trodas

Noproblemo! My best score with nVidia Vanta LT is, after going for 2 days, in 1170th frame out of 5200. I think I can wait... LOL Glad you respond, that counts! (I starting to think that time-lapse camera could be quite good thing there (but my old Nikon 4300 did not qualify, sadly), the total run lenght might get to 5 days or so) If you really need the Vanta LT to fix the wrapper, then I could borrow it to you for that

Bump for making the scorefiles!

free baloon rides

I don't consider cheating as normal. I can understand the need to "feel better that the others", yes, but go to the point of lying to yourself? Nah. Either I'm better, can mod, tweak and get higher score (or go thru plenty of rams/CPUs and boards to find the "golden ones"), or I'm just not. That it is. If you spot cheat, please report cheater. I believe that the competition is about reaching real speed, not imaginated. Sure it is not hard to bench with, for example, GeForce 2ti and pass the result as ATI Rage 128 Pro. But keeping that results in database is 1) insulting real benchers 2) it does damage on users, who cannot understand, how others get such results, when they - even when get to same clocks - cannot even get close to, much less get match or surpass. And I could do even better job that some cheaters do I will, for one, not do the mistake as to put higher FPS wher they are lower (Game 1 vs Game 2 test in 3DMark 99, for example) and I will never forget to match the test results - eg. when the selected tests mark some, they will have to be run (and they are not in the 1st 3DMark99 score on Rage 128 Pro: http://hwbot.org/submission/2248070_ ...so I think I could go way better job at cheating that these poor souls, lol. But maybe I'm entierly wrong and they do not cheat and I just overlook something? Dunno So I could use your opinion and advices there I don't want to see cheaters running rampant there.

Funny guy. Of course I have set Vsync off in the ATI panel settings If I don't, there will be just 15, 30 and 60fps results on the screen... and much lower score. Already learned that 3DMark99 like to use Vsync Now tell us, how there can be so huge difference between very similary clocked cards... because if you just throw some BS around, like "just turn off Vsync", then that is nowhere near to explain away these vastly different results. Either the ATI Rage 128 Pro is much more powerfull card that it show in all other benches, or you are cheated badly and boldly. Witch it is?

Don't punish all Italians, pls. That is bad. Ther are good Italians I trust, check the thread there: http://forum.hwbot.org/showthread.php?p=399955 So for few bad apples did not throw them all away. IMHO scores with different mems pictured that used should be discarded, when we talking about first 3 places in the world. It happens, that one can use older pictures of rig, when changing items... but this is IMHO not permisible for WR.

I take that offer too. Sending PM, reached 179 ATM after 12h+ long run. Very disapointed, wanted lower. Taking steps to change that...

Sure, I will install Win2k with biggest pleasure... and can use DX 8.1 or DX 9.0c and we see, who is cheating Or how the score change! I have nothing against you, if you did not cheat (BTW, this is lame excuse, because all the possible effectivity difference is erased by brute force of X6800 GPU at 3.4GHz...) And he wanted to discuss in THIS thread only the reasons why holding back on some of the tips, nothing else. Not to discuss specific submissions, but to discuss specific ram chips/types that are capable of 2-2-2 at past 200MHz marks. That is something interesting for everyone, I bet.

