The Official PC building thread -3rd Edition

creaky & sammorris,
I just ran the newest version of Auslogics defrager, and it automatically does all the drives at the same time unless you uncheck any of them. As far as skipping large files goes, it seems to only do that if you select it from the menu.



I ran it on the new eMachine, and it took a whole 3 min, 22 sec to do 160GB. That's outstanding. Thank you Creaky and Sam!

I just ordered my new Athlon II x4 630 Quad. Should be here tomorrow or Friday.

Russ

 

My apology! Yes, avoid big-endian processors (like the G4, though I love it AltVec coprocessor). Intel & other modern processors are little-endian. (Big-endian order was traditional for Unix - though not Linix, but this is more romantic than practical.)Software can't switch bytes, so before distributing software beyond my own house, I have to recompile it on an Intel machine. Some cross-compilers have been built, but it's more practical to build an Intel machine in the future. (By then I may be able to afford an old Intel Mac.

MacOSX deviated significantly from Unix, with its own virtual memory modifications, by the time the still-expensive Intel Mac appeared.

No, I really didn't know how to optimize a machine for *nix, but I assumed Apple's engineers did. The 2002 Quicksilver was designed for MacOSX 1.0 (a Unix) and was US $200. MacOSX offers a numerical linear algebra library, which I assume Apple engineers optimized for the AltVec short-word vector processor (akin to the Cray supercomputer).

About 'ominous': no, I refused to work for government labs. The algorithms implement new thermodynamic theorems, initially written in the 'J' (APL) computing language. The theorems allow one to determine much about the environment where granites & such formed. (They have no military applications I'm aware of

When students, we punched cards in assembler or computer-dependent Fortran, then stood in line to hand them to the mainframe operator. Then we fixed our comma & did it again. When Stallman released EMACS, I remember commenting 'Why should I want a monitor, let alone the ability to move about the screen?'

 

Defraggler user myself. It works well enough for me as I have no real fragmentation or storage issues ATM. It makes my games run smoother and my load times shorter and that's all I really care about. It's faster than Windows defrag and is very easy to use

 

http://www.fudzilla.com/content/view/16159/1/

not looking goof for win7 users who want to use an intel gen2 SSD

 

Forgive my messy post above: a spurious keystroke combination caused Afterdawn to post my initial draft. (Supporting the last statement, I suppose.) I offer the following as an apology.

Regularly, I either use a defragmenter or perform a file-by-file restore of the partition. (The latter is very convenient if Linux or Unix has been partitioned for this purpose.)

My need, with small disks, is to defragment free space, not files. This leaves sufficient contiguous free space for swapping applications.

However, if using a defragmenter to thoroughly tune a disk, I find it much faster to do it in three stages: (1) defragment the files themselves, leaving spaces between them; (2) compact the disk, eliminating these spaces; and (3) sort the directory structure, speeding the ability to find the pieces. Many defragmenters offer to do all the above at once, but I've found this often takes many times as long.

 

I've been playing with Auslogics lately and it supposedly can defrag your system files and optimise the drives at the same time. I love that it does multiple disks at once

Also it moves OS files to the outside of the disk for faster access times. I might do a before and after on HD Tune as I have hardly defragged on 7 and then only with Defraggler yet.

 

Shaff,
The thing I don't understand is why people rush to upgrade firmware and software every time there's something new. I guess I just don't understand the "Gotta-Have-It" mentality. I'm an "If it ain't broke, Don't fix it" person myself. I get stuff from customers like that all the time where someone has upgraded something, just because it's the latest, even though everything was working fine before they decided to change things. Hell, I crashed the bios in a BioStar G-Force MB, simply because someone at BioStar made a mistake and accidentally switched the G-Force and the T-force bios revisions when making up the download files, which rendered the bios useless. It cost me time, and money for shipping, not to mention that I had to pull everything apart and then put it all back together again! I've yet to upgrade the bios in my UD4P, and it works just fine. I have all the revisions in case I ever do need one, but I see no point in upgrading, just to have the latest (and supposedly the greatest) firmware or bios revision.

I also strongly believe that the manufacturers should give more info on what the different revisions actually do, or correct! Just my opinion of course!

Best Regards,
Russ

 

well this firmware has added ALOT to the performance. its one of the biggest updates intel had given.

this article shows the new firmware

http://www.anandtech.com/storage/showdoc.aspx?i=3667&p=8

butt hat specific page is fantastic! anandtech really are the kings of benching ssds

 

I've never updated a HDD firmware. But if a firmware addresses a huge performance gain...I'd probably do it LOL!

 

Oman7,
I would too, only I would let others be the Guinea Pigs for a month or so first, and wait for the feedback. Sometimes, we learn patience the hard way! Been there! Done that, the hard way! LOL!!

Russ

 

It doesn't make any difference to me. It wouldn't appear Samsung bother to release newer firmwares....

 

Oh no! I wasn't suggesting you a fool. In fact I agree with you completely. HDD firmwares are apparently not a laughing matter LOL! However i'll laugh about this mishap LOL!

