topbanner_forum
  *

avatar image

Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
  • Thursday March 28, 2024, 11:57 am
  • Proudly celebrating 15+ years online.
  • Donate now to become a lifetime supporting member of the site and get a non-expiring license key for all of our programs.
  • donate

Last post Author Topic: In search of ... RAMdisk opinions  (Read 33993 times)

barney

  • Charter Member
  • Joined in 2006
  • ***
  • Posts: 1,294
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #25 on: September 24, 2012, 04:28 PM »
P.S. If I had a dollar each time a critical backup (made using standard "enterprise grade" backup software) ended up being corrupted (and sometimes not recoverable) I'd have enough money to take a not overly modest vacation on the Continent.

Yea, verily!  Religiously backing up in obeisance to a god that we later discover does not exist or was naught but a false prophet.

f0dder

  • Charter Honorary Member
  • Joined in 2005
  • ***
  • Posts: 9,153
  • [Well, THAT escalated quickly!]
    • View Profile
    • f0dder's place
    • Read more about this member.
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #26 on: September 24, 2012, 04:29 PM »
Wrong  :down:.  Well, presumptive.
No, right :-) *tongue-in-cheek*. Keep in mind that I'm referring to the "put pagefile on ramdisk" as insanity, not general use of ramdisks!

So, isn't that 1.5G RAM wasted?  It's not being used by system or software.
Well, depending on OS and the memory reporting gadget, the 1.5gig of ram could be used as filesystem cache, which is reported as 'free memory' since the caches can be dropped/flushed as need be. Just a thing to keep in mind!

So why not make that a RAMdisk as mentioned previously, thus decreasing writes to HDD or SSD, thus decreasing wear & tear?
Sure thing, I'm all for that - and doing it myself. Just don't put the pagefile on ramdisk, as mentioned previously - it's better disabling it altogether if you got that kind of memory.
- carpe noctem

barney

  • Charter Member
  • Joined in 2006
  • ***
  • Posts: 1,294
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #27 on: September 24, 2012, 04:38 PM »
Sure thing, I'm all for that - and doing it myself. Just don't put the pagefile on ramdisk, as mentioned previously - it's better disabling it altogether if you got that kind of memory.

OK, that's been expressed a few times now, but I do not see the difference between running the pagefile in memory or on a RAMdisk.  Only difference I discern is the operating software for the RAMdisk.  Is there something else that makes this untenable?

f0dder

  • Charter Honorary Member
  • Joined in 2005
  • ***
  • Posts: 9,153
  • [Well, THAT escalated quickly!]
    • View Profile
    • f0dder's place
    • Read more about this member.
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #28 on: September 24, 2012, 05:02 PM »
Sure thing, I'm all for that - and doing it myself. Just don't put the pagefile on ramdisk, as mentioned previously - it's better disabling it altogether if you got that kind of memory.
OK, that's been expressed a few times now, but I do not see the difference between running the pagefile in memory or on a RAMdisk.
If you disable the pagefile (System Properties -> 'Advanced' tab -> 'settings' button in 'Performance' group -> 'Advanced' tab -> 'Advanced' tab -> 'change' button in 'Virtual memory' group -> set to 'no paging file' for all partitions), you aren't "running the pagefile in memory", you're disabling it entirely.

Not swapping at all is even faster than swapping to memory... and since you're not using memory for a pagefile ramdisk, you have more memory to allocate from before swapping would be necessary. It really should be common sense :-)

Do note that you shouldn't disable the pagefile unless you always have enough free physical memory, even under high load, or you at least know the implications of what turning off the pagefile means. Windows doesn't really like running out of memory (but at least it doesn't go about OOM-killing processes like Linux does by default).

- carpe noctem

40hz

  • Supporting Member
  • Joined in 2007
  • **
  • Posts: 11,857
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #29 on: September 24, 2012, 06:01 PM »
P.S. If I had a dollar each time a critical backup (made using standard "enterprise grade" backup software) ended up being corrupted (and sometimes not recoverable) I'd have enough money to take a not overly modest vacation on the Continent.