During preparation for super-slow runs, I took my ATI Rage 128 Pro card and run some tests. I was a bit concerned, how the card could reach results like 21 447, 14 078 and 13 128 marks in 3DMark 99: http://hwbot.org/submission/2248070_drswizz_3dmark_99_max_rage_128_pro_21447_marks'>http://hwbot.org/submission/2248070_drswizz_3dmark_99_max_rage_128_pro_21447_marks http://hwbot.org/submission/2496046_stermy57_3dmark_99_max_rage_128_pro_14078_marks'>http://hwbot.org/submission/2496046_stermy57_3dmark_99_max_rage_128_pro_14078_marks http://hwbot.org/submission/2349966_havli_3dmark_99_max_rage_128_pro_13128_marks'>http://hwbot.org/submission/2349966_havli_3dmark_99_max_rage_128_pro_13128_marks So my main question was - could physically the card handle it? It is possible to draw the game scenes at such high framerates to get so high results? After all, mine Rage 128 Pro can do w/o overclocking only 3 351 marks on default clocks: http://hwbot.org/submission/2937935_ Even when supported by 3.4GHz X6800 CPU. So, exhibit no. 1 - Rage 128 Pro at 118/140MHz do 3351 marks and 30.1 and 37.8 fps in game tests respectively: Is this normal? It is. Exhibit no. 2 - Rage 128 Pro doing 3DMark 99 bench on video: As you can see, around 30fps is normal result for stock Rage 128 Pro clocks. 2673 marks. So I was confronted from a bit shockingly high results - 3351 marks I have and next person (havli) is having 13128 marks. That is about 10 000 marks more. Could these poins stack so quickly when overclocked? We see! Exhibit no. 3 - Rage 128 Pro at 150/165MHz do 4984 marks and 47.8 and 52.1 fps in game tests respectively: Exhibit no. 4 - Rage 128 Pro at 150/170MHz do 5020 marks and 48.1 and 52.5 fps in game tests respectively: Please note that the performance scale with clocks up rather consistently. My Rage 128 Pro are (at these hot days) showing some little glitches in the image at the end of game 2 test, so maybe there are some errors that impair a bit on the speed. All in all, 150MHz for core is the maximum I could get out of the not modified card (except previous user stock some crazy heatsink on it, that it is). ************************************************** Now let's go to muddy watters there. Fake no. 1 - Rage 128 Pro at 155/166MHz do 13128 marks and 150.3 and 117.6 fps in game tests respectively: http://hwbot.org/submission/2349966_havli_3dmark_99_max_rage_128_pro_13128_marks Points to consider: - second game test is always and consistently faster, NOT slower - 150/170MHz do 48.1/52.5fps, but this 155/166MHz do 150.3/117.6fps?! - on the screenshot, the selected tests are not visible, hidden by the windows - is not that grounds for invalid test all by itself? Fake no. 2 - Rage 128 Pro at 177/190MHz do 14078 marks and 162 and 124.5 fps in game tests respectively: http://hwbot.org/submission/2496046_stermy57_3dmark_99_max_rage_128_pro_14078_marks Points to consider: - second game test is always and consistently faster, NOT slower - 150/170MHz do 48.1/52.5fps, but this 177/190MHz do 162/124.5fps?! Fake no. 3 - Rage 128 Pro at 125/143MHz do 21447 marks and 254.8 and 185.2 fps in game tests respectively: http://hwbot.org/submission/2248070_drswizz_3dmark_99_max_rage_128_pro_21447_marks Points to consider: - second game test is always and consistently faster, NOT slower - 150/170MHz do 48.1/52.5fps, but this 124/143MHz do 254.8/185.2fps?! - the tests selected include the filrates and textu rerendering speed, however in the resulting details are these tests shown as not run (N/A), so obviously some doctoring of the screenshot take place ************************************************** Now lets compare a good result from not cheating overclocker: Rage 128 Pro at 160/180MHz do 4756 marks and 47.9 and 47.3 fps in game tests respectively: http://hwbot.org/submission/2381644_skyline_3dmark_99_max_rage_128_pro_4756_marks Points to consider: - second game test is slower, but within the margin for error - 150/170MHz do 48.1/52.5fps, so 160/180MHz can do 47.9/47.3fps easily - he use the Pro version (registred), unlike all the fakers before - he show the GPU-Z screenshot, unlike all the fakers and I can confirm, that GPU-Z does not show actuall clock... (witch is why I made second screens with Powerstrip) - tests witch he choosed not to run won't run, unlike some fakers above So, on czech forum(s) I heard havli boast and gloat on others, trying to beat his scores, speaking that thanks to his "tweaks", he is unbeatable (he said it specifically to 3DMark 2001 score(s) ): http://pctforum.tyden.cz/viewtopic.php?f=76&t=202370&start=220 https://translate.google.com/#cs/en/Doc%3A%20ten%20quad%20hlavne%20preto%20ze%20chcem%20zautocit%20na%20Havliho%20rekord%20s%20X8x0XT%20PE%20%3Atwisted%3A%20a%20tiez%20preto%20ze%20je%20to%20Eng.%20sample.%20Inak%20tam%20dam%20bud%20tu%20X6800%20alebo%20E6700%2C%20podla%20stastia%20v%20aukcii.%0A%0A%2F%2Fje%20to%20nejaky%20starsi%203DMark%2C%20takze%20to%20CPU%20by%20tam%20mohlo%20pomoct%2C%20havli%20to%20robil%20na%20Phenome%20II%0A%0AJestli%20chces%20prekonat%20muj%20rekord%20v%203DMarku%2001%2C%20tak%20to%20budes%20mit%20hoodne%20tezky.%20%3Atwisted%3A%20http%3A%2F%2Fhwbot.org%2Fsubmission%2F2314592_hav%20...%206799_marks%0A%0AJe%20to%20totiz%20muj%20nejoblibenejsi%20benchmark%20a%20mam%20zmaknutych%20vetsinu%20tweaku.%20A%20to%20jeste%20tohle%20skore%20vyse%20neni%20uplne%20nejlepsi%2C%20co%20se%20da%20dosahnout%20-%20jsou%20tam%20rezervy.%0A%0AJinak%20co%20se%20tyce%20legacy%20benchmarku%2C%20tak%20ctyrjadro%20ma%20prinos%20jen%20v%203DM06%20%28a%20to%20vesmes%20jen%20v%20CPU%20testu%29%2C%20starsi%20jsou%20single-thread...%20takze%20tam%20dvoujadro%20bohate%20staci. So that leave us two possibilities: 1 - he (and the two others there) is cheating - not a big problem to run the tests with other GFX card and pass it as Rage 128 Pro, IMHO 2 - he (and the two others there) is using tweaks that boost the Rage 128 Pro speed up to the sky and that make these scores invalid, because they are no longer represent best hardware speed (like Skyline's results, 160/180MHz is damn impressive for Rage 128 Pro... even with replaced caps and peltier cooling I cannot hope for something like that, IMHO), but the baddest tweaks, that could be considered cheats, even if they did not meddle with the 3DMark 99 code directly... witch I doubt, because there is no way Rage 128 Pro can do 254fps in any kind of 3D action ...but maybe I will be proved wrong completely?

