Great benchs Muzz...
--fibo
Printable View
Great benchs Muzz...
--fibo
Thx fibo... just trying to dispell what I feel to be rumours is all......
The card kept freezing up at over 500 for whatever reason, it took me forever to figuer out it was the card:D It still can do about 508 solid, but not on the NF7 at 217 FSB:DQuote:
Originally posted by Xenogias
OPP, whats the deal with the Radeon. I seem to remember you running over 500mhz core and much higher on the mem than 358. Of course that was on Intel which may make a difference. Is it a different card? If not--why don't you push that Raddy some more? break 22k?
a little off topic, but while I'm at it--what tweak program/method do you use to OC your card
As far as the memory, it just gets worse and worse, I've had this card for while now, I'm running her down to the ground.
OPP
on my tests I did use the lower multis
On my 8k3a I got around 220~30 mrks for every 10mhz+ of fsb with my gf4 , with a ati9700 the scalling should be higher
for every 100mhz of cpu i´d get around 200mrks this still with a gf4
these are my benchs using single channel
(1833mhz)
11x166 12184mrks (8k3a 11x166 12080mrks)
10,5x174 12246mrks
10x183 12230mrks (double checked)
9,5x192 12329mrks
9x200 (1800mhz) 12399mrks
215mrks from 33mhz+ of fsb isnt something I should be happy about
9x200 (tweaked w/ wpcredit) 12549mrks
365mrks from 33mhz+ of fsb
maybe single channel produces dif results , I dunno when I have another good stick of ram ill test it again
Muzz
With all due respect saying something like this is a rumor must be from someone who owns an EPox for the first time ;)
which bios are you using as I want to try it?
just another thought you might thing that you are getting a good linear scalling but you might consider comparing 8rda+ at default with abits nforce2 or Asus , as from what i´ve seen around forums and review sites , both Abit and Asus are faster
so Abit mobos have always been faster for 3dmark and I dont think 8rda+ will be a first , but who knows
I said rumour as in this incarnation bro, I also stated what I have heard about previous boards, so as far as that goes............
I use the c10 bios, it seems to work well for me.
The under10 multi problem has been noted by alot of folks( luckily I'll never need that ;) ) , and I think that it may really be an NF2 problem, as I have seen some benches at lower than 10 at high fsb(EDIT: from other boards besides the epox), that just are not that great.
I just wanted you to know Skatan that this was not meant to be an attack on you, as I have stated it just seems that Epox is getting bashed, and from what I hear there may be a reason for that, as you are not the first person to bring it up....... that I can tell you for certain.
Hope there are no hard feelings.
muzz
I have np benching other brands, and if they score higher than so be it, I have seen some results that I looked at a little funny around the web.
All I know is 1 thing and 1 thing only.....
The 8RDA+ is a damn good board.
I totaly agree , and it will get even better all is needed is a bit a pacience until better bios come outQuote:
Originally posted by muzz
All I know is 1 thing and 1 thing only.....
The 8RDA+ is a damn good board.
right now is performing nicely but not optimal like abits :D
no hard feelings at all :)
good for you to be getting better scalling than me :toast:
im usinc c16
and I think ill go back to c10 :)
Well I am using dual, it does help in benches, but wouldn't see why it would have anything to do with it scaling.........
Guess I'm gonna have to hunt around for some info on the other brands........
I thought the reason Epox's performed a bit slower was because they were a tad more lax on the ram timings?
Reguardless, if you can get 5 to 10mhz FSB over the other boards, its not going to matter so much anymore.. 8-)
but asus at 204fsb is still fster than epox at 215fsbQuote:
Reguardless, if you can get 5 to 10mhz FSB over the other boards, its not going to matter so much anymore.. 8-)
:eek: too bad it died on me.
poor quad
:(
SKATAN,
You are suffering the under 10x mult slowdown @ those higher FSB's...this is part of your problem. Though I won't argue that EPoX relaxes the timings @ the higher FSB's...cuz if its anything like the 8K3A, I'm sure its being done :)
this is @10x215 with 8rda+ can any nf7s user beat this @215fsb?? (2422 dual channel)
Sorry i'am a lil late opp, but nice Clock m8 :D
i tested it again under winxp and got these results
and superpi
Can you complete 3d at that fsb using dual ch. ddr?
no:(
my max stable 24/7 use dual channel fsb sync is @213mhz
with one stick i can reach 225mhz sync stable.
