Radeon 9700 non-Pro owners: 9700 Pro overclocking results with 7 different BIOS's

MuFu · Feb 17, 2003, 09:05 AM

[waffle]

Late last Summer, it became clear that the memory clock limit of R300-based boards would likely be primarily dictated by stability of the memory interface itself. Because of the complexity of the 256-bit design, TL or internal logic issues would cap operating speeds at about 330-340MHz.

Around the time of the 9700 Pro launch, a few people speculated that the soon-to-be-released 9700NP could easily be overclocked to 9700 Pro speeds, despite having slower-rated RAM. It made sense - the clock limit of the memory interface would come into play before the RAM started to throw up errors (what with yields of 3.6ns BGA being so good these days - 350MHz+ is often attainable). They even went so far as to suggest that even if AIBs started to push 9700NP's without an external power connector (which really isn't needed until about 300MHz), you could probably solder on a unit from an old floppy drive yourself, flash the BIOS and be ready to roll in "9700 Pro mode". A tantalising prospect for any modder.

But the prediction turned out to be wrong - most 9700NP and 9500NP users reached (and continue to reach) much poorer memory overclocks than those with Radeon 9700 Pros. Memory ratings aside, it would seem that the only thing the 9700P and 9700NP reference boards don't have in common is their BIOS. An area for investigation, surely! This issue has been touched on so many times in various forums, but I've never seen anybody systematically produce a set of data using different BIOS's to try and give some insight into it (probably because it takes bloody ages!). Please excuse me if this has been done before. Here's the juice...

[/waffle]

I have a Connect3D Radeon 9700 Pro (BBA, 942 PCB v3.0). Using the ATiFlash utility, I flashed and tested 6 of warp11's modified BIOS's, as well as the original BIOS that came with my card. A maximum core/memory overclock was established using the first 3DMark2001SE benchmark (doesn't matter how well it shows up artifacts because we're interested in *relative* clock limits only) and I also ran the benchmark in its entirety just out of curiosity. Clean install/clean boot, Athlon 2200+, 512MB RAM, DX9, Cat 3.1's, default driver settings.

[BIOS name] - [default core/mem] - [3DMarks@275/275MHz] - [Max stable core/mem O/C]
[BIOS info]

97nptx.bin - 270.0/263.3MHz - 11923 - 351.0/303.8MHz
R300 Infineon DDR BIOS P/N 113-94216-102 - 2002/09/23 15:04 (NTSC)

97npHerc.bin - 276.8/270MHz - 11887 - 351.0/303.8MHz
R300 Hynix 9700 BIOS - P/N 113-94223-101 - 2002/10/24 15:01 (NTSC)

97npWarp.bin - 276.8/270MHz - 11935 - 351.0/303.8MHz
R300 Hynix 9700 BIOS - P/N 113-94223-101 - 2002/10/24 15:01 (PAL)

95np128.bin - 276.8/270MHz - 10198 - 351.0/310.5MHz
R300 4P Hynix Memory BIOS - P/N 113-94228-100 - 2002/10/28 16:24 (NTSC)

97npClub.bin - 276.8/270MHz - 11858 - 351.0/324.0MHz
Radeon9700 Infineon DDR BIOS B8480.002 - 2002/12/16 21:39 (PAL)

97npSap.bin - 276.8/270MHz - 11914 - 351.0/324.0MHz
R300 DDR 275/270 PAL BIOS P/N 133-PC0000-13 - 002/10/31 15:30 (NTSC)

My original BIOS - 324.0/310.5MHz - 11895 - 351.0/351.0MHz
R300 Samsung DDR BIOS P/N 113-94206-102 - 2002/10/17 13:33 (PAL)

All checked and double checked. As you can see, with the exception of the Hierarchial-Z disabled/4-pipe 9500NP config, all the cards produce scores +/-100 of each other. The big difference is in the memory clockspeed limits, with the 9700 Pro BIOS enabling the card to go far beyond the others.

In the past there have been examples of BIOS's that have limited memory overclocking by altering timing/latency settings (e.g. some 8500LE BIOS's). Setting these more aggressively will only have a negligible effect on performance (due to the nature of framebuffer reads/writes) yet produce errors at a lower clockspeed. That's probably what is happening here. ATi knew that the 9700NP's would have no problem reaching Pro speeds from an analog point of view, so there was some kind of marketing call that dictated they be "retarded" through BIOS settings.

