PM
              First Dunnington benchmark resultsRichard Cownie08/20/08 07:58 PM
                SLC vs. MLCDavid Kanter08/21/08 12:16 AM
                  SLC vs. MLCMatt Sayler08/21/08 05:25 AM
                  SLC vs. MLCRichard Cownie08/21/08 05:32 AM
                    SLC vs. MLCLinus Torvalds08/21/08 07:39 AM
                      SLC vs. MLCMichael S08/21/08 08:07 AM
                        SLC vs. MLCLinus Torvalds08/21/08 08:52 AM
                          SLC vs. MLCMichael S08/21/08 09:35 AM
                            OLTP appliance = mainframe? (NT)Potatoswatter08/21/08 10:44 AM
                              OLTP appliance = HP NonStop?Michael S08/21/08 11:03 AM
                                OLTP applianceJoe Chang08/21/08 02:33 PM
                                  OLTP appliancePotatoswatter08/21/08 02:59 PM
                            SLC vs. MLCAaron Spink08/21/08 12:29 PM
                          SLC vs. MLCDan Downs08/21/08 10:33 AM
                            SLC vs. MLCrwessel08/21/08 11:45 AM
                              SLC vs. MLCDan Downs08/22/08 07:21 AM
                            SLC vs. MLCAaron Spink08/21/08 12:34 PM
                          SLC vs. MLC vs DRAMpgerassi08/21/08 11:24 AM
                            SLC vs. MLC vs DRAMDavid Kanter08/22/08 12:31 AM
                          SLC vs. MLCGroo08/23/08 11:52 AM
                      SLC vs. MLCDoug Siebert08/21/08 05:14 PM
                        SLC vs. MLCLinus Torvalds08/22/08 07:05 AM
                          SLC vs. MLCDoug Siebert08/22/08 01:27 PM
                            SLC vs. MLCEduardoS08/22/08 05:26 PM
                            SSD Controller differentiationDavid Kanter08/22/08 08:35 PM
                              SSD Controller differentiationDoug Siebert08/22/08 09:34 PM
                                SSD Controller differentiation (supercaps, cost...)anon08/23/08 09:18 AM
                                  SSD Controller differentiation (supercaps, cost...)Doug Siebert08/23/08 09:40 AM
                            SLC vs. MLCLinus Torvalds08/23/08 09:50 AM
                              SLC vs. MLCLinus Torvalds09/08/08 11:03 AM
                                SLC vs. MLCMax09/08/08 12:51 PM
                                  SLC vs. MLCHoward Chu09/08/08 08:04 PM
                                    SLC vs. MLCMax09/08/08 09:29 PM
                                      SLC vs. MLCHoward Chu09/08/08 11:12 PM
                                        RAM vs SSD?Jouni Osmala09/09/08 12:06 AM
                                          RAM vs SSD?Max09/12/08 11:51 AM
                                            RAM vs SSD?EduardoS09/12/08 03:27 PM
                                            Disk cache snapshottingMax09/13/08 07:34 AM
                                              Disk cache snapshottingHoward Chu09/14/08 08:58 PM
                                                Disk cache snapshottingMax09/15/08 11:50 AM
                                        SLC vs. MLCLinus Torvalds09/09/08 06:43 AM
                                          SLC vs. MLCHoward Chu09/09/08 08:42 AM
                                            SLC vs. MLCLinus Torvalds09/09/08 09:39 AM
                                              SLC vs. MLCMichael S09/09/08 11:29 PM
                                                SLC vs. MLCanon09/10/08 01:51 AM
                                                  SLC vs. MLCMichael S09/10/08 02:09 AM
                                                    SLC vs. MLCMax09/10/08 03:48 AM
                                                      SLC vs. MLCMichael S09/10/08 04:52 AM
                                                        SLC vs. MLCMax09/10/08 05:28 AM
                                                    SLC vs. MLCMatt Sayler09/10/08 05:21 AM
                                                      SLC vs. MLCMichael S09/10/08 08:17 AM
                                                    SLC vs. MLCanon09/10/08 05:29 AM
                                                      SLC vs. MLCMichael S09/10/08 08:23 AM
                                                        SLC vs. MLCMatt Sayler09/10/08 09:45 AM
                                                SLC vs. MLCLinus Torvalds09/10/08 06:25 AM
                                                  SLC vs. MLCMichael S09/10/08 08:54 AM
                                                    SLC vs. MLCLinus Torvalds09/10/08 09:31 AM
                                                      Physical vs effective write latencyMax09/11/08 06:35 AM
                                                        Physical vs effective write latencyLinus Torvalds09/11/08 08:06 AM
                                                          Physical vs effective write latencyLinus Torvalds09/11/08 08:48 AM
                                                            Physical vs effective write latencyLinus Torvalds09/11/08 10:39 AM
                                                              Physical vs effective write latencyMark Roulo09/11/08 11:18 AM
                                                                Physical vs effective write latencyDoug Siebert09/11/08 04:59 PM
                                                                  Physical vs effective write latencyLinus Torvalds09/11/08 06:16 PM
                                                                    Physical vs effective write latencyDoug Siebert09/11/08 09:28 PM
                                                        Physical vs effective write latencyMS02/03/09 02:06 PM
                                                  SLC vs. MLC - the trick to latencyAnonymous09/11/08 11:39 AM
                                                    SLC vs. MLC - the trick to latencyanon09/11/08 12:17 PM
                                                      SLC vs. MLC - the trick to latencyAnonymous09/11/08 04:25 PM
                                                        SLC vs. MLC - the trick to latencyDoug Siebert09/11/08 04:47 PM