Where have I been? and how did I not see this sooner? ... Congrats OPP!
whoohoo great oc! happy new year everbody :D
hey opp...my a7n8x/raddy 9700pro for some reason or another dont like 200fsb......so i settled for 195...
still not too shabby on air
http://service.futuremark.com/compare?2k1=5500089
Quote:
Originally posted by N8
SKATAN,
You are suffering the under 10x mult slowdown @ those higher FSB's...this is part of your problem. Though I won't argue that EPoX relaxes the timings @ the higher FSB's...cuz if its anything like the 8K3A, I'm sure its being done :)
N8 you seem like a fair guy, and you know a bit........... does it seem that way here though?
OPPAINTER,
How high can you clock your CPU at stock Core Voltage, 1.65V ?
and at 1.85V ?
Thank you.
news is spreading..
Way to go Oppainter! :toast:
Amazing work once again Opp. I always find myself marveling in your magnificance!
I've broke 18000 3dmarks with my radeon/2400+/a7n8x dlx combo on AIR which was no easy feat but you are the man :) you should have no problem taking the top spot!
Thanks guys :)
Happy New year :toast:
Copenhagen,
I will check out the 1.65 volts in a bit. I think the 1.85 will do 2800 stable.
OPP
At 1.65 Volts the Chip will loop 3D Mark 2001 @ 2600 Mhz. I'll check it and see how high it will do Prime95.
OPP
Truly amazing. How many CPU's did you have to try out, before getting your hands on that one ?Quote:
At 1.65 Volts the Chip will loop 3D Mark 2001 @ 2600 Mhz. I'll check it and see how high it will do Prime95.
/Copenhagen
That chip is really amazing, you got a real good one good jobQuote:
Originally posted by OPPAINTER
At 1.65 Volts the Chip will loop 3D Mark 2001 @ 2600 Mhz. I'll check it and see how high it will do Prime95.
OPP
It was luck of the Draw :D This was the first XP2700 I bought, it was Box by the way.
OPP
Why didnt you publish the 21.7k run? That would put you in second and ruffle JC's feathers.. maybe wake him from his slumber? 8-)
What do ya mean? Do I think the timings are automatically set to a lesser aggressive setting automatically by the 8RDA @ higher FSB's?Quote:
Originally posted by muzz
N8 you seem like a fair guy, and you know a bit........... does it seem that way here though?
Well, I couldn't tell ya to be honest, as I don't own the board nor have I ever played with one. I was just sayin that I wouldn't put it past EPoX, being a 8K3A owner. I see your results & they look fine, I'd be happy with them if I was ya ;)
baller02:
Is that TwinMOS equipped with TwinMOS or Winbond chips?Quote:
with multiplier 10 my 8rda+ scales nicely up to 215 fsb sync dual channel with 2422 timings which is the limit for my twinmos pc3200. on my 8k3a+ this ram also doens't go higher.
What sizes are they, 256 MB DIMMs? (You use 2x right?)
:stick:
Well from my (less informed) point of view, with the bios set to optimal memory timings, as you increase FSB the memory timings do relax. If you set expert mode for memory timings, you can lock the faster timings at whatever FSB. With my pc2100 I must use expert memory timings so that I can run memory synchronously with the 166 FSB. PCI is alwasy locked to 33 Mhz BTW and AGP can be locked to whatever or be left related to the FSB speed.Quote:
Originally posted by N8
What do ya mean? Do I think the timings are automatically set to a lesser aggressive setting automatically by the 8RDA @ higher FSB's?
Well, I couldn't tell ya to be honest, as I don't own the board nor have I ever played with one. I was just sayin that I wouldn't put it past EPoX, being a 8K3A owner. I see your results & they look fine, I'd be happy with them if I was ya ;)
OPP I still wanna know....