Yea, verily!  Religiously backing up in obeisance to a god that we later discover does not exist or was naught but a false prophet.


LOL! I see you've dealt with MS Exchange backups too! 8) :Thmbsup:

40hz

  • Supporting Member
  • Joined in 2007
  • **
  • Posts: 11,857
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #30 on: September 24, 2012, 06:10 PM »
Do note that you shouldn't disable the pagefile unless you always have enough free physical memory, even under high load, or you at least know the implications of what turning off the pagefile means. Windows doesn't really like running out of memory (but at least it doesn't go about OOM-killing processes like Linux does by default).

Well Linux might have done it the same way...except Microsoft has a patent on the BSOD - so Linux can't just let itself fault-crash like Windows does. :P
-----------------------------------------------------------
@f0 - (Sorry. Just kidding. I couldn't resist!) ;) ;D

Tinman57

  • Charter Member
  • Joined in 2006
  • ***
  • Posts: 1,702
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #31 on: September 24, 2012, 07:06 PM »
  Back in my old Amiga days the RAM-Disk was pretty much a must, and it worked great on the AmigaDos.  Then when I went to W-95 I researched RAM-disk and found that back then, RAM-disk software was rare and you had to be pretty much a computer scientist to set it up.  lol  I haven't researched it since then and don't even know if it's worth it or not on Windows machines, I just doubled the RAM on my puter and kept on truckin'.....

barney

  • Charter Member
  • Joined in 2006
  • ***
  • Posts: 1,294
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #32 on: September 24, 2012, 08:14 PM »
LOL! I see you've dealt with MS Exchange backups too! 8) :Thmbsup

Oh, hell no!  This started way before Exchange was a gleam in some developer's eye.  MS-DOS v2.1, I thimk, introduced me to the concept of PC backups [conceptually deficient - backup worked, restore didn't] ... I've distrusted 'em ever since  ;D.  Granted, MS Exchange, if it doesn't take top spot, at least ranks in the top three (3) damnittohell recovery scenarios, but I've given up on backups, for the most part.  The backup takes time, and recovery is always questionable, no matter the source or methodology.

I've had better luck with disk images, but even there the results can be erratic.

MS Exchange is a latecomer to this party  :P.

barney

  • Charter Member
  • Joined in 2006
  • ***
  • Posts: 1,294
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #33 on: September 24, 2012, 08:23 PM »
Do note that you shouldn't disable the pagefile unless you always have enough free physical memory, even under high load, or you at least know the implications of what turning off the pagefile means. Windows doesn't really like running out of memory (but at least it doesn't go about OOM-killing processes like Linux does by default).

Well, I just disabled - I thimk - the pagefile.  Don't see any significant RAM usage above and beyond the norm.  But, then, I'm just doing forum posts at the moment.  However, I normally have a bunch of stuff loaded at boot, so the actual program load is pretty hefty.  Barring major malfunctions, I'll try this for a week or two (2), then reactivate the pagefile on a RAMdisk, see if there's a performance hit - or any other significant difference for that matter.

However, my system(s) be head-and-shoulders above what most of my compatriots use, so I still won't know whether it would make a difference to them.

tomos

  • Charter Member
  • Joined in 2006
  • ***
  • Posts: 11,959
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #34 on: September 25, 2012, 03:01 AM »
Do note that you shouldn't disable the pagefile unless you always have enough free physical memory, even under high load, or you at least know the implications of what turning off the pagefile means. Windows doesn't really like running out of memory (but at least it doesn't go about OOM-killing processes like Linux does by default).