                                                          SLC vs. MLC - the trick to latencyrwessel09/11/08 05:01 PM
                                                        SLC vs. MLC - the trick to latencyanon09/11/08 11:00 PM
                                                          SLC vs. MLC - the trick to latencyAnonymous09/12/08 07:52 PM
                                                            SLC vs. MLC - the trick to latencyanon09/13/08 09:06 AM
                                                            SLC vs. MLC - the trick to latencyUngo09/15/08 11:18 AM
                                                              To SSD or not? One lady's perspectiveDavid Kanter09/22/08 12:12 AM
                                                                To SSD or not? One lady's perspectiveHoward Chu09/22/08 03:02 AM
                                                                  To SSD or not? Real data..Linus Torvalds09/22/08 06:33 AM
                                                                    To SSD or not? Real data..Ungo09/22/08 11:27 AM
                                                                      4K sectorsWes Felter09/22/08 05:03 PM
                                                                        4K sectorsDaniel09/22/08 09:31 PM
                                                                      Reasons for >512 byte sectorsDoug Siebert09/22/08 08:38 PM
                                                                        Reasons for >512 byte sectorsrwessel09/22/08 09:09 PM
                                                                          Reasons for >512 byte sectorsHoward Chu09/23/08 01:50 AM
                                                                        Reasons for >512 byte sectorsDaniel09/22/08 09:40 PM
                                                                          Reasons for >512 byte sectorsrwessel09/23/08 08:11 AM
                                                                            Reasons for >512 byte sectorsDaniel09/23/08 11:10 AM
                                                                      HDD long sector size availabilityEtienne Lehnart09/23/08 04:32 AM
                                                                        HDD long sector size availabilityrwessel09/23/08 08:19 AM
                                                                          HDD long sector size availabilityEtienne Lehnart09/23/08 01:17 PM
                                                                    To SSD or not? Real data..Jouni Osmala09/22/08 10:16 PM
                                                                  To SSD or not? One lady's perspectiveWes Felter09/22/08 10:25 AM
                                                                How should SSDs be engineered into systems?Rob Thorpe09/22/08 01:01 PM
                                                                  How should SSDs be engineered into systems?Matt Craighead09/23/08 05:59 PM
                                                                    How should SSDs be engineered into systems?Matt Sayler09/24/08 03:17 AM
                                                            ATA/SCSIS, Write Flushes and Asych FilesystemsTruePath01/25/09 03:44 AM
                                                        SLC vs. MLC - the trick to latencyMichael S09/12/08 03:58 AM
                                                        overlapped erase and read Michael S09/12/08 03:59 AM
                                                          overlapped erase and read David W. Hess09/12/08 08:56 AM
                                                          overlapped erase and read Anonymous09/12/08 07:45 PM
                                                            overlapped erase and read Jouni Osmala09/12/08 10:56 PM
                                                            overlapped erase and read Michael S09/13/08 10:29 AM
                                                            overlapped erase and read Michael S09/13/08 11:09 AM
                                                              overlapped erase and read Linus Torvalds09/13/08 01:05 PM
                                                    SLC vs. MLC - the trick to latencyDoug Siebert09/11/08 04:31 PM
                                SLC vs. MLCEduardoS09/08/08 01:07 PM
                                  SLC vs. MLCLinus Torvalds09/08/08 01:30 PM
                                    SLC vs. MLCEduardoS09/08/08 03:01 PM
                                      SSD and RAIDJoe Chang09/08/08 06:42 PM
                                        SSD and RAIDDoug Siebert09/08/08 08:46 PM
                                          SSD and RAIDAaron Spink09/09/08 03:27 PM
                                            SSD and RAIDGroo09/10/08 12:02 PM
                              SLC vs. MLCJoern Engel01/06/09 09:22 AM
                                SLC vs. MLCLinus Torvalds01/06/09 01:04 PM
                                  SLC vs. MLCJoern Engel01/06/09 02:24 PM
                                    SLC vs. MLCrwessel01/06/09 03:47 PM
                                      SLC vs. MLCanonymous01/06/09 04:17 PM
                                        SLC vs. MLCrwessel01/06/09 04:58 PM
                                      SLC vs. MLCJoern Engel01/06/09 11:35 PM
                                    SLC vs. MLCLinus Torvalds01/06/09 04:45 PM
                                      SLC vs. MLCrwessel01/06/09 05:09 PM
                                        SLC vs. MLCLinus Torvalds01/06/09 06:47 PM
                                      SLC vs. MLCJoern Engel01/06/09 11:26 PM
                                    SLC vs. MLCanon01/06/09 07:23 PM
                                      SLC vs. MLCJoern Engel01/06/09 11:52 PM
                                        SLC vs. MLCanon01/07/09 01:34 AM
                                  SLC vs. MLCIntelUser200001/07/09 06:43 AM
                                    SLC vs. MLCLinus Torvalds01/07/09 09:28 AM