Vapokill or prometeia?
BTW my 8RDA+ runs fine at 200mhz plus..
http://homepage.ntlworld.com/mark.fa...t%20ownage.jpg
:D
http://service.futuremark.com/compare?pcm=892664
me 3dmark was done at lower speed, havent had time to update it, been busy with me R/C chopper :D
i think opp was most definatly running prometia..............:banana:
Galifrey,
Yea it was on Prometeia.
Maybe you can tell me, is that a 2700 chip? If so can you use the 13.5 and 14 multiplier with that mobo??
Thanks,
OPP
You can supposedly run 13 and 13.5 but ive not tried it myslef (I dont have a 2700+)
Hi Opp
yeah its a 2700+ and 13.5 and 14 mults work VERY intermittently
I have to mess with it for hours to get them to work, and then randomly on reboot it restarts as an xp1500+ (13x100)
Dont understand why, so have given them up...
Had it boot to windows logon at 15x204 tho :D crashed when the sound played... gah...
Sounds are evil!Quote:
Originally posted by Galifrey
Had it boot to windows logon at 15x204 tho :D crashed when the sound played... gah...
Opp, YGPM
i suggest you take that chip out for a while.. since its such a beauty and keep it stored away, until nvidia releases a gffx.....
Don't let it burn out etc.. (hope i didnt jinx it lol)
I step away from the forums for a little while and when I get back Opp's tearin up the performance world again. WOW!!!!!!!
Tiribulus:toast:
Its simple. Each succeding chip is 66.6666666666666MHz more than the last. the 1600+ ran at 1400. the 1700 ran at 1466.6666666. The 1800 ran at 1533.3333333333. The 1900 ran at 1600. So, take 3100MHz, subtract 2133.3333333333333 and divide by 66.666666666667.Quote:
Originally posted by shray
convert that from mhz to the amd rating scale?
Anybody have that link which converts?
Or set up a proportion...
1800 = ??
------ ----
1533 = 3100
i think .... lol
about 3600+ right??
3100-2133.3333333333333333333 = 966.666666666666666
966.66666666666666666666666666/66.666666666666666666 =
14.5
Add 14.5 to 27 to get 41.5= Athlon XP= 4150, :)
Oh, and to OPP:
:slapass: :smileysex :stick: :toast: :frag: :slash: :thumbsup: :hehe: :rocker: :banana:
OPP:)
On my 8RDA+ 13.5 and 14 DID NOT work...13 did as well as 15. This was a biggie for me as you know..I need 13.5x to achieve my max CPU Mhz and FSB with the chip I've got.
Randi:D
Wrong.Quote:
Originally posted by Penguin4x4
Its simple. Each succeding chip is 66.6666666666666MHz more than the last. the 1600+ ran at 1400. the 1700 ran at 1466.6666666. The 1800 ran at 1533.3333333333. The 1900 ran at 1600. So, take 3100MHz, subtract 2133.3333333333333 and divide by 66.666666666667.
1600+ = 1.40 GHz
1700+ = 1.47 GHz
1800+ = 1.53 GHz
1900+ = 1.60 GHz
2000+ = 1.67 GHz
2100+ = 1.73 GHz
2200+ = 1.80 GHz
So far so good, but as the 66 MHz per 100 rating scales much faster than a P4 does, AMD is risking lying more as they keep ramping up the QS ratings without raising MHz by more than 66 per step, so here comes some adjustments:
2400+ = 2.00 GHz
This one gets 200 rating on 200 MHz!
Now comes the next cut in the line, the 333 MHz FSB:
2600+/333 = 2.08 GHz
That's 83 MHz for 200 rating, eh! The FSB jump increases the QS rating more, though. Here's the two next CPUs:
2700+ = 2.16 GHz
2800+ = 2.23 GHz
That's 83 MHz per 100 QS rating. You can figure out that this might hold for future Tbreds on 333 MHz bus also, but soon you'll see a Barton 512 L2 cache coming... with more cuts in the QS rating scheme.