Well, I just disabled - I thimk - the pagefile.  Don't see any significant RAM usage above and beyond the norm.  But, then, I'm just doing forum posts at the moment.  However, I normally have a bunch of stuff loaded at boot, so the actual program load is pretty hefty.  Barring major malfunctions, I'll try this for a week or two (2), then reactivate the pagefile on a RAMdisk, see if there's a performance hit - or any other significant difference for that matter.
Let us know how you get on barney :)

Re the Pagefile:
it's (supposed to be) for when you run out of memory.
Therefore it seems counterproductive to use up some of that same memory in order to create a pagefile. Cause then you'll be forcing computer to use the pagefile sooner than normally otherwise necessary.
Unless, of course as said above, the machine is 32bit and has memory over & above whatever the limit is for 32bit
Tom

f0dder

  • Charter Honorary Member
  • Joined in 2005
  • ***
  • Posts: 9,153
  • [Well, THAT escalated quickly!]
    • View Profile
    • f0dder's place
    • Read more about this member.
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #35 on: September 25, 2012, 10:19 AM »
Well, I just disabled - I thimk - the pagefile.  Don't see any significant RAM usage above and beyond the norm.  But, then, I'm just doing forum posts at the moment.  However, I normally have a bunch of stuff loaded at boot, so the actual program load is pretty hefty.
I started disabling the pagefile when I had 1gig ram, I belive back in the win2k days - or perhaps early XP. A few things back then did cause problems, most notably I had to shut down every other process to play some games. With 2gig and XP (definitely before SP3, though!) I don't recall running into problems. Of course the amount of ram necessary to disable the pagefile will depend on your workloads.

Barring major malfunctions, I'll try this for a week or two (2), then reactivate the pagefile on a RAMdisk, see if there's a performance hit - or any other significant difference for that matter.
I really don't see why anybody in their right mind would do this, though, considering the reason(nings) I've posted previously, and tomos' in the post right above this one as well.

Yes, if you're on a 32bit OS and have unadressable RAM, you could do it, and it might even sorta make sense - but unless you have stuffed a lot of unadressable RAM in your box (and then I'd be questioning your sanity :P), you'd still a disk-based pagefile, since the meager amount that's normally unadressable wouldn't suffice for a system that actually needs to swap...
- carpe noctem

cranioscopical

  • Friend of the Site
  • Supporting Member
  • Joined in 2006
  • **
  • Posts: 4,776
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #36 on: September 25, 2012, 10:48 AM »
Fortunately, at DoCo we don't get too many "forum-cops"
I think somebody already said that…

Renegade

  • Charter Member
  • Joined in 2005
  • ***
  • Posts: 13,288
  • Tell me something you don't know...
    • View Profile
    • Renegade Minds
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #37 on: September 25, 2012, 11:00 AM »
Fortunately, at DoCo we don't get too many "forum-cops" or "discussion historians" who just loooove to point out there's already been a thread about that "new" topic you just posted. ;D
I think somebody already said that…

And I think that most people here actually have lives and simply aren't petty enough to bother wasting however long to drag it up and link as a kind of nyah-nyah-nyah-nyah-nyah. Well, beyond pointing out that there's another thread, as that is actually helpful in many cases. :P

Did I already say that? Getting a deja vu here... :P

Did I already say that? Getting a deja vu here... :P

Did I already say that? Getting a deja vu here... :P

Did I already say that? Getting a deja vu here... :P

Did I already say that? Getting a deja vu here... :P

Did I already say that? Getting a deja vu here... :P

Did I already say that? Getting a deja vu here... :P

:P
Slow Down Music - Where I commit thought crimes...

Freedom is the right to be wrong, not the right to do wrong. - John Diefenbaker

barney

  • Charter Member
  • Joined in 2006
  • ***
  • Posts: 1,294
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #38 on: September 25, 2012, 06:29 PM »
I really don't see why anybody in their right mind
Hey!  You leave my mind out of this! It's already unstable enough as it is  :P!