I completely agree, that sharing information helps comunity as whole. I completely despise guys, who claim that their scores are "because of their tweaks" unbeatable, when all what they did was just cheating. I mean... Rage 128 Pro doing 185fps in 3DMark 99 "3D person shooter" game 2 test? You gotta be kidding me that this is because of "tweak." That is BS. The card really did not have such power at all, even paired with X6800 Extreme CPU at 3.4GHz it cannot get over 52.5fps no matter what (150/170MHz). If anyone want seriously suggest, that a 155/166MHz clocked Rage 128 Pro can do 117.6fps there, then he damn well better show these "tweaks", or have get banned for cheating... But that is for different thread to start... (there: http://forum.hwbot.org/showthread.php?p=399955 ) Except sharing direct code/executable (for example I get the idea, that the PCMark04 grammar test could be made to run on multiple cores (and even utilize them for great gains!) by blocking the shared cache after the call for the test is made... so when executable is made, then sharing it will be probably going too far, but spelling out the idea is IMHO a good thing for others to catch on...) I would consider not sharing informations as too secretive and harmfull for society. ... I was just recently read that something like waza exist! I was thinking that benching is about the HW, not that much about SW )) But I learn that I could get better results by just setting realtime priority to SuperPi and ending explorer task (not to mention having optimized windows counts as well)... but not for all benches. PCMark 04, for example, get worser when running on maximum priority, because the tests of other programs will suxx then... So I completely agree that THIS IS THE PLACE to discus these things, even that some I would call almost cheating (tweaking D3D settings for 3DMark to run better) and I do not use them.

Only Elpida MGH-E Hypers can do on AMD 6-6-5-15-26 1T @877MHz Congratulations. Your english is good, hoooray to Chile for Hypers! I need 1.72V to get Corsair Dominator GT over 833MHz with Hypers at 6-6-6 stable, but this is triple channel and Intel X58, so... different beast

Well, Jetway recognize the board: http://www.jetway.com.tw/jw/motherboard_view.asp?productid=29&proname=V600DAP And it is Jetway. ECS KT600 (KT600-A and variants) is quite different board: http://www.ecs.com.tw/ECSWebSite/Product/Product_Detail.aspx?DetailID=360&MenuID=24&LanID=0 http://www.newegg.com/Product/Product.aspx?Item=N82E16813135143 http://akebonotown.com/bouken/pc_KT600A.htm It might be similar by chipset, but certainly it is not and nowhere near same mobo. So, since I know on what mobo I benching... could you please change the scores mainboard vendor? I can edit the type, but entering "V600DAP" is no use, becuse the vendor is locked to be ESC, witch is not the case Thanks!

Thank you, much nicer!

Okay, I was a bit affraid, but I tried it. Yes, the double notches are universal cards and they run into 1.5V AGP only boards w/o a glitch. So ASRock 775i65G with Ati Rage 128 Pro works well. Not exactly good GFX card for way more modern (Core 2 Duo) board, but that is usefull for some other stuff... Who says I must bench at Jetway V266B ( http://hwbot.org/submission/2468900_ ), when there is a better choice?

I would like to know, if I'm not mistaken and AGP card with two notches in the connector is universal AGP 3.3/1.5V card and therefore CAN be used on boards with WARNING about not using AGP 3.3V cards. So is this card: ...universal card, as from this source: http://www.playtool.com/pages/agpcompat/agp.html ...it looks like? ... Bottom line - Ati Rage Pro (AGP v2, 4x speed, universal card?) is usable in ASRock 775i65G mainboard (AGP v3, 8x speed) witch does not support AGP 3.3V cards, when the card have two notches, right or wrong?