I heard that the latest (now official) 2C31 BIOS has fixed some of the multipliers issues...Quote:
Originally posted by MrIcee
OPP:)
On my 8RDA+ 13.5 and 14 DID NOT work...13 did as well as 15. This was a biggie for me as you know..I need 13.5x to achieve my max CPU Mhz and FSB with the chip I've got.
Randi:D
--fibo
Well, yeah, its 83 MHz per 100 QS rating when running on the 333MHz DDRFSB, but its still 66 MHz per 100 QS rating when runnin 266MHz DDRFSB, no?:D:D:D:DQuote:
Originally posted by Pjotr
Wrong.
Now comes the next cut in the line, the 333 MHz FSB:
2600+/333 = 2.08 GHz
That's 83 MHz for 200 rating, eh! The FSB jump increases the QS rating more, though. Here's the two next CPUs:
2700+ = 2.16 GHz
2800+ = 2.23 GHz
That's 83 MHz per 100 QS rating. You can figure out that this might hold for future Tbreds on 333 MHz bus also, but soon you'll see a Barton 512 L2 cache coming... with more cuts in the QS rating scheme.
No, the 2400+ is a 266 fsb CPU and is 200 MHz faster than the 2200+
Thanks for this link! I hope it solves the bios temp alarm going off after changing the multiplier from default (*15 xp2400+) to *14 with a FSB increase to 166Mhz. Otherwise it stays off. I also wish the temp cut -off options extended alot lower.Quote:
Originally posted by fibonaccov
I heard that the latest (now official) 2C31 BIOS has fixed some of the multipliers issues...
--fibo
I checked 2C31 bios and it allowed my 2400 with a multi of 13 whereas with previous bioses it wouldn't have worked...fyi.Quote:
Originally posted by megatron
Thanks for this link! I hope it solves the bios temp alarm going off after changing the multiplier from default (*15 xp2400+) to *14 with a FSB increase to 166Mhz. Otherwise it stays off. I also wish the temp cut -off options extended alot lower.
--fibo
Its the same as c16 for me. Its a keeper though. I meant that I cant go straight to 166*14 @1.925v. I must make an imtermediate o/c of 13* 166 @1.85v and after that no problem with reboots or coldboots.
It depends on the cpu - if u cant run at 14x166 then increase voltage a bit more - running a 2400 at that speed can require 2.0 to 2.1v - my 2400 will need 2.1v to approach 2.35 so u might be in the same boat - nothing to do with the bios (totally CPU dependent) - try more voltage until it is stable with a limit of 2.1v I'd say...Also, you might want to increase the FSB higher rather than staying at 166...just my 0.02 cent...
--fibo
/me looks at above and :confused: :doh: /Quote:
Originally posted by fibonaccov
It depends on the cpu - if u cant run at 14x166 then increase voltage a bit more - running a 2400 at that speed can require 2.0 to 2.1v - my 2400 will need 2.1v to approach 2.35 so u might be in the same boat - nothing to do with the bios (totally CPU dependent) - try more voltage until it is stable with a limit of 2.1v I'd say...Also, you might want to increase the FSB higher rather than staying at 166...just my 0.02 cent...
--fibo
Thanks for your advice. Will someone misinterpret my last sentance too? I seem to not be making much sence to you guys. My last two posts have been about explaining what I meant in the previous post. :D
I have my CPU @1.925v rock solid stable. I have got into windows @ 14.5 * 166Mhz @ 1.975v, its couldn't even start the splash screen for 3dmark2001se 330. The temp seemed to be rising too. My watercooling setup is quite good but not that good. If I was still running a pelt I could, but pelts and my CPUs have taken a dislike to one another.
My memory is fine @ 166Mhz. I am thinking of buying some OCZ PC3200 that is E105 (euros) for one 256Mb stick (want two). I think pc3500 is just not going to benifit much more since my board apparently doesn't like high FSB so much - and I think the extra cost is not worth the extra theoretical performance. Plus this memory is on sale and looks like the current sweet spot for memory prices at the moment.
To explain why my system's alarm goes off: Its due to a buggy bios.
I do what I said above to get round this buggy bios, not because I have too low a voltage. 1.9v was too low because I could get into windows but games would crash.