Rationale is as follows:
  • I've had the pagefile run away with me on some systems, not always identifiable as to why.
  • I always limit the size of the pagefile ... call it quirky, but I do.
  • Mindful of SSD stuff I've read recently, a RAMdisk pagefile reduces writes, although 'twould seem the jury is still out on that one.
  • I've had the pagefile on a USB stick - not fast, but no noticeable deterioration in performance.  This is just another test.
  • Finally, I won't know for certain until I try it myself, in real life, regardless the mentors that exist here.

If I'm even possibly likely to help someone else with this, I kinda need to understand the workings from experience, not just conversation no matter how erudite and knowledgeable the contributors to the conversation might be.

[Sidebar]
After I eliminated the pagefile, my memory usage decreased by seven (7) to ten (10) percent.  Not certain just what that implies, but I wouldn't have know w/o the experimentation.
[/Sidebar]

[Addendum]
If that memory is not otherwise being used, why not put it to use?
[/Addendum]

40hz

  • Supporting Member
  • Joined in 2007
  • **
  • Posts: 11,857
    • View Profile
    • Donate to Member
Re: In search of ... RAMdisk opinions
« Reply #39 on: September 25, 2012, 07:58 PM »
    Finally, I won't know for certain until I try it myself, in real life, regardless the mentors that exist here.[/li]
    [/list]

    If I'm even possibly likely to help someone else with this, I kinda need to understand the workings from experience, not just conversation no matter how erudite and knowledgeable the contributors to the conversation might be.

    @barney - Good man! That's the right way to approach this stuff. Sure, do the research and homework. But hands-on is everything. Best way to learn how a computer system works is to set something up you don't mind breaking every so often. Then break it every so often.

    If nothing else, you'll learn what not to do - and also how to recover a borked system.  8) :Thmbsup:

    « Last Edit: September 25, 2012, 08:05 PM by 40hz »

    barney

    • Charter Member
    • Joined in 2006
    • ***
    • Posts: 1,294
      • View Profile
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #40 on: September 25, 2012, 09:30 PM »
    If nothing else, you'll learn what not to do - and also how to recover a borked system.  8) :Thmbsup

    Hee, hee.  Been doin' that for the last twenty (20) or thirty (30) years  ;).  What can I say, I'm a slow learner  :P.

    IainB

    • Supporting Member
    • Joined in 2008
    • **
    • Posts: 7,540
    • @Slartibartfarst
      • View Profile
      • Read more about this member.
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #41 on: September 28, 2012, 09:58 PM »
    I've been following this discussion thread with interest but without so far being able to contribute much that I thought could be of real use/help to it, until, perhaps, this last couple of weeks.
    The subject could perhaps be more accurately restated as a general "Making greater use of available RAM for efficiency and performance iprovement", rather than just focussing on one approach to doing this - ie., RAMdisk.

    The main advantage of a RAMdisk (or RAMdrive) is performance improvement. Though I had used memory optimisers and the IBM DOS RAMdisk in laptops since years ago, I have trialled them but not really used them (seemed too kludgy or flaky or of no real advantage) in later Windows GUI operating systems. In DOS they certainly did have a noticeable effect on performance, largely because they could reduce latency by reducing the HDD read/write activity in HDD TEMP and paging operations, and they also took advantage of any unused "upper memory". So, I knew that RAM drives could offer real potential performance benefits under some circumstances.

    For some time I have coincidentally been trialling potential performance-improving tweaks - hardware/software. The single most significant performance improvement achieved so far in this (and which I posted about in detail in the DC Forum as Test: Does latency reduction via RAM upgrade lift software performance?) came about after bumping up the DDR3 RAM to the max 8Gb (on available slots) on two laptops.
    Suffice it to say that the performance improvement was significant in terms of raising the WEI (Windows Experience Index) "Memory operations per second" subscore, but not the overall Base Score (which is determined by the lowest subscore).

    However, from a user perspective
    ..."everything seems to run faster"...I have also now turned on (previously turned off most) all the settings for max graphics quality - on both laptops - which, theoretically should place extra load on the CPUs. However, the user experience is that display quality/resolution has improved and there has been no perceptible waiting/latency from any processes.

