Z77 Pro4-M - new INF driver update?

[highcountryrider]

Re: Z77 Pro4-M - new INF driver update?

Ignorance is not a sin.....stupidity is!!
"A man's got to know his limitations."
You've been asking excellent questions in your threads.
Clearly you know your limitations and appreciate the wisdom of seeking help before your system gets upgemucked**

**Sorry for using yet another high tech term.

You've proven that you are NOT a member of any of the following clubs:

PEBKAC ("Problem Exists Between Keyboard And Chair")
PEBCAK ("Problem Exists Between Chair and Keyboard")
PICNIC ("Problem In Chair Not In Computer")
PIBCAK ("Problem Is Between Chair And Keyboard")
ID-10T error ("Idiot error")

User error - Wikipedia

profJim, one thing I have always known is my limitations! I've sort of made the jump now from just wanting a system that works well to taking control and tweaking it to maximize performance. Obviously, the majority of that is because it is my system; I've built it. That means I got to choose what went into it and now how to set it up. While that's the good part, it's also the scary part. My appoach: baby steps and ask a lot of quetions.

I used to think I knew a fair amount about computers. I am smart enough to know now how little that actually was. Hang in there with me prof and please let me know when I'm about go down in flames.

[profJim]

Re: Z77 Pro4-M - new INF driver update?

"limitions" are one thing, limitations are a whole 'nother thing.
Don't feel bad, FireFox underlines my misspellings while I'm writing a post and I still don't catch them all.
The problem you're having is what we all go through when learning new things; for every answer we get, we then have several new questions questions. parsec and synack can best assist you with overclocking because they have first hand experience with ASRock systems. Even when I think that I know the answer, I usually try not to advise unless I have first hand experience and can try out some settings on my equipment. I'll be watching from the sidelines and butting in from time to time when I have something to further confuse you and others with the facts.

With an Ivy Bridge cpu and mobo, getting to about 4.40GHz stable should be fairly easy. The main barrier to reaching the upper 4GHz range is when cpu temperatures rise very quickly as you apply more voltage. You can search the web for ASRock Ivy Bridge overclocking guides or tutorials to start to learn some of the basics. Sin0822 has posted some outstanding Gigabyte z77 motherboard reviews that include overclocking suggestions. Don't get hung up in all the details, the important thing is to understand the main concepts. Some of the terminology will be different, but it's similar to the difference between American and Canadian English.

I'm sure that synack will be commenting on my last sentence.
Life is a journey, enjoy the ride.

[parsec]

Re: Z77 Pro4-M - new INF driver update?

Honestly that makes absolutely no sense to me. Swap RAM to RAM?!?! Why on earth would you ever want to do that? If you have the RAM you don't need to swap... if you don't have the RAM, no RAM to swap to...
Unless the RAM drive pre-allocates the memory, even if it's not being used. but in that case, swapping out RAM to RAM would produce page faults and would REDUCE overall system performance.

Syn

(HCR, the following is an example of the "experts" discussing a subtle detail (HA!) that one of them noticed that could negate the need and advantage of doing something (putting the pagefile on a RAM Disk), and another realizing he is right, but then explains the motivation for doing so, while hoping that goal remains worth while.)

Syn, what you are saying is correct, if I understand your statement, which IMO needs this preface: A Windows pagefile, which is normally a file on a disk, is/can be used as a replacement for a lack of system/RAM memory on a PC. BTW, yes, the RAM drive pre-allocates the memory.

If so I see your point, which is, why reduce the amount of system memory available, by creating a RAM disk, which then potentially could cause the need for a program to use a pagefile as replacement memory.

To top that all off, we then put the pagefile in the RAM disk (which would be unnecessary if we had not created the RAM disk in the first place), and the OS must perform the extra work of reading the pagefile on the RAM Disk, instead of just directly using system memory in the first place!

