Intel Application Accelerator 2.3 is out...OFFICIAL

[SupermanCK]

Here you go

 

[Shagga]

Yeah using this now. Apparently it's solved the defaulting to 8mb cache bug, which has in turn boosted speed quite a bit by all accounts. I have not noticed any difference from 2.22 and I've not run any benchmarks...Thumbs up so far...

 

[sean2002]

cool find, thanks

 

[JuicyFruit]

Does anyone know if this will affect EAC and Lame? I had problems with the Accelerator that came with my P4B533-E.

And what does "defaulting to 8mb cache bug" mean?

Thanks.

 

[crapito]

thanks for the update info.

JuicyFruit: I don't know if this newer version will solve your problem, as I haven't installed it yet, but installing newer ASPI drivers allowed EAC to work fine with the older Intel App Acc on my computers.

 

[JuicyFruit]

I swear I did that. I forgot the link but I downloaded them from someplace (Nero.com maybe?).

Any benefits from installing this particular application accelerator?

 

[zCypher]

Why do you need an application accelerator? Going too slow..?

 

[labooboo]

I will not install the application accelerator if I am using Windows XP with Service pack one
really not that helpful

 

[JellyBaby]

Originally posted by: zCypher
Why do you need an application accelerator? Going too slow..?

Benchmarks are everything I guess.

 

[Nothinman]

Would be nice if Intel could just release a processor that worked well, instead of releasing software to cover up for it's bad design decisions. Seems a lot like VIA and their 4-in-1s.

 

[sean2002]

Do you even know what your talking about ??????

 

[CZroe]

Originally posted by: Nothinman
Would be nice if Intel could just release a processor that worked well, instead of releasing software to cover up for it's bad design decisions. Seems a lot like VIA and their 4-in-1s.

These are IDE driver enhancements, not "compatability fixes," and they are entirely optional (Unlike 4-in-1's).

Via is the king of compatability problems (Only because SiS is so small).

Intel has previously been known to set the standards that others attempt to clone (x86 ring a bell?). We practically only see "early adopter" compatability issues with their products.

If AMD were able to bring 64-bit chips to the consumer market with a 64-bit consumer version of XP earlier the tables would be turned, but it sounds more and more like they're just waiting for Intel to catch up.

 

[Nothinman]

These are IDE driver enhancements, not "compatability fixes," and they are entirely optional (Unlike 4-in-1's).

I realize they're not required. If they're IDE drivers, why aren't they labled as such?

Intel has previously been known to set the standards that others attempt to clone (x86 ring a bell?).

Just because it's in widespread use, doesn't mean it's good. i.e. McDonalds sells the most hamburgers, but they sure as hell don't make the best food.

If AMD were able to bring 64-bit chips to the consumer market with a 64-bit consumer version of XP earlier the tables would be turned,

Why? What need does any regular consumer for 64-bit computing? MMX has done fp math with 80-bit precision for years.

 

[ElDonAntonio]

Originally posted by: Nothinman

Why? What need does any regular consumer for 64-bit computing? MMX has done fp math with 80-bit precision for years.

I think MMX registers are integer registers and they're 64-bit wide. For FP, you probably mean SSE 1/2, and they're 128 bits wide.

 

[Nothinman]

I think MMX registers are integer registers and they're 64-bit wide. For FP, you probably mean SSE 1/2, and they're 128 bits wide.

I thought the MMX registers were 80-bit, but I'm no authority on it =)

 

[CZroe]

In reference to Intel Application Accellerator:

I realize they're not required. If they're IDE drivers, why aren't they labled as such?

Because WindowsXP's standard IDE drivers work fine. IAA, like Hypertheading doesn't help performance with all applications.

In reference to Intel's x86 standard:

Just because it's in widespread use, doesn't mean it's good. i.e. McDonalds sells the most hamburgers, but they sure as hell don't make the best food.

Not the point. The point is that others strive to be compatable with the market standard. Hamburgers do not work that way Of course it's not "good." It's freakin' ancient. That's what we get using a reverse-compatable architechture.

In reference to AMD's 64-bit standard:

Why? What need does any regular consumer for 64-bit computing? MMX has done fp math with 80-bit precision for years.

It has nothing to do with that. Ever heard of "Yamhill?" If AMD has their 64-bit standard thouroughly entrenched in the consumer market before Intel, then Intel will be in the same position regarding x86-64 as AMD was regarding x86. Intel would be making AMD clones. I'm NOT preaching the benefits of 64-bit computing. I'm talking about products, standards and clones of products in reference to compatability.