    Thus, as a general rule, anything you can do to push operations requiring physical HDD activity into RAM are going to reduce latency and enable an improved real speed of operation from the user perspective.

    The laptops are:
    • a DELL Inspiron with an AMD Phenom II X3 N850 (triple-core) CPU and a 5,400RPM HDD, OS Win7-64 Home Premium.
    • an HP ENVY 14 with an Intel i7 (4x2=8-core) CPU and a 7,200rpm HDD, OS Win7-64 Home Premium.

    With all that extra RAM, I have been trying to apply the above general rule by seeing what other operations I can push into RAM - for example, having the Firefox cache in RAM:
    So much opinion! So, in the absence of facts, I take a suck-it-and-see approach. My HP ENVY's Firefox about:config thus currently has been (and is still being) tweaked to something similar to these trial settings for RAM/disk caching utlilisation: (caveat - might not suit everybody's needs though)

    Firefox RAM and disk cache settings in about-config.png

    Also, over the last couple of weeks, what I have been doing (in piecemeal fashion, because it is "when I have the time") is trialling the standard Windows ReadyBoost facility in Win7. This was prompted by my reading @Stephen66515's very informative post (displaying most of the Windows Help file details on ReadyBoost): Using memory in your storage device to speed up your computer.
    There are several other posts in the DC Forum referring to the ReadyBoost Windows feature.

    Whilst I have so far made no objective measures of the effects of using ReadyBoost, I can report that it certainly seems to work as it should, and that it seems to provide some latency reduction, though I do not yet understand how to make the most of this. Presumably the max read/write speed of the USB RAM ("memory stick" or whatever you might call it) is a constraint, and so measuring that would seem to be useful.
    Gaining a better understanding of the constraints/limitations should enable me to better understand how to take best advantage of any ReadyBoost potential benefits.

    NB: Whilst one of the potential disadvantages of a RAMdisk or RAM caching might be (for some) the loss of the RAM-based contents on power-down (Shutdown), it is not necessarily so relevant for laptop users who keep the laptop in "Sleep mode" most of the time when not using it, rather than Shutdown. Also, for those that might want it, there will probably be ways to make a backup HDD copy of files in a RAM disk/cache - if necessary - before terminating the process or before Shutdown. (Though I haven't played around with this yet.)

    barney

    • Charter Member
    • Joined in 2006
    • ***
    • Posts: 1,294
      • View Profile
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #42 on: September 29, 2012, 04:31 PM »
    There seem to be, basically, two (2) facets of the RAM question.  And not just on DC - I've seen this semi-battle for a decade or better.

    Some say to hold it in reserve, save it for a rainy day, so to speak.

    Others say to use it, max it out.  Unused RAM is a waste.

    I don't know that I agree with either camp.  The opinion here on DC seems equally divided:  do it, don't do it, and don't know.  However, barring catastrophic failure, four (4) weeks - two (2) each - should provide at least some decision points.

    I hope  :P.

    [Sidebar]
    It took an inordinate amount of time to do this on a Nexus 7.  If tablets are to replace PC/Laptop units,  a vastly improved input system will have to be devised  :o - and voice is not always a viable, much less preferable, choice  :huh:.
    [/Sidebar]

    Curt

    • Supporting Member
    • Joined in 2006
    • **
    • Posts: 7,566
      • View Profile
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #43 on: September 29, 2012, 05:46 PM »
    ..., I take a suck-it-and-see approach.

    Unless one is an arctic monkey, the term is suck in and see.

    Thank you for the link to explaining Ready Boost - I never got it to do anything noticeable for me. Instead I have been using eBoostr, and is satisfied with this. Of course only because I merely have 4GB RAM - on my next PC I sincerely hope to access much more memory.

    IainB

    • Supporting Member
    • Joined in 2008
    • **
    • Posts: 7,540
    • @Slartibartfarst
      • View Profile
      • Read more about this member.
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #44 on: September 30, 2012, 06:33 AM »
    ..., I take a suck-it-and-see approach.