I suggest people with 9700NP's flash their cards with the Pro BIOS and see if it allows them to overclock their memory further. 3.6ns memory should really be able to reach at least 325MHz with no problems. There have been reports of 9500NPs with 2.86ns memory that haven't been able to break the 300MHz memory mark - surely that's a clear example of the BIOS holding things back. Maybe one of you BIOS editing guys can get creative and produce an "overclocking" BIOS for each configuration by splicing code. It'd be nice to have a non-Pro binary that sets a conservative default clock (275/275MHz) but has similar overclocking characteristics to the 9700 Pro BIOS (I guess this can be done using RadEdit).

Starting to pick up on the same things that have come to light in TriC's 9500 Pro BIOS and warp11's 9700NP BIOS threads now. What isn't really clear to me from those threads is whether people with Hynix RAM are getting artifacts etc when using the Infineon DDR BIOS's. Is that the case?

Cheers,

MuFu.

Skid · Feb 17, 2003, 11:54 AM

Fascinating experiment, thanks for sharing.

I've only tried three, the original unlocked 9500 n-p bios on my 128-meg Sapphire, the 97npSap and the 97npClub.

My lowly 3.3 ns Infineon has been stable up to 297 with all of them. The Club bios, which has an Infineon tag, lets me stay longer at 303, but eventually it will go belly up at that setting -- even with ramsinks and a fan blowing across them.

I may try your pro suggestion when I get some playtime. Meanwhile I'd love to hear from anyone else who has tried a 9700 Pro bios on a modded 9500.

I've been reluctant to try because I assumed the memory timing would be more aggressive, and because my card works so well with what I've got now ;>)

Vially · Feb 17, 2003, 11:54 AM

Very good stuff there, I never would have thought someone will do that, well, I was wrong

. Thank U !

There is one more question though: Where the heck do we find the BIOS-es for the Pro versions ?

I personally have a 9700np card, but I don't know if it's an ATI one or Sapphire, It was bulk. It's on red PCB with a square heatsink with an ATI logo. Anyone ?

MuFu · Feb 17, 2003, 12:01 PM

The BBA and red Sapphiretech boards are identical, don't worry about it.

Here is my BIOS if you fancy trying it. It's PAL ~Edit - no, it's NTSC~, with a default clockspeed of 325/310MHz. Should be ok to get into Windows from where you can reclock.

I am quite curious to see what the results will be.

MuFu.

Rene B · Feb 17, 2003, 01:48 PM

Just flashed your bios and now running at 346/319 completed 3Dmark01 without a hitch(12857) seems to run ok will test it with some gaming before going any further.

MuFu · Feb 17, 2003, 01:55 PM

Quote:

Originally posted by Rene B
Just flashed your bios and now running at 346/319...

Sounds encouraging! So you could reach 310MHz before, huh?

MuFu.

Vially · Feb 17, 2003, 02:05 PM

Have to find me a floppy here...

Rene B · Feb 17, 2003, 02:13 PM

Quote:

Originally posted by MuFu
Sounds encouraging! So you could reach 310MHz before, huh?

MuFu.

Yeah would start crapping out if I took it any higher than arround 310-315 before, gonna try and find max now

Doomtrooper · Feb 17, 2003, 02:13 PM

Mufu spent alot of time on this guys, alot of effort and good round of thankyou's are in order

Rene B · Feb 17, 2003, 03:17 PM

Ok here are the final(for today atleast) results from me, my card is now running at 348,5/331,4 without any issues at all (made it to a 3DMarks01 score of 13063), I tried 336.3 but that resutled in some minor artifacts so I took it back down to 331.4 for some more testing tomorrow...

Quote:

Mufu spent alot of time on this guys, alot of effort and good round of thankyou's are in order

How true, thank you for your hard work MuFu, because of guys like you we all get the most out of our hard earned money a BIG salute from Denmark to you

You get my vote for member of the week that's for sure

huwwatkins · Feb 17, 2003, 03:32 PM

I have a Sapphire 9700np that has 2.8samsung ram, I can only get 310mhz on it, ill try your bios as soon as i can

Huw

MuFu · Feb 17, 2003, 03:39 PM

Thanks for the feedback.

