If you are the lucky owner of one of these Intel motherboard chipsets, you can now install the Intel Application Accelerator 2.0, a new performance software package for Intel-based desktop PCs. This new version brings the following enhancements: faster boot time, accelerated disk I/O for games, graphics Applications, disk utilities, and edia authoring applications, performance-enhancing data pre-fetcher for Intel Pentium 4 processor-based systems, support for 137+ GB IDE hard drives. Check a benchmark here. The download supports Windows NT4/2000, Windows 98/SE/Me & WindowsXP Home/Pro.
Intel Application Accelerator 2.0 Released
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I want I want. But is there an AMD equivalent?
No. If there was one, VIA should have supplied the utility, not AMD. It is a chipset enhancement, not a CPU one.
I was discussing that with my husband over dinner tonight, and he is thinking that the “trick” that Intel may be using in the utility, is a new (unused yet by OS developers) feature found on the i8xx series of Intel chipsets. So, the trick is that the utility “talks” to the hard drive directly through a special API, without going through the PCI bus, which is the “traditional” way of doing things. Since the i8xx series of chipsets came out, they support such direct communication with the IDE, which makes the IDE hard drives fly on any PII/Celeron/PIII/P4 machine that features http://support.intel.com/support/chipsets/iaa/suppchip.htm“>one chipsets (and in conjuction with the Pentium 4’s new set of MTRRs enabled with that utility, (if you have a Pentium4) the gain is even bigger)!
Older Intel chipsets do not support that feature. And I think none of the VIA/SiS chipsets support that either.
Personally, I am still with my Abit BP6 (dual Celeron 533 Mhz) which employees the older BX chipset, so I can’t use that utility either.
Ok, I found some more information of what it is actually doing on Anandtech:
“IAA enables some software level prefetching that looks at data access patterns and prefetches data from the hard disk to main memory. This speeds up functions such as compressing files and even loading levels in a game (Intel gave Quake III Arena as an example). These speed benefits are offered only for IDE drives running off of the chipset’s I/O Controller Hub.
Boot time is also decreased but currently the majority of the performance boost comes under Windows 2000. Intel did inform that upon the final release of Windows XP we should see a more tuned version to work with XP’s inherent prefetching mechanisms to offer a performance improvement there as well.
In terms of compatibility, the IAA enables support for 48-bit LBA enabling support for ATA drives larger than 137GB. The final feature of IAA is that it will be the driver used to support Serial ATA devices.”
I am using Windows XP Pro on P4 2.0 256 RD RAM, 850 chipset, and I installed this patch, the speed I gained is incredible!! It does boost performance indeed!
—————–
Eugenia wrote:
Boot time is also decreased but currently the majority of the performance boost comes under Windows 2000.
—————–
In that case, imagine the speed that I will get when the one which is designed for WinXP only, gets released!
I already gained good performance increase with the current one! I will get even more speed with the next one
With all these antitrust rulings in favor of the states I was begining to worry about the wintel dynasty. It’s hard to really blame Intel, they know if they release a patch for Windows they’ve covered >90% of their market, but at the same time it’s still discouraging for alternative OS users.
Goody, I have one of those chips, must try this out. Has anyone tried this yet with a intel 810, I’m curious what this will do for the graphics chip. I’ve found XP pro runs quite well with my 500 celeron which is down near the base requirements for XP (not that i really follow requirements). Can’t imagine it booting much faster, it boots in about 25ish seconds. But I have found that it suffers sometimes coming out of sleep mode trying to draw everything. sometimes it takes about 15 seconds for the deskbar to come back to life. Maybe this will help. Course some extra ram might be good to, 128 just isn’t as much as it used to be.
Well that was less than steller, It now takes about a 50 seconds from a cold start to be in windows, and from the time windows starts to boot about 35-40 seconds. loss in time
. I rather enjoyed when I first installed XP how it booted slightly faster then BeOS did, I was amazed. Guess that’s no more. And so far windows still sometimes are suffering when trying to draw 
. Oh well, still will take it anyday over linux were you can get up go the bathroom and get some coffee before it even gets to where it ask for a password. Maybe there will be some speed gain someplace and i’m just not seeing it yet.
Damn…lets see if it works on the ‘ol Bluelight PC ….lol
Dano
So I don’t care about that
booting my system (with SCSI bios enabled)
took me something BEFORE loading any OS.
Anyway IF that trick is just use some chipset feature I hope that some os may use it too.
Is there any logic behind all those hardware manufacturer hiding the specs for their chips? I mean think about it, open up the specs and you will have like 100 times more developpers working on developing drivers and support for your hardware, you will sell more.
The only reason I see for this is that they want to keep things working in Windows only. I doubt that hiding their specs save their edge on the competition, and if that’s the problem, why don’t they release the specs after a period of 6 months? At the speed things goes nowadays that should be more than enough for the competitor to be face-to-face anyway.
Come on Intel, don’t be lame, put out the algo you used and let the Linux kernel hackers port that boost.
I am sorry to hear that Brad. However on my p4 system as I described in my previous post above, I really DO get performance increase. I am not sure why it didn’t happen to your computer. May be it works differently on different chipsets.
And when it comes to AMD users, well, I don’t think AMD will give up this easily. As it was said above, this has nothing to do with AMD, but it might have to do with AMD in the near future.