    Unless one is an arctic monkey, the term is suck in and see.

    Thank you for the link to explaining Ready Boost - I never got it to do anything noticeable for me. Instead I have been using eBoostr, and is satisfied with this. Of course only because I merely have 4GB RAM - on my next PC I sincerely hope to access much more memory.

    The term I used was correct use of English for the purpose intended:
    suck it and see
    UK informal
    Definition
    to try something to find out if it will be successful
    I'm not sure whether this paint is the right colour for the bedroom - we'll just have to suck it and see.

    (Definition of suck it and see from the Cambridge Advanced Learner's Dictionary & Thesaurus © Cambridge University Press)
    http://dictionary.ca...tish/suck-it-and-see

    I don't think I had heard of the Arctic Monkeys' song of the same name before.

    Thank you for the mention of eBoostr - I don't think I had heard of that before, either. Am doing some research on it now...

    EDIT: 2012-10-01 1552hrs (NZT)
    With all this talk of "sucking", I wonder if it is worth referring also to the old adage about "Teaching your grandmother to suck eggs."?...    :P
    « Last Edit: September 30, 2012, 09:55 PM by IainB »

    f0dder

    • Charter Honorary Member
    • Joined in 2005
    • ***
    • Posts: 9,153
    • [Well, THAT escalated quickly!]
      • View Profile
      • f0dder's place
      • Read more about this member.
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #45 on: September 30, 2012, 07:29 AM »
    I really don't see why anybody in their right mind
    Hey!  You leave my mind out of this! It's already unstable enough as it is  :P!
    :D

    • I've had the pagefile run away with me on some systems, not always identifiable as to why.
    • I always limit the size of the pagefile ... call it quirky, but I do.
    Humm, back when I used pagefiles, I used to set a largeish minimum bound, so that under normal operation and moderate load, the pagefile size would stay constant (and unfragmented!), whereas exceptional situations could grow as large as needed. That would still be my recommendation - but if you have badly written software that runs amok, putting an upper bound might be reasonable...

    • Mindful of SSD stuff I've read recently, a RAMdisk pagefile reduces writes, although 'twould seem the jury is still out on that one.
    Entirely disabling the pagefile entirely disables writes, though :-)

    • I've had the pagefile on a USB stick - not fast, but no noticeable deterioration in performance.  This is just another test.
    Personally wouldn't do this, because of how flaky usb sticks are - using ReadyBoost might be a somewhat safer middle ground.

    [Sidebar]
    After I eliminated the pagefile, my memory usage decreased by seven (7) to ten (10) percent.  Not certain just what that implies, but I wouldn't have know w/o the experimentation.
    [/Sidebar]
    Memory management is complex - some applications tune their usage depending on "free memory", and what "free memory" means is also up for debate. If you go by "unused physical RAM", you ignore the fact that filesystem cache can be quickly discarded to serve large memory requests.

    [Addendum]If that memory is not otherwise being used, why not put it to use?[/Addendum]
    Indeed. It's worth noting, though, that memory used for a RAMdisk can't be use for normal filesystem cache - and if you allocate a large ramdisk but only use part of it, technically that's still unused memory. It's all about striking a balance :-)

    I've got 16 gigs of memory in my current rig, but only have a 1 gig permanent ramdrive. That's large enough for (most) %TEMP%, firefox profile+cache, WebsiteWatcher profile+cache, and scratch space (it's usually ~50% full). Anything larger than that would be wasted memory - but for special occasions, I can quickly add a temporary disk.

    Suffice it to say that the performance improvement was significant in terms of raising the WEI (Windows Experience Index) "Memory operations per second" subscore, but not the overall Base Score (which is determined by the lowest subscore).
    Slight nag: putting more RAM in a machine doesn't give you more memory operations per second, unless the RAM you're putting in is faster. (there's an exception if you go from single to two sticks because of dual-channel memory architecture, though). Unfortunately, for some reason the WEI score does seem to limit the score to 5.9 unless you have >4GB of memory (no matter how fast the RAM is) - which is plain misleading, IMHO.