Quote:

Originally posted by Rene B
Ok here are the final(for today atleast) results from me, my card is now running at 348,5/331,4 without any issues at all (made it to a 3DMarks01 score of 13066).

That's pretty cool. Doomtrooper just informed me that he has a 9700NP running with my BIOS now at 330MHz memory clock also (up from 303MHz with the standard one). He has 2.86ns RAM though, so the result isn't that significant.

What memory do you have, Rene B?

MuFu.

P.S. Turns out my BIOS is NTSC, not PAL! No idea why or how since I am in the UK. I'll post some 275/275MHz default, PAL and NTSC versions in a few minutes once I've tested them.

Rene B · Feb 17, 2003, 04:34 PM

I got Hynix 3.6ns ram and a good airflow

MuFu · Feb 17, 2003, 05:01 PM

Quote:

Originally posted by Rene B
I got Hynix 3.6ns ram and a good airflow

Oh man, that's awesome - exactly the kind of result I wanted to see, cheers.

MuFu.

huwwatkins · Feb 17, 2003, 05:43 PM

Bios allowed me to increase mem speed from 310 to 317, not much, but an increase non the less.

Update: memory type acrifacting occurs about half way through the nature test.

Huw

MuFu · Feb 17, 2003, 05:55 PM

Quote:

Originally posted by huwwatkins
Update: memory type acrifacting occurs about half way through the nature test.

At 317MHz? What do you see?

Strange - I would have bet good money that you'd be able to push 330MHz with the Pro BIOS. :-\

MuFu.

thagame24 · Feb 17, 2003, 06:05 PM

this bios actually gives me WORSE results, it artifacts on anythin over 320 core 305 memory, i was runnin warp's perfect at 330/307.

huwwatkins · Feb 17, 2003, 06:06 PM

at 317 i see large tearing, sort of objects that appear that shoudn't. BUT the fact that it only started articating halfway through the demo would suggest that its a heat related problem.

Artifacting doesn't occur during the lobby demo on 3dmark 2001, only nature

It had been noted that there was an additional resistor on the back side of the card where the GPU is located, it seems that this is also another difference between the 9700np and 9700pro, although we don't know what it does

see thread:http://www.rage3d.com/board/showthre...post1331708106

MuFu · Feb 17, 2003, 06:10 PM

Quote:

Originally posted by thagame24
this bios actually gives me WORSE results, it artifacts on anythin over 320 core 305 memory, i was runnin warp's perfect at 330/307.

Bizarre that the core overclock has been affected. What memory do you have?

MuFu.

MuFu · Feb 17, 2003, 06:11 PM

Quote:

Originally posted by huwwatkins
at 317 i see large tearing, sort of objects that appear that It had been noted that there was an additional resistor on the back side of the card where the GPU is located, it seems that this is also another difference between the 9700np and 9700pro, although we don't know what it does see thread:http://www.rage3d.com/board/showthre...post1331708106

Thanks, I'll check it out.

MuFu.

thagame24 · Feb 17, 2003, 06:12 PM

samsung 2.8 ns

freshbrain · Feb 17, 2003, 07:03 PM

I am not sure if i got it right, please help me get it clearer.

freshbrain · Feb 17, 2003, 07:05 PM

i have a 9500NP working as a 9700, Hynix 3.6 memory here, so, should I flash it to the 9700Pro bios, or is it only for 'original' 9700s?

MuFu · Feb 17, 2003, 07:08 PM

Quote:

Originally posted by freshbrain
i have a 9500NP working as a 9700, Hynix 3.6 memory here, so, should I flash it to the 9700Pro bios, or is it only for 'original' 9700s?

How far can you overclock at the moment?

Try the Pro BIOS I posted first. The whole point of this thread is to establish whether the BIOS's on 9700NP's are holding back memory overclocking.