As I'm sure you know, putting the pagefile on the RAM Disk is done to prevent the pagefile from being created on a SSD OS drive, to reduce the writes performed on the SSD. That of course is part of the "wear out my SSD" paranoia that still exists among many SSD users (including myself), as well as the "you must have a pagefile!" warning that haunts those that want to get rid of it.

When the actual usage of a pagefile as pseudo-memory is started is unknown to me, as in what threshold or situation causes it to begin. I have 16GB of RAM, and even with a 4GB RAM Disk, rarely use more than 50% of that 16GB. So I don't see myself causing the situation you described. HCR has 8GB, so using 4GB of it for a RAM Disk does increase the potential for the situation you described to occur. That is at least partially why Prof. Jim suggested creating a smaller RAM Disk, IMO.

There are other side affects of putting the pagefile on a RAM Disk, such as the loss of storing the memory image when a system crash/BSOD occurs, which Windows would read from the pagefile on the next startup, if the RAM Disk is not backed up.

HCR, no harm done so not to worry, at worst you might lose a few milliseconds, and use a few extra microwatts of power. And as always, "experts" are only human...

[synack]

Re: Z77 Pro4-M - new INF driver update?

I see your point on not wanting the swap on SSD.
But, how many writes happen to a page file if you have little to no memory contention? I can't image very many.

I understand that the page file is to free up inactive RAM to make more room for active RAM. Which of course is really only an issue when memory is in contention. Yes, I do know that Microsoft, in their infinite wisdom, force swap out parts of the system and will have issues if it can't. Real dumb if you ask me... But Microsoft never seems to lol...


I just still don't like the idea of putting the pagefile in RAM. page faults, context switches, yuck... if it were me running windows, (thankfully not! haha) I would configuring the RAM drive and subsequent page file as small as possible. I think 4/2GB of 16GB in your case parsec, would be my upper limit. With 8GB of RAM I'd be reluctant to size it at 2/1GB and would probably try for 1/.5GB.

Windows is a pig with RAM, add to that file caching in memory, DLL preloads (even worse if you have MS office installed), and you have burnt through 2-4GB of RAM.

My opinion is that if you NEED a swap file, You need RAM.

Anyway.. sorry to sidetrack the thread here. It's just that when I saw the SWAP on x: (ramdrive), my first reaction was WHAAAAAT?!?!?! lol


Syn

[parsec]

Re: Z77 Pro4-M - new INF driver update?

Syn, I totally agree with you, I'm glad you brought up that topic, since it got me thinking about the relevance of it all.

Frankly, I could care less about the pagefile existing at all, but you mentioned the reality of it in Windows. I've run PCs without one at all for weeks, no problem.

That 2GB pagefile is what the ASR utility sets up by default, and I haven't played with it, but HCR said it has a mind of it's own when creating the pagefile size... or is that Windows too... hmm...

The most enjoyable usage I get from the RAM Disk is, I set it as my download target drive. So I download a driver or program installation file right into memory. Then run it from there, now that is fast! Can't really do that with really big files, but otherwise works great. Copy the file to disk if I want to save it.

Thanks for tolerating us Windozers...

[highcountryrider]

Re: Z77 Pro4-M - new INF driver update?

Whooooosh! That's the sound of what most of what you guys said going right over my head but I sort of get the gist of the pros and cons. Right now the ASR RAM drive I created is 3 GB with 1.5 GB being automatically allocated to the pagefile. So far I haven't noticed any adverse effects (other than a couple of small anomolies) of running the RAM drive but conversely I'm not aware of any noticable advantage.

One of the main reasons I decided to experiment with having a RAM drive was parsec's mention of a lot of SSD user's paranoia about reducing the amount of writes to the SSD in order to increase longevity. Being new to SSDs, and given the relative price difference between SSDs and traditional platter HDDs, I assumed that the RAM drive option would be advantageous in achieving a longer life span for my SSD. Although I obviously don't understand (yet) a lot of Syn's objection to using RAM for RAM (is that right?) I do see that there are some potentially adverse effects of using a RAM drive. This is especially true of his mention of DLL preloads (again, no idea what that means) relative to MS Office which I will be using.