 

I mentioned last week about flashing my OCZ SSD's firmware which for the most part was the same reason as this Intel story, Win7 TRIM capabilities. I'm glad I didnt have the same problem.
There is not a huge performance gap between between the Intel and most newer generation SSD's. Sustained read and write times are comparable but random, smaller writes which a Windows OS is good at are much faster on an Intel. They bogged down the early compition and caused some stuttering but that has gone away with newer controllers and firmware revisions. Tweaks that a person with enough geek in them to early adopt an SSD, are easily found online and can minimize the slow writes soooo once you minimize the slow random writes you are left with 2 drives with comparable sustained read and writes.

 

Nice! Can one really do such things these days? When I stopped consulting, I stopped studying improvements in hardware design; but I thought that HDD drivers were, essentially, deep mysteries: we no longer knew how our hard disks were organized, where the partitions went, or even where the files went within the partitions. Are these still controllable by us?

There were great varieties in the above; and Unix usually placed its 'directory' in the center of the disk, Microsoft on the outside of the disk's platter. Because each user's habits are individual, no HDD can optimize the placement of partitions & files. (The BSD Unix disk cache & virtual memory learned habits as it went, but forgot all when the computer shut down.) I always partitioned clients' drives twice. First was a partition for swap and one for everything else. Then I installed a program that studied the history of which files were opened when, and studied the startup sequence. After two weeks, I used this information for a second partition and file placement.

Because it was based upon actual use, it greatly sped and smoothed disk activity. (The startup on an OS/2 computer was decreased from 2 minutes to 20 seconds.) Though it required work, it was done only once. Does one still have such control over placement of files and information about access & read times?

Also, because reliability was of utmost importance, all hardware & software I purchased was the previous versions: all the bugs had been fixed, and the price was reduced. I agree that none but hobbyists should buy the latest & greatest.

Regularly, I update the firmware on my optical drives, to speed writing or write to new media; but I've never 'flashed' an ATA drive. (Be sure you have an alternative boot device & ability to reinstall your old firmware.) Are you speaking of old PATAs or SCSIs, or of a newer drive type, such as SATA? (I'm about to buy a new PATA drive from one of the major companies.) Installing MacOSX 10.2 (not 10.1) allows all 120 GB to be addressed by it; and I'm hoping I can snuggle a Linux boot partition (in ext3) within such a Mac partition, after a Mac boot partition (in HFS+) with no difficulties. I would buy a larger drive if I knew there would be no difficulties.

Finally, I also agree with the person who wrote that computer companies have an obligation to let us know the content of a patch. MacOSX 10.4.1 flew on my wife's G3 iBook. By 10.4.9, she had to wait (and glare at me) as a new window opened. (Remember, Apple sells hardware.) I needed only bug fixes & security patches, not new frills. I've decided to give her a nice variety of Linux, a secure WiFi USB dongle, and let the Quicksilver handle its security, mailboxes, &c. That way she can bank over it and shop by web page.

In short, I agree with everyone and think these hardware topics are related.

 

Thanks for the link for Auslogics, just tried it on one machine and is now running on the other 4.Awesome! Thanks again!

 

speaking of OS's
what is the difference in the so called OEM,Pro,Ultimate version???
ive had Xp home OEM(which im back to running now) and XP Pro. i didnt see any difference in them other then when loggin on the one has a green thingy across the screen and the other has blue? i couldnt tell the difference in them? is it more software related with the one over the other? the only thing ive had a problem with the xp home oem is it doesnt have any SP's installed to it. so with my mobo now i need at least SP1 preinstalled to install it on my mobo. i dont know if thats the case with all mobo or not.. well it cant be cause ive let someone install my OEM version on his P35 mobo and it worked fine then i was able to install my mobo drivers and such once it had the SP update....

 

Gneiss1,

That is true to a degree. Microsoft does in fact put the OS software on the outer part of the disk, because the data density is higher at any given rpm, therefore faster to execute. Thats why you see some files that are unmovable. They also allocate a portion of the disk or partition for internal temporary files, expressly so the heads don't have to skip around as much accessing the data. It will also change the size of the temporary file to suit the need, up to the point that there is contiguous room for them, at which point they will be continued on another portion of the boot drive, linked to the original file. It can't readily move a partition because all the information is stored on a chip physically in the HDD. It can be done with software, but I would never recommend it because if something happens where the file that controls it becomes corrupt, you loose everything on the hard partition that is written partially to the soft one. It almost always results in the computer being unable to boot, or just crash. The files would become corrupted because of the loss of the algorithms and would no longer exist as far as the computer was concerned.

This was a large problem back when HDDs suddenly became so large that the operating system of the day could not recognize them. You had to install a file for the OS to be able to see it correctly. Everything was good unless something happened to that file. Then it became a nightmare, as you could no longer access your data. You generally couldn't even boot the computer again, once it crashed.

With AusLogics defrag, there is a setting where when you defrag, it moves all the system files to the beginning of the disk, and when it's defragged, it basically moves the calls to the exe files right to the head of the line to speed the computer bootup. Thank God those days are gone! LOL!!

Best Regards,
Russ

 

Shaff: Then just don't update the firmware. This is my mentality all along. I have no qualms about buying a Gen2 SSD, they're still far and away the best drives - just never update the firmware, the patches are even worse than Asus BIOSes!

 

firstly, TRIM is very important for an SSD or as the SSD is used the performance hit is alot.

secondly, never has a problem with any asus bios, inc modded ones.

 

I AGREE 100%!!!!!!!!