MuFu.

freshbrain · Feb 17, 2003, 08:20 PM

Works great for me!!!, thanks MuFu, great job. I could get to 335 with the 9500NP bios, now 351 with no artifacts. 17123 3dmarks at 410/351 with my voltmodded 9500np and nforce2

Razor6 · Feb 17, 2003, 09:16 PM

What would happen if I flashed the 9700 pro bios to my 9500 np that has not been modded. It doesn't work as a 9700 so I'm only looking at possibly better memory overclocking. As it is, it gets to 378/290, which is really good for the core but crap for the memory seeing as how it's rated for 303. (3.3ns Infineon)

exesspeed · Feb 18, 2003, 01:05 AM

i used warps herculees modded bios for the 9700pro on my non pro herc 9700 see my post by exesspeed overclocks like hell had core to 400 no probs but mem wont go higher than 340 without artifacts , tearing so its on the limit, just wanted to see how far it would go its runnin at 425 / 310 now been on the net for 5 hours playin ghost recon no probs also completed unreal 2 the awakening in 2 days nprobs at all and the heat sinks are cool , im a happy chappy ,thanks warp

TrickerZ · Feb 18, 2003, 01:56 AM

I got a Saphire 9500 I'm looking to try and get a 9700 pro out of. I got the big ol' Zalman heatpipe on the core and can overclock it like nothing, but my memory doesn't do quite as well, will my card run the 9700 pro bios ok with the resistor modded? I want to make sure so I don't boot and can't do anything. And also, how do I save my bios incase it doesn't work well?

exesspeed · Feb 18, 2003, 02:16 AM

mmmmmmmmmmm dont know about the resistor mod, better do some home work on it firsr to be safe,

huwwatkins · Feb 18, 2003, 07:49 AM

That resistor on the back of the Sapphire non pro's is a real mystery, for one, becuase its only only the Sapphire np's, other np's don't have it. However i'm not brave enough to remove it without having some idea of what it does.

Huw

Feb 17, 2003, 09:05 AM	#1	Advertisement (Guests Only) Login or Register to remove this ad
MuFu should be at a lecture Join Date: Mar 2002 Location: UK Posts: 1,476	Radeon 9700 non-Pro owners: 9700 Pro overclocking results with 7 different BIOS's [waffle] Late last Summer, it became clear that the memory clock limit of R300-based boards would likely be primarily dictated by stability of the memory interface itself. Because of the complexity of the 256-bit design, TL or internal logic issues would cap operating speeds at about 330-340MHz. Around the time of the 9700 Pro launch, a few people speculated that the soon-to-be-released 9700NP could easily be overclocked to 9700 Pro speeds, despite having slower-rated RAM. It made sense - the clock limit of the memory interface would come into play before the RAM started to throw up errors (what with yields of 3.6ns BGA being so good these days - 350MHz+ is often attainable). They even went so far as to suggest that even if AIBs started to push 9700NP's without an external power connector (which really isn't needed until about 300MHz), you could probably solder on a unit from an old floppy drive yourself, flash the BIOS and be ready to roll in "9700 Pro mode". A tantalising prospect for any modder. But the prediction turned out to be wrong - most 9700NP and 9500NP users reached (and continue to reach) much poorer memory overclocks than those with Radeon 9700 Pros. Memory ratings aside, it would seem that the only thing the 9700P and 9700NP reference boards don't have in common is their BIOS. An area for investigation, surely! This issue has been touched on so many times in various forums, but I've never seen anybody systematically produce a set of data using different BIOS's to try and give some insight into it (probably because it takes bloody ages!). Please excuse me if this has been done before. Here's the juice... [/waffle] I have a Connect3D Radeon 9700 Pro (BBA, 942 PCB v3.0). Using the ATiFlash utility, I flashed and tested 6 of warp11's modified BIOS's, as well as the original BIOS that came with my card. A maximum core/memory overclock was established using the first 3DMark2001SE benchmark (doesn't matter how well it shows up artifacts because we're interested in relative clock limits only) and I also ran the benchmark in its entirety just out of curiosity. Clean install/clean boot, Athlon 2200+, 512MB RAM, DX9, Cat 3.1's, default driver settings. [BIOS name] - [default core/mem] - [3DMarks@275/275MHz] - [Max stable core/mem O/C] [BIOS info] 97nptx.bin - 270.0/263.3MHz - 11923 - 351.0/303.8MHz R300 Infineon DDR BIOS P/N 113-94216-102 - 2002/09/23 15:04 (NTSC) 97npHerc.bin - 276.8/270MHz - 11887 - 351.0/303.8MHz R300 Hynix 9700 BIOS - P/N 113-94223-101 - 2002/10/24 15:01 (NTSC) 97npWarp.bin - 276.8/270MHz - 11935 - 351.0/303.8MHz R300 Hynix 9700 BIOS - P/N 113-94223-101 - 2002/10/24 15:01 (PAL) 95np128.bin - 276.8/270MHz - 10198 - 351.0/310.5MHz R300 4P Hynix Memory BIOS - P/N 113-94228-100 - 2002/10/28 16:24 (NTSC) 97npClub.bin - 276.8/270MHz - 11858 - 351.0/324.0MHz Radeon9700 Infineon DDR BIOS B8480.002 - 2002/12/16 21:39 (PAL) 97npSap.bin - 276.8/270MHz - 11914 - 351.0/324.0MHz R300 DDR 275/270 PAL BIOS P/N 133-PC0000-13 - 002/10/31 15:30 (NTSC) My original BIOS - 324.0/310.5MHz - 11895 - 351.0/351.0MHz R300 Samsung DDR BIOS P/N 113-94206-102 - 2002/10/17 13:33 (PAL) All checked and double checked. As you can see, with the exception of the Hierarchial-Z disabled/4-pipe 9500NP config, all the cards produce scores +/-100 of each other. The big difference is in the memory clockspeed limits, with the 9700 Pro BIOS enabling the card to go far beyond the others. In the past there have been examples of BIOS's that have limited memory overclocking by altering timing/latency settings (e.g. some 8500LE BIOS's). Setting these more aggressively will only have a negligible effect on performance (due to the nature of framebuffer reads/writes) yet produce errors at a lower clockspeed. That's probably what is happening here. ATi knew that the 9700NP's would have no problem reaching Pro speeds from an analog point of view, so there was some kind of marketing call that dictated they be "retarded" through BIOS settings. I suggest people with 9700NP's flash their cards with the Pro BIOS and see if it allows them to overclock their memory further. 3.6ns memory should really be able to reach at least 325MHz with no problems. There have been reports of 9500NPs with 2.86ns memory that haven't been able to break the 300MHz memory mark - surely that's a clear example of the BIOS holding things back. Maybe one of you BIOS editing guys can get creative and produce an "overclocking" BIOS for each configuration by splicing code. It'd be nice to have a non-Pro binary that sets a conservative default clock (275/275MHz) but has similar overclocking characteristics to the 9700 Pro BIOS (I guess this can be done using RadEdit). Starting to pick up on the same things that have come to light in TriC's 9500 Pro BIOS and warp11's 9700NP BIOS threads now. What isn't really clear to me from those threads is whether people with Hynix RAM are getting artifacts etc when using the Infineon DDR BIOS's. Is that the case? Cheers, MuFu. Last edited by MuFu : Feb 20, 2004 at 12:12 PM.