When I bought my computer, I was gonna buy Athlon 1.3 but it was not available in the computer store so that’s why I went with P4 2.0. I really wanted to have the fastest computer possible. My very first computer was AMD 133 MHz which was VERY slow. The second one was Celeron 433 which was again slow, and for the third one (yep, I got a job and I collected money) I said, this time, I am gonna get the fastest one I could get! Athlon 1.3 was not available in the store and I was in-patient to wait any longer so I went with P4 2.0 256MB RDRAM. But now, I think I am glad I went with P4
.
Dave,
I think there is some reason for it. 1) like you mentioned they don’t want the compitition to see how they do things, even though there’s a good chance it’s very similar. The big one I think is this. They don’t want 100’s of people tweaking with it. They want the only drivers out there to be theirs. If someone starts hacking a driver and it might be good at something but has a problem theirs doesn’t and people use them for OS’s not supported by that company the people who use those drivers and have issues may very well look down at the company who made the hardware. Maybe they don’t relize it’s not a factory driver, maybe they just want to blame someone. When a company has control of their drivers they can have better quality control cause they know whats out there. I don’t think many companies would be found of people messing with their stuff. Actully if you look at most products in the world from your car to your computer to you whatever companies tend to make their stuff so it’s hard to mess with to avoid people messing with it then compaining. Yes it’s the persons fault but to many people would tweak and then blame the company. Also it can come down to the code is something the company paid to create, and put many hours in and money. Thats not something they would just want to give away. Yes the open driver would most likely help sales, but i think they would rather market a driver for something themselves rather than have bob’s universal video driver hack getting their stuff to quasi work on something freaky.
my dream would be a driver less world were there is a standard hardware to OS interface for each type of component. the hardware maker just has to make there stuff hook up to the set universal and set standard open protocalls and the OS maker only has to have the other end of the protocol written once and every brand /model / whatever of that type of componet works. I guess USB is sorta like this. Imagine if when writting an OS you only had to write basicly one driver type thing and never have to do anything thing else and all the worlds hardware worked. Probly a few dozen of these protocalls (SP?) would handle every class of hardware there is, HD, CPU, Video, sound, printer … People could switch OS’s with out every worring about compatibilty. well I can dream.
If I had a time machine and could follow through on my dream to travel through time and club on the head people as they were about to come up with a stupid idea, I think the guy who thought up drivers as they are would be high on the list.
…it’s called the driver and it’s pretty much different for every OS there is. I’m sure you meant a standard that the hardware would conform to. While this can be good in some respects, you really can lose alot due to baggage, sticking with this interface. Just look at the Intel architecture, while they’ve been able to do phenomenal things with the architecture, it’s still held back in a lot of ways but it’s nearly 20 year old ISA. But they don’t ditch it because of compatability. I think it’s better to have a very good driver model in an OS then to have a strict hardware interface that the hardware must conform to.. Try using VESA or VGA graphics sometime….it’s ugly and slow.
-James
Installed but when I start the Intel application accelerator utility it says that the driver is not deteced.
I have a dell notebook, C810 is the model number and I think it should be supported with XP Pro.
Let me know
Intel should release something for linux too
1 If an opensource driver has issues they will e fixed as soon as they appear.
2 If an opensource implementation anybody can see what is going on and find out that who is to blame.
3 Making an opensource drive doesn’t mean you are about to lose your IP since apply rev en on a closed source driver, or an offuscated source driver or a open source driver for steal tecnology is in any case a copyright infringement.
4 Opensource driver can boost the developement of it since is possible to find more willing testers and is possible to collect bug request and fix proposals in a faster way.
What’s wrong with open source driver?
1 If they are done with lesser HW spec they may have a slow start.
2 They may be released as alpha quality and used as complete release (don’t blame the developers, blame the user), but that may appen and apply to the closed source drivers (ATi weekly drivers leack? Not completely working Detornator from nVidia?).
Anyway an Open source user has more money to spend on High quality HW since he has not to spend much for software (I pay for the commercial games I like but not for the OS or the Office automation programs, bet I can buy more RAM a better CPU and a good GPU (as I did)).
Any counterpoint?
I hope they do release something for linux. I tried their
compiler for linux (icc) and it was pretty cool.
First of all Christian, make sure you already have that .inf file that Intel requires before you install the driver. Read the details on their web site. If you do not have that .inf file already, the driver won’t install.
As for hiding specs, I do not think that this is true. The specs are available, freely or by request. Problem is, this thing is pretty new, so OS manufactures have not get their brain around it yet, or it was a low priority for them. Maybe we see that by default in every OS in the next few years.
I installed this thing on my Sony VAIO GR114EK and it gave me much better HD response times with my installation of W2K Professional. Things such as hibernating, restarting etc were all much quicker than before. I only wish I had done some HDTach benchmarks before to give you guys some figures but I was too impatient
This deserves two thumbs up to Intel from me!
~Tangy
I’ve used the chipset installation utility before .. its setup said I had all the .inf files up to date anyway the intel acceleration utility installs with no evident error and then the utility says no drvier detected.
If you succed installing it ona dell c810 please tell me
thanks
I was wondering does anyone know what chipset the 5445 pavilion uses i want to install the accelerator on my laptop but im not sure if it is compatible?
Chris,
Well if your in windows go into the control panel/system/devise manager then check out system devises, it will tell you your chips. Hope that helps
Where is the chipset in the device manager?
there will be a bunch of numbers in there probly start with 80 then a 3 more numbers, you will see a pattern in there with the same number, thats your chipset. i think it tells you like northbridge southbridge, video, IO. I don’t know what CPU you have so i can’t get you in the ball park.
Hey brad can u email me please its [email protected] i have a p3 1.1ghz