                                      drop data filesystem semanticDoug Siebert01/09/09 11:21 AM
                                      FTL and FSiz01/09/09 06:49 PM
                                        FTL and FSLinus Torvalds01/09/09 08:53 PM
                                          FTL and FSiz01/10/09 01:09 AM
                                            FTL and FSMichael S01/10/09 02:19 PM
                                              compiling large programsiz01/10/09 04:51 PM
                                                compiling large programsLinus Torvalds01/10/09 06:58 PM
                                                  compiling large programspeter01/11/09 04:30 AM
                                                  compiling large programsAndi Kleen01/11/09 12:03 PM
                                                  The File AbstractionTruePath01/25/09 05:45 AM
                                                    The File AbstractionHoward Chu01/25/09 12:49 PM
                                                      The File AbstractionLinus Torvalds01/26/09 08:23 AM
                                                        The File AbstractionMichael S01/26/09 12:39 PM
                                                          The File AbstractionLinus Torvalds01/26/09 01:31 PM
                                                            The File AbstractionDean Kent01/26/09 02:06 PM
                                                              The File AbstractionLinus Torvalds01/26/09 03:29 PM
                                                                The File AbstractionMark Christiansen01/27/09 08:24 AM
                                                                  The File AbstractionMark Christiansen01/27/09 09:14 AM
                                                                  The File AbstractionLinus Torvalds01/27/09 09:15 AM
                                                                    The File Abstractionslacker01/27/09 10:20 AM
                                                                      The File AbstractionLinus Torvalds01/27/09 12:16 PM
                                                                        Attributes All The Way DownMark Christiansen01/27/09 01:17 PM
                                                                        The File Abstractionslacker01/27/09 04:25 PM
                                                                          The File AbstractionLinus Torvalds01/28/09 07:17 AM
                                                                            The File Abstraction: API thoughtsCarlie Coats01/28/09 08:35 AM
                                                                            The File Abstractionslacker01/28/09 09:09 AM
                                                                              The File AbstractionLinus Torvalds01/28/09 12:44 PM
                                                                                Programs already 'hide' their metadata in the bytestream, unbeknownst to usersanon01/28/09 08:28 PM
                                                                                The File Abstractionslacker01/29/09 09:39 AM
                                                                                  The File AbstractionLinus Torvalds01/29/09 10:08 AM
                                                                                    The File AbstractionDean Kent01/29/09 10:49 AM
                                                                                      The File AbstractionHoward Chu01/29/09 01:58 PM
                                                                                        The File Abstractionrwessel01/29/09 03:23 PM
                                                                                    Extended Attributes in Actionslacker01/29/09 02:05 PM
                                                                                      Extended Attributes in Actionstubar01/29/09 03:49 PM
                                                                                      Extended Attributes in ActionLinus Torvalds01/29/09 04:15 PM
                                                                                        Like Duhanon01/29/09 06:42 PM
                                                                                          Like Duhanon01/29/09 08:15 PM
                                                                                            Like Duhanon02/01/09 06:18 PM
                                                                                              Double Duh.Anonymous02/01/09 09:58 PM
                                                                                                Double Duh.anon02/02/09 01:08 AM
                                                                                                  Double Duh.Anonymous02/02/09 04:11 PM
                                                                                                    Double Duh.anon02/02/09 06:33 PM
                                                                                              Like DuhDavid Kanter02/01/09 10:05 PM
                                                                                                Like Duhpeter02/01/09 10:55 PM
                                                                                                Like Duhanon02/02/09 12:55 AM
                                                                                                Xattrs, Solar power, regulation and politicsRob Thorpe02/02/09 03:36 AM
                                                                                                  Terminology seems too fuzzy to mehobold02/02/09 05:14 AM
                                                                                                    Terminology seems too fuzzy to merwessel02/02/09 11:33 AM
                                                                                                      good summaryMichael S02/03/09 01:41 AM