This legendary card has no image! http://hwbot.org/hardware/videocard/rage_128_pro/ That should be fixed sooon Thanks! :celebration:

Thanks for the tip. I run some super-low Aquamark scores and then set it up on another bench table, where it can sit for 4+ days, lol. Maxmem low? How to set it? Where? I never thought about it. I was just trying to take a look, what scores will HWbot accept as these super-low ones (Special Achievements, Submit 5 SuperPI 32M runs which took over 12 hours to finish.) and I had a good way to produce that result, based on very special downclock of Duron. Yes, one can just use original Pentiums... but when I stumbled upon a bug, that produce about 30MHz FSB on one MSI board: http://valid.canardpc.com/evdcpe ...then I was like - that is the interesting way to do it Sadly I never managed to run the x5 Duron yet into the mainboard, witch is kinda sad. But I will keep pushing and trying to figure out what went wrong... And even 30x7.5 was too fast and Duron made these SuperPi 32M in around 8 hours or so... There it is, 8h 32min: http://hwbot.org/submission/2435812_ So w/o working multiplier x5 I cannot manage that low scores, even having very very low FSB 30MHz

Bump with another Aquamark score going to wrong mainboard: http://hwbot.org/submission/2937525_ ...when I submit the mainboard precisely, I get it to recognize: http://hwbot.org/submission/2937532_ ...but Aquamark wrapper insist on wrong mobo. Maybe it is time to update the wrapper with new CPU-Z detection database, when Franck update the detection... Information (CPU-Z report.txt) was send today anyway. And another one in need of editing: http://hwbot.org/submission/2937575_

placebo steroid pills

After twaking a bit with my Duron 750 (now x5 multi), I set it slowest as it could go in bios and let it run SuperPi 32M overnight. Yep, booting to WinXP took about 15min alone, but I was patient. Now in the morning, I saw this: So it took 1h 11min 44.375sec for the Duron to come up with just the "initial value." What will be the total time of the test? Could I roughly estiminate, that when doing on same board SuperPi 1M test, then it took with AXP 47.765 sec ( http://hwbot.org/submission/2462459_ ), while 0.625 sec took the initial value... So that is 76.424x longer for the vhole result, witch could push the time to 5502.528 minutes = 91 hours = 3.8 days? Could that be somewhat roughly accurate?

Stunning 2-2-2 rams ... SDRAM at 219MHz doing 2-2-2-7 is really something to behold! Some pictures could be handy, too Probably also pushed to 3.7V at least for the rams, as gradus, yep?

I have a first testing Aquamark score on my Jetway V600DAP: http://hwbot.org/submission/2937449_ And the score was moved under ECS mobo: http://hwbot.org/hardware/motherboard/kt600_8237/ ... Now this is wrong and I don't see, how can I edit that 1) HWbot recognize the mobo: http://hwbot.org/hardware/motherboard/v600dap/ 2) CPU-Z does not recognize the mobo: http://valid.canardpc.com/1r0xd3 ...but it is certainly NOT ECS KT600-8237 mobo, I can prove that by pictures and other info: Could some mod move this test to appropriate mainboard? (yep, there will be more to come...)

Looks like the closest is the Samsung SH-224DB white drive: (Samsung SH-224DB "white" drive, I fear it will be more baige-like...) I got also warning, that: ...but hopefully that would be fixed, as Samsung is releasing plenty of firmware updated. Worser is, that there is no RPC-1 firmware (region free). Older Samsung drivers are covered: http://codeguys.rpc1.org/firmwares_samsung.html But not yet this one.

Laughable. Same CPU settings, just 2T 9-9-9 rams and it is 8.13: http://hwbot.org/submission/2934676_ That means shock, because 1T 6-6-6 rams give only 0.01 points increase...

Rams from Wizerty recieved today. Very nicely packaged, each ram with sticker that it was tested OK and date of the testing. Professional Thx Wiz for the nice MSI bracelet, lol MY GF seems to like it... but I bet she more like that I will no more canibalize "her" PC, when I need rams to test No damage, MSI X58 Pro-E posted with these triplechannel rams right away and just fine. Wizerty seller reputation +1 Proofs: http://valid.canardpc.com/g8j4hi (need 1.72Vdimm to be benched at 6-6-6-18 timings, tough) SuperPi 6-6-6-18 1T 1.72Vdimm http://hwbot.org/submission/2935826_trodas_superpi___1m_xeon_x5650_11sec_203ms Cinebench 11.5 (32bit only OS&test, so score suxx) 6-6-6-18 1T 1.72Vdimm http://hwbot.org/submission/2935830_trodas_cinebench___r11.5_xeon_x5650_8.14_points ... On more serious note, they can do 7-7-7-20 1T @ 903MHz: And the look is impressive: So, dealing with Wizerty was friendly (he even send the rams when I send at first only about half the payment, as there was not enought money on the PayPal account) and I could only recommend him. ...now off to play with new things :nana: Thanks, Wizerty!