I'm just now ordering another 8 GB of DDR3 SDRAM ($34 - can't believe how inexpensive it's getting!) which will up my total to 16 GB. I wil then re-evaluate whether to use the RAM drive option.

Syn, as both parsec and profJim are aware, I suffer from BCD (Brain Cell Depletion) which is triggered by an overload of new information. This condition has mysteriously accelerated since I built this new system but this type of discussion is well worth the loss of few hundred thousand BCs.

[synack]

Re: Z77 Pro4-M - new INF driver update?

LOL.. I only have a few BCs left, so I need to be cautious of my info intake. :)

Just to clear up a few points, and hopefully not too technical.

RAM drives = Good.
Faster "disk" access is always good. Not good if it causes memory contention and the system needs to page in/out.

Page file = Bad for performance, Good for one off situations. Like an insurance for your RAM. Insurance that comes at a cost of performance. The more you need it, the more it costs.

Page file on RAM drives = Bad, mostly
It does have the benefit of reducing disk accesses on the SSD.

Preloads/caching = Generally good.
Most modern operating systems allow you to preload into memory and cache dynamic stuff like DLLs, files etc. This comes as a cost to RAM, but the performance to cost ratio is normally worth it. preloads however cause issues where memory is short. That is because when memory is short, the page file is required, and that is bad.

Context switching
This is something that is required due to multitasking operating systems. When a processor is working on one thread, it is in that "context" when it switches to another process it switches to that context. Windows "virtually" allocates RAM (4GB for 32bit process, 8TB for 64bit processes) which get mapped to physical RAM in "pages." When memory contention is high, least active pages get sent to swap. Now think about a really busy multitasking system that switches between processes and is forced to swap in and out "pages" based on what is the current context... which is why I say "yikes"

If you have ever run out of RAM on a busy system and see the hard drive light pinned? that's what's happening. page "faults" (fault means that the page was not in ram when required and the memory manager needs to possibly free a chunk of RAM and swap in the required page from "disk".) combined with context switches when memory is low will kill any system.

When a RAM drive on Windows allocates a chunk of memory (say 4GB) it is not available for other applications or the system to use. Less memory = greater need for swap file (see swap file = bad above)
Once swap is required the system needs to page out from RAM, to the RAM drive (which is RAM...) so that is cycles lost to redundancy.

And it appears the way the RAM drive was architected, that you could potentially get into the situation where you swap out the RAM drive to the swap file on the RAM drive...(swap ram to ram drive to ram drive to ram drive etc...) which will immediately crash your system.
It's all a cyclic redundancy that should be in my opinion avoided.

So as long as your Memory is larger than you need, and the RAM drive is sized appropriately, it is a good solution.
Making the RAM drive too big, with the page file on it, is asking for trouble.

I'm not a RAM drive hater, I use one built into Linux. That one however does not pre-allocate all the RAM (if only 15k of files are on it, it's only 15k large), it's only mapped to "temp folders" and no swap files are on it (actually I don't run a swap file at all)

Final thought: You will gain more performance letting the system do "it's thing" by preloading and caching, than you will by adding in another layer of complexity. As long as you realize that the RAM drive is not a performance enhancer but simply a tool to fix a problem (like eliminate SSD accesses for swap) or have a fast temp download place (lolz at parsec, you do know that files are cached in memory right? except maybe really big ones).
It does not speed up the system as whole, and has a very real risk of actually slowing it down if not implemented carefully.

phew.. that was longer than I anticipated haha

Syn

[highcountryrider]

Re: Z77 Pro4-M - new INF driver update?

OK, now I'm really sorry I asked the question! Did I ask the question? Just kidding....a very good explanation of some processes and RAMifications (OK, a bad pun!) that I didn't realize were all connected to using a RAM drive the way I'm using it.