                                                                                                        good summaryMark Christiansen02/03/09 08:57 AM
                                                                                                          good summaryHoward Chu02/03/09 09:21 AM
                                                                                                            good summaryMark Christiansen02/03/09 10:18 AM
                                                                                                              good summaryHoward Chu02/03/09 11:00 AM
                                                                                                                good summaryMark Christiansen02/03/09 11:36 AM
                                                                                                          good summaryRagingDragon02/03/09 09:39 PM
                                                                                                            good summaryrwessel02/03/09 10:03 PM
                                                                                                              good summaryRagingDragon02/03/09 10:46 PM
                                                                                                      Terminology seems too fuzzy to meslacker02/04/09 04:06 PM
                                                                                                        Terminology seems too fuzzy to meMichael S02/05/09 12:05 AM
                                                                                                          Terminology seems too fuzzy to meUngo02/05/09 12:15 PM
                                                                                                          Terminology seems too fuzzy to meslacker02/05/09 01:19 PM
                                                                                                            Terminology seems too fuzzy to meHoward Chu02/05/09 03:44 PM
                                                                                          Like Duhiz01/30/09 01:03 AM
                                                                                          EAs (security labels) hosed me badlyanon01/30/09 08:48 PM
                                                                                        Extended Attributes in ActionRagingDragon01/29/09 08:31 PM
                                                                                      Extended Attributes in Actionanonymous01/29/09 07:13 PM
                                                                                      Extended Attributes in ActionHoward Chu01/29/09 08:38 PM
                                                                                        Extended Attributes in Actionslacker01/30/09 10:24 AM
                                                                                          Extended Attributes in Actionanon01/30/09 04:50 PM
                                                                                      Extended Attributes in ActionEtienne Lehnart01/29/09 11:22 PM
                                                                                      Extended Attributes in ActionRob Thorpe01/30/09 11:39 AM
                                                                                        Extended Attributes in Actionslacker01/30/09 12:16 PM
                                                                                          Extended Attributes in Actionanon01/30/09 05:03 PM
                                                                                          Extended Attributes in ActionHoward Chu01/30/09 10:22 PM
                                                                                            Extended Attributes in Actionrwessel01/30/09 11:08 PM
                                                                                              Extended Attributes in Actionanonymous01/30/09 11:22 PM
                                                                                                Extended Attributes in Actionrwessel01/30/09 11:56 PM
                                                                                                  ScalingDean Kent01/31/09 08:04 AM
                                                                                                    ScalingRob Thorpe02/02/09 01:39 AM
                                                                                                      Scalingrwessel02/02/09 10:41 AM
                                                                                                        ScalingHoward Chu02/02/09 11:30 AM
                                                                                                          ScalingDean Kent02/02/09 01:27 PM
                                                                                                            ScalingRob Thorpe02/03/09 04:08 AM
                                                                                                              ScalingDean Kent02/03/09 06:38 AM
                                                                                                              Scalingrwessel02/03/09 01:34 PM
                                                                                                                ScalingRagingDragon02/03/09 09:46 PM
                                                                                                    in defense of software that does not scaleMatt Sayler02/03/09 10:27 AM
                                                                                                      in defense of software that does not scaleHoward Chu02/03/09 11:03 AM
                                                                                                        in defense of software that does not scaleMatt Sayler02/03/09 11:17 AM
                                                                                                        in defense of software that does not scaleRagingDragon02/03/09 10:00 PM