    As for disabling firefox disk cache, ho humm. Might be an advantage, but not having stuff cached between sessions? I chose the third way: moving firefox profile + cache to RAMdisk. Best of both worlds :)

    Whilst I have so far made no objective measures of the effects of using ReadyBoost, I can report that it certainly seems to work as it should, and that it seems to provide some latency reduction, though I do not yet understand how to make the most of this. Presumably the max read/write speed of the USB RAM ("memory stick" or whatever you might call it) is a constraint, and so measuring that would seem to be useful.
    Well, yes and no.

    ReadyBoost only serves "random" I/O requests (i.e., smallish scattered-all-over reads), whereas large linear sequences are served from harddrive. It's done this way because normal flash memory (usb pendrives, SD cards etc., not full-on SSDs) is quite a bit slower than harddrives - I have a decent USB2 corsair pendrive that does ~20MB/s and ~24MB/s reads, whereas my Velociraptor does ~140MB/s. But once you start doing highly random 4kb I/O, those figures change dramatically - even a high-end HDD like the velociraptor drops in below 1MB/s, whereas flash ram holds up much better (contrary to common belief, they aren't 100% directly addressable, though, so there will be some performance drop).

    RB is somewhat intriguing, since it isn't just "more memory for the pagefile", it's actually a full disk caching layer - so unless you run an SSD, you might see some benefit from it even if you have large enough amounts of RAM (that would mostly be after a cold system boot, though, as memory filesystem cache would then get filled, and memory is insanely much faster than flash memory). I don't think I'd ever use RB, though - flash memory is too fragile, and enough ram + SuperFetch should be better.

    Also note that ReadyBoost does both compression and AES128 encryption of the cache files, so there will be some CPU overhead.

    Also, for those that might want it, there will probably be ways to make a backup HDD copy of files in a RAM disk/cache - if necessary - before terminating the process or before Shutdown. (Though I haven't played around with this yet.)
    More than one RAM disk offers persistance, saving the memory content to disk on shutdown. SoftPerfect RAM disk might be worth checking out :)

    - carpe noctem

    IainB

    • Supporting Member
    • Joined in 2008
    • **
    • Posts: 7,540
    • @Slartibartfarst
      • View Profile
      • Read more about this member.
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #46 on: September 30, 2012, 09:37 PM »
    Suffice it to say that the performance improvement was significant in terms of raising the WEI (Windows Experience Index) "Memory operations per second" subscore, but not the overall Base Score (which is determined by the lowest subscore).
    Slight nag: putting more RAM in a machine doesn't give you more memory operations per second, unless the RAM you're putting in is faster. (there's an exception if you go from single to two sticks because of dual-channel memory architecture, though). Unfortunately, for some reason the WEI score does seem to limit the score to 5.9 unless you have >4GB of memory (no matter how fast the RAM is) - which is plain misleading, IMHO.
    As for disabling firefox disk cache, ho humm. Might be an advantage, but not having stuff cached between sessions? I chose the third way: moving firefox profile + cache to RAMdisk. Best of both worlds :)

    I think I would probably prefer your "best of both worlds" approach too.
    However, I are now confuzzled: What you say about one/two sticks seems to make sense, but it doesn't seem to tally with the summary results I posted regarding the RAM upgrade:

    DDR3 RAM upgrade performance table.jpg
    ____________________________________
    Whilst I have so far made no objective measures of the effects of using ReadyBoost, I can report that it certainly seems to work as it should, and that it seems to provide some latency reduction, though I do not yet understand how to make the most of this. Presumably the max read/write speed of the USB RAM ("memory stick" or whatever you might call it) is a constraint, and so measuring that would seem to be useful.
    Well, yes and no.
    ReadyBoost only ...
    RB is somewhat intriguing...
    ...ReadyBoost does both compression and AES128 encryption of the cache files, so there will be some CPU overhead.