 

[Nothinman]

Of course it's not "good." It's freakin' ancient

Age has nothing to do with it. If it was old and it worked, why replace it? But there's a reason that a number of colleges teach assembly with MIPS instead of x86. Unix is ~20 years old and it's still in widespread use and is still magnitudes better than the alternatives in a lot of ways.

I'm talking about products, standards and clones of products in reference to compatability.

Which translates into marketing and the selling of products to people that have no idea what they're buying.

But x86-64 hasn't a chance unless some big company like Dell gets behind them and starts pushing their products. And pretty much Dell is all that's left now that Compaq has been bought by HP, I doubt anyone will notice if Gateway start selling their products and IBM has their own 64-bit processor line to push.

One good thing is that with Linux gaining popularity processor architecture is becoming less important. I can put Linux on x86, IA-64, Alpha, UltraSparc, Arm, MIPS, etc and it works exactly the same (well there's some difference in the bootloaders, but that's about it). At this point I'd like AMD to give Intel a run for their money in the 64-bit market, but no matter who comes out on top I'm sure all of my software will run.

 

[CZroe]

We all know that newer more efficient CPU architechtures exist. My 97MHz N64's SGI chip blew away any 233MHz Pentium MMX of the day for that reason. SGI workstations were not compatable w/ x86 PC software, so it made no difference. Attacking x86 for that reason makes no sense. So let's stop talking about that. The topic is not "Architechture vs. Architechture" or "Via vs. Intel" anyway

Oh and regarding Linux for multiplatform software:

Actually, nearly all Linux software I've ever been interested in enough to run is x86 only Some are compiled for other platforms but you can't just port Assembler code between totally different CPUs.

My hopes rest with Wine and WineX's eventual "completion" for Linux's multi-platform software explosion

 

[Nothinman]

Actually, nearly all Linux software I've ever been interested in enough to run is x86 only

I run, maybe, 2 applications that are binary only for x86 on my Linux box, all the rest have the source available and run just fine on non-x86 arches.

Some are compiled for other platforms but you can't just port Assembler code between totally different CPUs.

What applications are you using that are written with large amounts of asm?

My hopes rest with Wine and WineX's eventual "completion" for Linux's multi-platform software explosion

That's a poor place to put your hopes. Wine is a decent temporary solution, but depending on it for anything more than trivial apps and games is bad, we need native ports.

 

[manly]

Originally posted by: Nothinman

My hopes rest with Wine and WineX's eventual "completion" for Linux's multi-platform software explosion

That's a poor place to put your hopes. Wine is a decent temporary solution, but depending on it for anything more than trivial apps and games is bad, we need native ports.

I mostly agree. WINE will never actually be completed, and it's not very interesting unless you really need to run native Office/financial apps on the commercialized Crossover Office. In this sense, it's complete enough since the most important applications will run on Crossover Office (and others on Crossover Plugin). Don't get me wrong, I think a solid Win32 API library for *nix would be a good thing, but the Win32 API itself has proven to be more effort than it's really worth to reengineer.

As for WineX, I think (again for a small customer base) it's actually more useful than Wine. Linux will NEVER be a gaming platform, and we're lucky to see a handful of native ports of first tier games in any calendar year. John Carmack has already said any future id software ports to Linux will be charity jobs as there is NO viable commercial Linux gaming market, even if native ports were shipped on the same day as the Windows retail box. The choice for Linux users really is between virtually no commercial games or a decent number of WineX-enabled games.

As for the rest of the interesting thread, I totally disagree that AMD's late release of x86-64 has any meaningful effect on the PC desktop market, but I'll leave it to Nothinman to provide the proof argument.

 

[tart666]

So is this what the "Bus Master driver" called now?

 

[CZroe]

For a time, they remained seperate from the "Bus Master" drivers, and I believe they still are (There haven't been any recent BM driver updates).

 

[tart666]

My scores went up by 0.5%.

Does this mean Dell had these drivers pre-installed on my 4550 ?

 

[Chad]

These drivers waxed my Windows XP. Epox 4g4a with Intel P4 1.6 @ 2.4 good luck if you try them.

 

[JackBurton]

Would be nice if Intel could just release a processor that worked well, instead of releasing software to cover up for it's bad design decisions. Seems a lot like VIA and their 4-in-1s.

I'm going to ask what someone else asked, "do you even know what you are talking about?" Please list the "bad designs" of Intel processors... I just can't wait to here this. I also like the fact that you come on in here and slam Intel and don't even know what the software is or does.