Feb 17, 2003, 12:01 PM	#4
MuFu should be at a lecture Join Date: Mar 2002 Location: UK Posts: 1,476	The BBA and red Sapphiretech boards are identical, don't worry about it. Here is my BIOS if you fancy trying it. It's PAL ~Edit - no, it's NTSC~, with a default clockspeed of 325/310MHz. Should be ok to get into Windows from where you can reclock. I am quite curious to see what the results will be. MuFu. Last edited by MuFu : Feb 17, 2003 at 03:50 PM.

Feb 17, 2003, 01:48 PM	#5
Rene B Radeon HD 4850 Join Date: Oct 2000 Location: Svendborg, Denmark Posts: 201	Just flashed your bios and now running at 346/319 completed 3Dmark01 without a hitch(12857) seems to run ok will test it with some gaming before going any further. __________________ Leftie: AMD Athlon 2500+ Barton, Abit NF7-S mb, Radeon9700np, 512 MB PC3500, 2x80Gb SATA Barracuda's, Antec TP 480w, Iiyama WM Pro 454, SuSE Pro 9.3 Laxmi: P III 633MHz , Radeon 7000, 512MB PC133, Lotsa diskspace, Debian Linux 2.6.6

Feb 17, 2003, 03:32 PM	#11
huwwatkins Radeon HD 4850 Join Date: Jan 2002 Location: UK Posts: 267	I have a Sapphire 9700np that has 2.8samsung ram, I can only get 310mhz on it, ill try your bios as soon as i can Huw __________________ 2400+ @ 2700+ Epox 8RDA+ (nForce2) WD800JB - 8 mb cache Coolermaster HHC-L61 HeatPipe Radeon 9700 128DDR 512mb Twinmoss PC3200 WD800JB - 8 mb cache Mitsubishi Diamond Pro 740 SB