Since I've ordered 8 GB of additional RAM I think what I might do is modify the RAM drive and increase it's size to 4 GB, then remove the pagefile from it and only use it for System and User Temp files and also for IE and Firefox cache files. Of course, I would then have to put the pagefile back on the SSD as I originally had it. I understand I could also put it on my WD HDD (Drive E). Would that be a better solution than having it on the SSD? With the new RAM I would still have 12 GB available for the OS on the C drive which should be more than adequate for what my needs are right now.

I'm afraid what I've done here is forgotten the basic axiom KISS (keep it simple stupid) in my enthusiasm for getting the maximum performance out of my new system. I'm stepping into configurations that I'm not really qualified to be doing, all with potentially negative effects. Of course, I could ask how can I improve my computing skills if I don't try new things? Darn, sometimes I hate this stuff!

[synack]

Re: Z77 Pro4-M - new INF driver update?

I'm not saying DON'T EVER do it. Just be aware of what you are doing, why, and the risks.
My shock was more to do with ASRock doing it by default and forced if I read you guys correctly. Without any sort of warning or whatever.

Go ahead and setup the ram drive with the pagefile. but if your system performance degrades, you know why and possibly how to fix it. (by maybe decreasing the size or disabling etc.)

As I said with parsec before, the point of moving the page file off of SSD and onto RAM is valid. Just understand why and the risks.
Putting it on a standard hard drive if you have one is also valid and good practice anyway.

Syn

[parsec]

Re: Z77 Pro4-M - new INF driver update?

Now Syn, of course I know files are cached in memory, although is the term "cached" appropriate any time something is put into memory? But I see your point.

My use of the RAM Drive for downloads of driver and other installation files, is really mainly used during a new OS installation. Since I always install Windows with ONLY the OS target drive powered up, as otherwise Windows can put the MBR on another disk, I would use a USB flash drive to hold all my drivers, etc, for installation. That is slow to read of course, and some programs will not let you run them from a flash drive.

So I used the RAM Drive to run installation programs from. And yes, that no doubt too reads it from RAM Drive memory, into standard memory for execution. But, that technique is faster than any SSD, and it rips through the installations in nothing flat. Also, no read and write thrashing that would occur if the installation program was running on the same drive you are installing to.

[highcountryrider]

Re: Z77 Pro4-M - new INF driver update?

I'm not saying DON'T EVER do it. Just be aware of what you are doing, why, and the risks.
My shock was more to do with ASRock doing it by default and forced if I read you guys correctly. Without any sort of warning or whatever.

Go ahead and setup the ram drive with the pagefile. but if your system performance degrades, you know why and possibly how to fix it. (by maybe decreasing the size or disabling etc.)

As I said with parsec before, the point of moving the page file off of SSD and onto RAM is valid. Just understand why and the risks.
Putting it on a standard hard drive if you have one is also valid and good practice anyway.

Syn

OK, thanks for the advice. I'll wait until my new RAM is installed and make a decision then. Appreciate, as always, the replies.

[JAlford]

Re: Z77 Pro4-M - new INF driver update?

I have a similar question about new Z77 drivers. I have just installed Windows 8 upgrade from an XP OS on a Z77 Pro4M with a Gigabyte Radeon HD7770 video card. The question is: should I go ahead and install the new INF drivers from the AS Rock site or leave it alone if it is operating? I do have a couple of problems; one is not be able to wake up the system after the monitor goes to sleep and the USB wireless not coming back on when I reboot. I have to unplug and re-plug the wireless USB. I did install the new drivers for the video card and SATA drivers. Windows 8 already had the basic Intel video and sound drivers.

[parsec]

Re: Z77 Pro4-M - new INF driver update?

The inf files are very basic and necessary, they are the first thing I install for a new Windows installation. They're also not dangerous, and actually other drivers depend on the inf information to install correctly. Who knows what state they are in after your upgrade to Windows 8. I would not worry about running the inf file installer.