The question is rather simple isn`t it? However I ask it because I failed to success using Catalyst 4.4 so I decided that searching through all posible versions of catalyst inst a good ide and so I came here.. And although this doesn`t refer directly to Rivatuner - How it is posible to tell wether my Radeon is 128 - bit or 256 bit, thus should I try to get 9700 level from it or be satified with 9500pro..
All currently available drivers up to 4.4. If you've failed with this driver- read FAQ included in RT
hmm.. I read the manual.. kind of.. What I did was downloaded the original catalyst 4.4 and dearchiwed(unziped) in some specially created folder.. Then I run the RivaTuner and launched the according script for softmod win2kxp.. Then I found the file that was mentioned in manual.. (Interestingly that in drivers I had both files ati2mtag.sys and ati2mtag.sy_ in my drivers{I choosed file ati2mtag.sy_ }).. But no matter which file i selected the RT reported that this file ins`t supported(or something like that) and that thus that I may have problems if I choose to run script.. Well as I couldn`t find nothing about which drivers are supported in RT manual I decided to launch the scriopt anyway.. It reported that everything is ok however and the script has moded the drivers correctly.. But after I installed them and started in windows I was not even posible to launch 3DMark2001 (it reported that it doesn`t recognise my adapter or something like that{hence the driver problem}).. May be what I did wrong was installing those modified drivers over another instaled draivers.. (may be I had to remove the previous drivers first?) An yes another thing... There is a site http://www.ocfaq.com/softmod/ which offers allready [patched drivers with softmod.. the question is - Is this softmod diferent from softmod offered in RT, and may I get diferent results form each one of them? currently I use softmod offered in that site..
Windows File Protection did its' 'job'. RTFM, discussed in FAQ They cannot give different results. And it is discussed in FAQ too. RTFM.
if it is isn`t diferent sofware and is taken form your RT softmod then of course it cant give diferent results.. But if it is diferent sofware then I think it is perfectly posible to return diferent results.. I didn`t find in their page any reference to RT... I read the FAQ, but who knows may be I missed something.. How about those 256 - bit radeon 9500 - do they come with 256 bus by default or I see 256 bit bus only after the softmod?
The approach is different because it uses DevID swapping instead of DevID forcing added by RT. But it cannot give different result and I'm perfectly familiar with internals of both ways. BTW it ts explained in RivaTuner's FAQ. I won't repeat RTFM once agian so do it at last. The only thing that can cause differences is user. How do you imagine unlocking 256-bit bus on 128-bit board? It is technically impossible, and softmod has no effect on memory bus type at all.
>How do you imagine unlocking 256-bit bus on 128-bit board? It >is technically impossible, and softmod has no effect on memory >bus type at all. Well It is posible only if manufacturer has artificially disabled half of the bus.. Imagine something like dual-chanel memory subsystems available on for example n-Force.. It is perfectly posible to work with 2 memory modules either in dual-chanel mode or in single-chanel mode.. I just thought that probably just half of the chanells for 128-bit radeons could be artificially disabled, but physically available.. But I guess you are right.. Only - I have never seen any 256-bit Radeon 9500 for sale.. BTW thanks for information..
Disabling 'half of the bus on 256-bit board' (or more preciesly 2 of 4 available memory channels) will also disable half of memory chips installed on the PCB.
I of course trust you on word, but I would like to add that this may depend from arhitecture for example - If you have two 256Mb DDR memory modules on n-Force chispet you can still switch to single-chanel mode and still have accsess to all 512Mb of memory.. Of course it seems that in videocards arhitecture those chanels are more independent..
It's nice that you can make theories and know that nForce2 memory controller is able to run in both dual/single-channel mode, you can even forget that you've to reconfigure memory slots manually to switch from single to dual-channel mode, but how does it relate to R3x0 memory controller? Once again, it's technically impossible to 'unlock' 256-bit bus on 128-bit R3x0 unless you've early 64MB R9500 based upon R9700 PCB and you're able to solder missing 64MB on it. I won't repeat it anymore
You misunderstood me.. Of course I understand now and believe that it ir not posible to "unlock" any additional bus on 128-bit R9500`s once they: 1. phisically miss two memoryy chanels and 2. R300 architecture has independent memory chanels I just wanted to say that it is not so obvious and may depend on architecture and thus if you must know the architecture of R300 to know if I can "tune up" the 128-bit bus to 256-bit.. Of course you It seems to me you know R300 architecture pretty well so it may be obvious to you.. But For me it became so obvious only after you explained some details of R300 arhitecture... So I wrote previous posts not to show that I disagree with you, but just to show that my question was reasonable in my situation.. Thanks for your time and answers