                                                                                                          in defense of software that does not scaleMichael S02/04/09 05:46 AM
                                                                                                            in defense of software that does not scaleRagingDragon02/04/09 08:33 PM
                                                                                                      in defense of software that does not scaleDean Kent02/03/09 11:17 AM
                                                                                                        in defense of software that does not scaleMatt Sayler02/03/09 11:24 AM
                                                                                                        in defense of software that does not scaleVincent Diepeveen02/04/09 09:43 AM
                                                                                                      in defense of software that does not scalerwessel02/03/09 01:44 PM
                                                                                                        in defense of software that does not scaleanon02/04/09 01:35 AM
                                                                                                          in defense of software that does not scaleCarlie Coats02/04/09 04:24 AM
                                                                                                      Scaling with time vs. scaling from the beginning.mpx02/05/09 12:57 AM
                                                                                                  Extended Attributes in ActionMichael S01/31/09 09:33 AM
                                                                                                  Extended Attributes in Actionanon01/31/09 09:37 PM
                                                                                          Extended Attributes in ActionJasonB01/31/09 07:11 AM
                                                                                            Extended Attributes in ActionHoward Chu01/31/09 10:43 AM
                                                                                              Extended Attributes in ActionJasonB01/31/09 03:37 PM
                                                                                                Extended Attributes in ActionHoward Chu02/02/09 01:42 PM
                                                                                                  Extended Attributes in ActionHoward Chu02/02/09 01:44 PM
                                                                    The File AbstractionRob Thorpe01/27/09 10:20 AM
                                                              The File AbstractionHoward Chu01/26/09 11:28 PM
                                                                The File AbstractionMichael S01/27/09 02:00 AM
                                                                The File AbstractionDean Kent01/27/09 07:30 AM
                                                        The File AbstractionAndi Kleen01/27/09 01:05 AM
                                  SLC vs. MLCMichel01/12/09 05:54 PM
                                    SLC vs. MLCLinus Torvalds01/12/09 06:38 PM
                                      SLC vs. MLCrwessel01/12/09 11:52 PM
                                      SLC vs. MLCUngo01/13/09 02:04 PM
                                        SLC vs. MLCWes Felter01/13/09 04:42 PM
                                      SLC vs. MLCTruePath01/25/09 04:05 AM
                    SLC vs. MLCUngo08/21/08 11:54 AM
                    SLC vs. MLCAaron Spink08/21/08 12:20 PM
                  MLC vs. SLCMichael S08/21/08 07:57 AM
                First Dunnington benchmark resultsrwessel08/21/08 09:40 AM
              First Dunnington benchmark resultsAaron Spink08/21/08 02:18 AM
      First Dunnington benchmark resultsEtienne Lehnart08/20/08 03:38 AM
  Will x86 dominate big iron?Tom W08/19/08 09:10 PM
    Will x86 dominate big iron?Jesper Frimann08/19/08 11:28 PM
      Will x86 dominate big iron?Tom W08/20/08 02:42 PM
        Will x86 dominate big iron?David Kanter08/21/08 12:13 AM
          Will x86 dominate big iron?Joe Chang08/21/08 05:54 AM
            Will x86 dominate big iron?asdf08/22/08 12:18 PM
              Will x86 dominate big iron?Dean Kent08/22/08 06:54 PM
        Will x86 dominate big iron?Jesper Frimann08/22/08 08:48 AM
          Will x86 dominate big iron?Tom W08/24/08 12:06 AM
            Will x86 dominate big iron?Michael S08/24/08 03:19 AM
              Will x86 dominate big iron?Dean Kent08/24/08 08:30 AM
                Will x86 dominate big iron?Paul08/24/08 10:16 AM
                  Will x86 dominate big iron?Dean Kent08/24/08 11:37 AM
                Will x86 dominate big iron?Michael S08/24/08 11:53 PM
        Will x86 dominate big iron?someone08/22/08 09:19 AM
          Will x86 dominate big iron?aaron spink08/23/08 01:56 AM
            Will x86 dominate big iron?Michael S08/23/08 08:58 AM
              Will x86 dominate big iron?someone08/23/08 12:51 PM
            Will x86 dominate big iron?someone08/23/08 12:55 PM
              Will x86 dominate big iron?Aaron Spink08/23/08 03:52 PM
                Will x86 dominate big iron?anonymous08/23/08 04:28 PM
                  Will x86 dominate big iron?Dean Kent08/23/08 05:12 PM
                    Off road and topicEduardoS08/23/08 05:28 PM
                  Will x86 dominate big iron?someone08/23/08 05:26 PM
                    Will x86 dominate big iron?Dean Kent08/23/08 08:40 PM
                    Will x86 dominate big iron?anonymous08/24/08 12:46 AM
                  Off road and topicDavid W. Hess08/24/08 02:24 AM
                    Off road and topicAaron Spink08/24/08 03:14 AM
  Beckton vs. DunningtonMr. Camel08/22/08 05:30 AM
    Beckton vs. Dunningtonjokerman08/22/08 11:12 AM
    Beckton vs. DunningtonMr. Camel05/29/09 09:16 AM
Reply to this Topic
Name:
Email:
Topic:
Body: No Text
How do you spell green?