Feb 17, 2003, 04:34 PM	#13
Rene B Radeon HD 4850 Join Date: Oct 2000 Location: Svendborg, Denmark Posts: 201	I got Hynix 3.6ns ram and a good airflow __________________ Leftie: AMD Athlon 2500+ Barton, Abit NF7-S mb, Radeon9700np, 512 MB PC3500, 2x80Gb SATA Barracuda's, Antec TP 480w, Iiyama WM Pro 454, SuSE Pro 9.3 Laxmi: P III 633MHz , Radeon 7000, 512MB PC133, Lotsa diskspace, Debian Linux 2.6.6

Feb 17, 2003, 11:54 AM	#2
Skid Radeon HD 4830 Join Date: Apr 2002 Posts: 157	Fascinating experiment, thanks for sharing. I've only tried three, the original unlocked 9500 n-p bios on my 128-meg Sapphire, the 97npSap and the 97npClub. My lowly 3.3 ns Infineon has been stable up to 297 with all of them. The Club bios, which has an Infineon tag, lets me stay longer at 303, but eventually it will go belly up at that setting -- even with ramsinks and a fan blowing across them. I may try your pro suggestion when I get some playtime. Meanwhile I'd love to hear from anyone else who has tried a 9700 Pro bios on a modded 9500. I've been reluctant to try because I assumed the memory timing would be more aggressive, and because my card works so well with what I've got now ;>)

Feb 17, 2003, 11:54 AM	#3
Vially Radeon HD 4600 Join Date: Jan 2003 Posts: 87	Very good stuff there, I never would have thought someone will do that, well, I was wrong . Thank U ! There is one more question though: Where the heck do we find the BIOS-es for the Pro versions ? I personally have a 9700np card, but I don't know if it's an ATI one or Sapphire, It was bulk. It's on red PCB with a square heatsink with an ATI logo. Anyone ?

Feb 17, 2003, 02:05 PM	#7
Vially Radeon HD 4600 Join Date: Jan 2003 Posts: 87	Have to find me a floppy here...

Feb 17, 2003, 02:13 PM	#9
Doomtrooper Rage3D Veteran Join Date: Aug 2000 Location: The Great White North Posts: 5,340	Mufu spent alot of time on this guys, alot of effort and good round of thankyou's are in order

Feb 17, 2003, 05:43 PM	#15
huwwatkins Radeon HD 4850 Join Date: Jan 2002 Location: UK Posts: 267	Bios allowed me to increase mem speed from 310 to 317, not much, but an increase non the less. Update: memory type acrifacting occurs about half way through the nature test. Huw __________________ 2400+ @ 2700+ Epox 8RDA+ (nForce2) WD800JB - 8 mb cache Coolermaster HHC-L61 HeatPipe Radeon 9700 128DDR 512mb Twinmoss PC3200 WD800JB - 8 mb cache Mitsubishi Diamond Pro 740 SB

Feb 17, 2003, 06:05 PM	#17
thagame24 Radeon HD 4600 Join Date: Jan 2002 Posts: 51	this bios actually gives me WORSE results, it artifacts on anythin over 320 core 305 memory, i was runnin warp's perfect at 330/307. __________________ Athlon XP Barton 2500+ @ 3200+ Abit NF7 mobo Windows XP Professional SP1 Sapphire Radeon 9700 np @ 340/290 Maxtor 80 gig 7200 rpm Samsung 160 gig 7200 rpm 8 mb cache 512 MB corsair xms LL3200 DDR

Feb 17, 2003, 06:06 PM	#18
huwwatkins Radeon HD 4850 Join Date: Jan 2002 Location: UK Posts: 267	at 317 i see large tearing, sort of objects that appear that shoudn't. BUT the fact that it only started articating halfway through the demo would suggest that its a heat related problem. Artifacting doesn't occur during the lobby demo on 3dmark 2001, only nature It had been noted that there was an additional resistor on the back side of the card where the GPU is located, it seems that this is also another difference between the 9700np and 9700pro, although we don't know what it does see thread:http://www.rage3d.com/board/showthre...post1331708106 __________________ 2400+ @ 2700+ Epox 8RDA+ (nForce2) WD800JB - 8 mb cache Coolermaster HHC-L61 HeatPipe Radeon 9700 128DDR 512mb Twinmoss PC3200 WD800JB - 8 mb cache Mitsubishi Diamond Pro 740 SB Last edited by huwwatkins : Feb 17, 2003 at 06:10 PM.