    What you write there is very interesting. Thanks for the info.    :Thmbsup:
    I am now in a position to be able to add some objective info/metrics as regards how ReadyBoost is currently configured/running on my laptop:

    The Flash RAM I used is an oldish 4Gb Toshiba USB stick that I had lying around with some folders/files already saved to it. ReadyBoost configured itself and grabbed the spare 3.2Gb for its cache. The RAM is one which has an LED on it which flashes when the RAM is in use. It's been flashing on/off a lot. HDSentinel can see it and view its performance:

    ReadyBoost - HDS view 01.png

    ReadyBoost - HDS view 02.png

    ReadyBoost - HDS view 03.png
    ____________________________________
    Also, for those that might want it, there will probably be ways to make a backup HDD copy of files in a RAM disk/cache - if necessary - before terminating the process or before Shutdown. (Though I haven't played around with this yet.)
    More than one RAM disk offers persistance, saving the memory content to disk on shutdown. SoftPerfect RAM disk might be worth checking out :)

    Yes, thanks. Interesting. As I read up on RAMdisk software, I realise that some of them offer pretty sophisticated solutions - e.g., ImDisk also looked interesting...
    ____________________________________

    Curt

    • Supporting Member
    • Joined in 2006
    • **
    • Posts: 7,566
      • View Profile
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #47 on: October 01, 2012, 02:46 AM »
    ... eBoostr - I don't think I had heard of that before, either. Am doing some research on it now...

    I should say it hasn't been updated for quite a while. Not that it needs to, I am just not sure it's being maintained / developed any more.

    ---
    Thank you for correction on how to suck...

    Curt

    • Supporting Member
    • Joined in 2006
    • **
    • Posts: 7,566
      • View Profile
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #48 on: October 01, 2012, 10:59 AM »
    accidentally, SoftPerfect and Imdisk are the two  S L O W E S T  ram disks...! How would I know? Of course, I don't! But because of this thread, I have taken the time to read what Raymond.cc has to say about this subject, and it turns out he has a very informing TEST of 12 of the ram disks available:

    http://www.raymond.c...ead-and-write-speed/

    The two fastest RAM disks are both shareware:
    http://www.romexsoft...amdisk/features.html
    http://www.gilisoft..../product-ramdisk.htm

    Some may want to test the weird interesting Bond http://bonddisc.com/ - it is something different from the others; a ram disk is merely a part of it.

    But go to Raymond and read the test.

    UPDATE (23 October 2012):
    However, see >f0dder's tests< for different test results!
    « Last Edit: October 22, 2012, 05:11 PM by Curt »

    f0dder

    • Charter Honorary Member
    • Joined in 2005
    • ***
    • Posts: 9,153
    • [Well, THAT escalated quickly!]
      • View Profile
      • f0dder's place
      • Read more about this member.
      • Donate to Member
    Re: In search of ... RAMdisk opinions
    « Reply #49 on: October 01, 2012, 05:34 PM »
    I think I would probably prefer your "best of both worlds" approach too.
    However, I are now confuzzled: What you say about one/two sticks seems to make sense, but it doesn't seem to tally with the summary results I posted regarding the RAM upgrade:
    Yep, as I mentioned: " Unfortunately, for some reason the WEI score does seem to limit the score to 5.9 unless you have >4GB of memory (no matter how fast the RAM is) - which is plain misleading, IMHO."

    It's a darn silly thing to do, since "amount of installed ram" is a metric that's entirely different from "speed of installed ram", and is easy to obtain.

    Curt: those ramdisk benchmarks look very interesting - would be interesting to see if it's something one can repeat. And I really do wonder what the heck causes those pathologically speeds, if the benches are true. Drivers are hard to program - apparently even something as conceptually trivial as a ramdisk. Wish I had time to bring out my trusty disassembler and go code spelunking :-)
    - carpe noctem