Feb 17, 2003, 06:12 PM	#21
thagame24 Radeon HD 4600 Join Date: Jan 2002 Posts: 51	samsung 2.8 ns __________________ Athlon XP Barton 2500+ @ 3200+ Abit NF7 mobo Windows XP Professional SP1 Sapphire Radeon 9700 np @ 340/290 Maxtor 80 gig 7200 rpm Samsung 160 gig 7200 rpm 8 mb cache 512 MB corsair xms LL3200 DDR

Feb 17, 2003, 07:03 PM	#22
freshbrain Radeon HD 4550 Join Date: Feb 2003 Posts: 48	I am not sure if i got it right, please help me get it clearer. __________________ DFI P35 blood iron Intel e2140 @ 3.3Ghz 2x1GB Crucial ballistix ATI hd3850

Feb 17, 2003, 07:05 PM	#23
freshbrain Radeon HD 4550 Join Date: Feb 2003 Posts: 48	i have a 9500NP working as a 9700, Hynix 3.6 memory here, so, should I flash it to the 9700Pro bios, or is it only for 'original' 9700s? __________________ DFI P35 blood iron Intel e2140 @ 3.3Ghz 2x1GB Crucial ballistix ATI hd3850

Feb 17, 2003, 08:20 PM	#25
freshbrain Radeon HD 4550 Join Date: Feb 2003 Posts: 48	Works great for me!!!, thanks MuFu, great job. I could get to 335 with the 9500NP bios, now 351 with no artifacts. 17123 3dmarks at 410/351 with my voltmodded 9500np and nforce2 __________________ DFI P35 blood iron Intel e2140 @ 3.3Ghz 2x1GB Crucial ballistix ATI hd3850

Feb 17, 2003, 09:16 PM	#26
Razor6 Radeon HD 4890 Join Date: Feb 2001 Location: Houston, TX Posts: 938	What would happen if I flashed the 9700 pro bios to my 9500 np that has not been modded. It doesn't work as a 9700 so I'm only looking at possibly better memory overclocking. As it is, it gets to 378/290, which is really good for the core but crap for the memory seeing as how it's rated for 303. (3.3ns Infineon)

Feb 18, 2003, 01:05 AM	#27
exesspeed Newbie Join Date: Feb 2003 Location: uk Posts: 5	i used warps herculees modded bios for the 9700pro on my non pro herc 9700 see my post by exesspeed overclocks like hell had core to 400 no probs but mem wont go higher than 340 without artifacts , tearing so its on the limit, just wanted to see how far it would go its runnin at 425 / 310 now been on the net for 5 hours playin ghost recon no probs also completed unreal 2 the awakening in 2 days nprobs at all and the heat sinks are cool , im a happy chappy ,thanks warp

Feb 18, 2003, 01:56 AM	#28
TrickerZ Radeon HD 4350 Join Date: Sep 2000 Location: Florida Posts: 23	Saphire 9500 -> 9700pro I got a Saphire 9500 I'm looking to try and get a 9700 pro out of. I got the big ol' Zalman heatpipe on the core and can overclock it like nothing, but my memory doesn't do quite as well, will my card run the 9700 pro bios ok with the resistor modded? I want to make sure so I don't boot and can't do anything. And also, how do I save my bios incase it doesn't work well?

Feb 18, 2003, 02:16 AM	#29
exesspeed Newbie Join Date: Feb 2003 Location: uk Posts: 5	mmmmmmmmmmm dont know about the resistor mod, better do some home work on it firsr to be safe,

Feb 18, 2003, 07:49 AM	#30
huwwatkins Radeon HD 4850 Join Date: Jan 2002 Location: UK Posts: 267	That resistor on the back of the Sapphire non pro's is a real mystery, for one, becuase its only only the Sapphire np's, other np's don't have it. However i'm not brave enough to remove it without having some idea of what it does. Huw __________________ 2400+ @ 2700+ Epox 8RDA+ (nForce2) WD800JB - 8 mb cache Coolermaster HHC-L61 HeatPipe Radeon 9700 128DDR 512mb Twinmoss PC3200 WD800JB - 8 mb cache Mitsubishi Diamond Pro 740 SB

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode