View Full Version : the 4ghz challenge discussion thread
so we will start the threaad once we have a few things sorted out, like what benchs what proof and that kind of thing.
so any ideas on what benchmarks would be good to show just cpu power? things that dont favour green or blue sides would be nice and one which are indicative of real world performance would be good too.
of course we can do the more run of the mill things like super pi or sandra and such but there has to more than just that.
what about proof? cpuz with 2 pages for cpu and ram speeds, maybe a task man to show uptime and loads?
i guess the other question is how many people would play? if its only me and martin i guess we could just pm each other and leave it at that but if others are going to throw som numbers up we could start a results thread to chart how different systems go.
what about leagues/sub sets for single/dual and quad core cpu but anything else?
I can get my E8400 to 4Ghz and 100% stable, so I could play if I get the time to do it. Also, (and I know it doesnt sound half as cool :lol:), there are far more people who can hit 3.8Ghz ish than 4Ghz. Is it worth thinking about making it something like that to include more users? :)
It's kinda weird 4Ghz cause an AMD 965 can hit 3.8 on stock volts and needs like 1.5v+ to get 4Ghz :p
It's kinda weird 4Ghz cause an AMD 965 can hit 3.8 on stock volts and needs like 1.5v+ to get 4Ghz :p
Which is my point, a truly stable 4GHZ AMD is next to impossible.
However, when I opened my window and load temps maxed at 44C, I was 1.4V stable.
Which is my point, a truly stable 4GHZ AMD is next to impossible.
However, when I opened my window and load temps maxed at 44C, I was 1.4V stable.
So you decided to do a Will? :p
What are your load temps normally and what cooling are you using?!
At 3.8Ghz 1.55v my Phenom II 720 is at 39C :clap:clapI should really unlock my 4th core and prime that and see what temps I get.
I didn't do a Will lol.
If I'd taken my PC into the snow, I'd probably get like 1.375V stable at 4GHZ.
I should really unlock my 4th core and prime that and see what temps I get.
http://webinstituteforteachers.org/%7Emjackson/sun.jpg
:chairfall::clap:clap
I think my motherboard might be holding me back seeming that it's AM2+ and apparently they don't overclock as well as AM3 boards. My board hates FSB frequencys above 250 :thumbsdn:
I think once I was at 270 x 14 @ 1.51v which is about 3.7Ghz and it said boot error etc so I went into the bios and on system information it said:
Phenom II x3 720
Speed: 4320MHz :surprised:
weird.
it sounds like it went back to default multi m8, and for that speed you would need lots of voltage.
it sounds like it went back to default multi m8, and for that speed you would need lots of voltage.
Default multi is 14 though....my board is just strange.
14 x 200 = 2800 = stock
If I'm at 250FSB but HT Link and northbridge are at 2500MHz...fair enough. If I got to 251+MHz (FSB) then my northbridge will continue to increase in speed (2510MHz+) but my HT Link multi (which there isn't even an option for in my BIOS) will drop right back to x4 1004MHz+
Which is my point, a truly stable 4GHZ AMD is next to impossible.
However, when I opened my window and load temps maxed at 44C, I was 1.4V stable.
Not next to impossible, but can be hard on air. I have mine stable at 4.1Ghz on WC'ing and my WC'ing is basic in the radiator department.. just a 120mm rad and a cheap pump/res combo.
:chairfall::clap:clap
I think my motherboard might be holding me back seeming that it's AM2+ and apparently they don't overclock as well as AM3 boards. My board hates FSB frequencys above 250 :thumbsdn:
I think once I was at 270 x 14 @ 1.51v which is about 3.7Ghz and it said boot error etc so I went into the bios and on system information it said:
Phenom II x3 720
Speed: 4320MHz :surprised:
weird.
Who said AM2+ boards don't OC as well as AM3? My board currently using WC'ing can go as far as 350FSB.. thats on £57 board..
I've had other cheapo boards going as far as 305 too.
Next to impossible on air I meant, I should have explained further lol.
Not next to impossible, but can be hard on air. I have mine stable at 4.1Ghz on WC'ing and my WC'ing is basic in the radiator department.. just a 120mm rad and a cheap pump/res combo.
Who said AM2+ boards don't OC as well as AM3? My board currently using WC'ing can go as far as 350FSB.. thats on £57 board..
I've had other cheapo boards going as far as 305 too.
350FSB on my board would make it erupt into a super nova :p
Lol, get my board instead :P Aria don't stock it though. Probably the best bang for buck AM2+ board out there.
Lol, get my board instead :P Aria don't stock it though. Probably the best bang for buck AM2+ board out there.
Sorry I'm ASUS loyalist :p - quick connectors ftw :p
I'll probably get a ASUS M4A79-T Deluxe :) (just to let you in a on a secret...when I bought my M3A79-T Deluxe from **** £62 brand new :D I actually thought I'd bought a M4A79-T and had just gotten confused haha) or see what comes with the new AMD chipset.
If you get this sorted, ill play :D
And a small suggestion - rather that 4Ghz - why not do it as a percentage - that is then not excluding as may people and making a slighly more level playing field?
Andrew Moore
10-01-10, 14:38
That sounds good jon especially considering both you and I can hit 4.5ghz without issue.
Andy
There should be separate Intel and AMD ones :)
I thought the point was to see what CPU was quickest at 4Ghz, rather than what cpu within a brand.. There should at least be an overall table?
thats the idea, @4ghz (or lower if you cant hit that) which is best and at what?
I presume we already know this....Core i7 - i5 -core2quad - Phenom II x4 - i3 - Phenom II x3 - core2duo - Phenom II x2 and the other ones :rolleyes: :P
the idea is to try and see what does best at what, martin seems to think his can beat an intel at rendering, which imo is intels strog side but the only way to find out is for us test our stsems under the same (ish) conditions to see.
so what should we use to test? some things that show some real world ish erformance as well as synthetic numbers as i think we need a bit of both to see if they talley up.
i dont think we should be striping down our os either to get the most out of it either but this will be down to trust as we cant really do much to prove/dis this.
Andrew Moore
10-01-10, 16:47
Custompc pc bench suit ! It's a rounded everyday suit, perfect to see what's going on at 4ghz.
Andy
the idea is to try and see what does best at what, martin seems to think his can beat an intel at rendering, which imo is intels strog side but the only way to find out is for us test our stsems under the same (ish) conditions to see.
so what should we use to test? some things that show some real world ish erformance as well as synthetic numbers as i think we need a bit of both to see if they talley up.
i dont think we should be striping down our os either to get the most out of it either but this will be down to trust as we cant really do much to prove/dis this.
Everybody should have to use a 64-bit OS :) proved by a 'my computer' screenshot.
or have seperate tables for 32 and 64 bit os?
Andrew Moore
10-01-10, 17:49
As previously said. The cpc bench mark give numbers as well as overall results.
Andy
Cinebench for rendering, that gives scores for a single core and multi core run? Also has x64 and x32 .exe's
http://www.maxon.net/en/downloads/downloads/cinebench.html
I will pass on this little "game" :).
Cinebench for rendering, that gives scores for a single core and multi core run? Also has x64 and x32 .exe's
http://www.maxon.net/en/downloads/downloads/cinebench.html
I will pass on this little "game" :).
Surely it's just having a 64-bit OS that matters not using a 64-bit program to benchmark.
Surely it's just having a 64-bit OS that matters not using a 64-bit program to benchmark.
:surprised:
Think about what you just said.
you would run out of room fast is your sig'd all will's classic quotes :D y we love him :D
cpc suite is a good one, thats on the list.
cine bench is another good one but iirc thats in the cpc suite isnt it?
@ Marsey, No idea mate, seen it being used about various places, mucked about with it myself on the odd occasion.
@ Will heres a run using the 32bit version and 64 bit version on Vista 64. As you can see the 64 bit version does considerably better than the 32 bit version on a 64 bit OS ;)
http://i880.photobucket.com/albums/ac6/Aleksandr_Orlov/screenshots/Cinebench32bit.jpg
http://i880.photobucket.com/albums/ac6/Aleksandr_Orlov/screenshots/cinebench64bit.jpg
i know it used to be in the old suite but i never ran the last 1.
what i dont know is where to get it now as they dont have their own site anymore and i cant find it on the bit tech site???
Surely it's just having a 64-bit OS that matters not using a 64-bit program to benchmark.
:chairfall::chairfall::clap:clap
:chairfall::chairfall::clap:clap
:surprised:
Think about what you just said.
Guys :p
I didn't mean productivity wise...I mean stability wise. It will be more stable running a 32-bit OS than a 64-bit OS.
So how do you plan to fully test 64bit processors if you aren't using a 64bit OS?
So how do you plan to fully test 64bit processors if you aren't using a 64bit OS?
Run a 64-bit OS obviously. It's just unfair if people are running 32-bit OS...jezzzzz how hard :lol:
run a 32bit app twice?
if only you were being serious :chairfall:
I can I use my 4th core? :D should be stable enough to render on Cinebench :)
run a 32bit app twice?
:lol: :lol:
Run a 64-bit OS obviously. It's just unfair if people are running 32-bit OS...jezzzzz how hard :lol:
But then that doesnt give a fair test either! :lol: :lol:
Run a 64-bit OS obviously. It's just unfair if people are running 32-bit OS...jezzzzz how hard :lol:
Which is why i said Cinebench as i knew it had both 32bit and 64bit exe's if its used then you would need 2 tables, one for each type of OS. As it stands if it is used there will need to be different charts for numbers of cores so why not OS?
I can I use my 4th core? :D should be stable enough to render on Cinebench :)
Down to the guys running the show i would imagine, if there is a rule stating it need to be stable for XX amount of time proven by running whichever stability test then possibly not, but if it only needs to complete the given benchmark then possibly yes.
I'm 64 bit stable..
32 Bit is easy to get 4GHZ stable lol, I could do that on 1.425V.
I'm 64 bit stable..
32 Bit is easy to get 4GHZ stable lol, I could do that on 1.425V.
This is exactly what I'm referring too. People who have access to 32-bit OS will be able to reach higher clock speeds which is unfair to those that don't.
My old C3 965 was easy stable at 4GHZ on 32 bit, and that had a higher stock voltage than this one.
Im on 64Bit and im stable 4.1Ghz. Thats a C2 Ph.II.
is it really easier to hit higher overclocks using a 32bit OS?
That just strikes me as extremely odd when you all talk of stability and yet a 32bit write to ram (smaller number of data in one block) which is smaller passes validations easier than 64bit. If it is the case, then I would class your high overclock as not stable if it fails in ANY os.
DT.
Im on 64Bit and im stable 4.1Ghz. Thats a C2 Ph.II.
You're on water, they scale better to cold than they do volts.
is it really easier to hit higher overclocks using a 32bit OS?
That just strikes me as extremely odd when you all talk of stability and yet a 32bit write to ram (smaller number of data in one block) which is smaller passes validations easier than 64bit. If it is the case, then I would class your high overclock as not stable if it fails in ANY os.
DT.
That was my old C3, and it's way worse than this one, I'm on 64 bit now.
My E8400 is stable at 4Ghz on 64bit.. I never really even figured it would make much of a difference..
is it really easier to hit higher overclocks using a 32bit OS?
That just strikes me as extremely odd when you all talk of stability and yet a 32bit write to ram (smaller number of data in one block) which is smaller passes validations easier than 64bit. If it is the case, then I would class your high overclock as not stable if it fails in ANY os.
DT.
Not always possible to test for though
My E8400 is stable at 4Ghz on 64bit.. I never really even figured it would make much of a difference..
Neither did I...Rawz let me into that secret but kinda makes sense as 64-bit OS use more memory...so I would think it puts more stress on the memory controller.
The 2007 CPC Bench Suite can be downloaded HERE (http://movies.custompc.co.uk/custompc/CPCSuite2007.zip) if anyone is interested :D
will run it later m8 if i get chance :)
i have seen more and more people posting things about, i was stable on 32bit and now on 64 i bsod and it makes me wonder was you really stable?
i have had 64bit os on my for years now and wouldnt use 32bit on it as i would lose some ram and i have gotten mine stable upto 4.4ghz
Yeah ive downloaded and will run when i get back from work :D
Well im Linx stable at 4.2 with 12GB of RAM on a 64Bit OS, so as far as im concerned im (well my PC is) stable :P
Im on 64Bit and im stable 4.1Ghz. That's a C2 Ph.II.
I was trying to get 3.8GHz stable yesterday and didn't manage it. 3.72GHz is stable 240 x 15.5 but couldn't get 200x19 = 3800MHz stable :( any tips master overclocker on voltages etc?
At 1.55v the CPU is at 42C. It will prime 95 (blend) for 10-15 minutes then go BSOD.
Here are my voltage settings:
http://img198.imageshack.us/img198/3244/dsc02663e.jpg
The first column is stock then the 2nd and 3rd are fails :p
I also tried 1.45 NB Vid, 1.2 NB, 1.26 SB and 1.55 vcore
All at 200 x 19 = 3800MHz
Hmmmmmmmmmm... Stop PM'ing me Will! ..Jokes :P
What are you trying to achieve? Just a clock of 3.8Ghz stable or everything?
If just a stable 3.8Ghz, i suggest drop the multi/speed of the HT & NB down so that they stay around 2000Mhz mark with no voltage increase. Thats a good start. Thats if you have OC's using the BUS and not just the CPU Multi alone. I find it better to OC using the BUS more than the Multi. Play around a bit with it more.
Set the Vcore at 1.475v. Thats a nice level and see how far you can go stable from that. Try 3.6Ghz at that first with OCCT for 15mins (rough guide). If it passes, try 3.7Ghz with the 1.475v and let me know.
Hmmmmmmmmmm... Stop PM'ing me Will! ..Jokes :P
What are you trying to achieve? Just a clock of 3.8Ghz stable or everything?
If just a stable 3.8Ghz, i suggest drop the multi/speed of the HT & NB down so that they stay around 2000Mhz mark with no voltage increase. Thats a good start. Thats if you have OC's using the BUS and not just the CPU Multi alone. I find it better to OC using the BUS more than the Multi. Play around a bit with it more.
Set the Vcore at 1.475v. Thats a nice level and see how far you can go stable from that. Try 3.6Ghz at that first with OCCT for 15mins (rough guide). If it passes, try 3.7Ghz with the 1.475v and let me know.
I want the maximum performance from my CPU/NB
It's 3.72Ghz stable at 1.51vcore. I will get back to you on the 1.475vcore. Doesn't the NB Vid help overclocking?
I cant get mine past 3.5ghz, with extra core enabled or not...
OK ran the bench, straight out the box no tweaks and ............
http://img51.imageshack.us/img51/8638/capturecopy.png
No good if thats good bad or indifferent TBH ?
I cant get mine past 3.5ghz, with extra core enabled or not...
Kinda reminds me of my old retarded CPU that wouldn't pass 3.2Ghz :lol: sold it on the bay for £75 ;)
Even if you just increase multi, you still need to raise CPU NB voltage, this was proven by Gareth who managed 3.8GHZ on stock voltage on his 1.4V vid C3, but only after raising the CPU NB voltage.
And 1.5V is seriously high for 3.7GHZ =/.
Try it at 1.45V at 3.6GHZ with 1.3V CPU NB voltage at 2.4GHZ and see what happens after 10 runs of IBT maximum.
Even if you just increase multi, you still need to raise CPU NB voltage, this was proven by Gareth who managed 3.8GHZ on stock voltage on his 1.4V vid C3, but only after raising the CPU NB voltage.
And 1.5V is seriously high for 3.7GHZ =/.
Try it at 1.45V at 3.6GHZ with 1.3V CPU NB voltage at 2.4GHZ and see what happens after 10 runs of IBT maximum.
I have a C2 phenom.
I'll tell you whats been stable with leaving the CPU-NB voltage at 'auto'.
3.625GHz @ 1.45v NB@2500MHz
3.72GHz @ 1.51v NB@2400MHz
Ok non of what you guys said seemed to work...tried 1.55v on CPU and CPU-NB and can't get 3.8GHz.
Heres some of my settings:
http://img94.imageshack.us/img94/4503/dsc02665x.jpg
Could be the board buddy.
Could be the board buddy.
I would blame the chip? It's only a handful that can reach 3.75GHz and you never see any at 3.8GHz - 4GHz (bar those on water). It's a 790FX board...top of the range :D
What are all your voltage settings? Could you take a screen shot on AMD overdrive?
file:///C:/Users/Will/AppData/Local/Temp/moz-screenshot-2.png
C2/C3 makes no difference, rules are the same.
Actually, could just be your chip.
HT 2.4GHZ with maxed SB Voltage.
CPU NB 2.6GHZ with 1.5V on the CPU NB
19 multi with 1.55V on the core
If that doesn't work it's the chip.
C2/C3 makes no difference, rules are the same.
ok then but yours running on 4Ghz is a 600Mhz increase. Mine running at 3.7GHz is a 900MHz increase...isn't it probable that my chip has hit the ceiling?
ok then but yours running on 4Ghz is a 600Mhz increase. Mine running at 3.7GHz is a 900MHz increase...isn't it probable that my chip has hit the ceiling?
Chip increase amount doesn't mean much tbh lol, as they're the same chip, just different multi's/ voltage.
Then again, my old C2 940 maxed 3.7.
Chip increase amount doesn't mean much tbh lol, as they're the same chip, just different multi's/ voltage.
Then again, my old C2 940 maxed 3.7.
I would expect your chip to have a higher binning process (i think that's the technical term). They don't do it like this chip can run at 3.4GHz it will be boxed as a Phenom II 965 cause then they've have really bad overclocking chips if it's maximum speed was 3.4GHz. All the 965 chips are ones capable of doing 4GHz or whatever.
I would blame the chip? It's only a handful that can reach 3.75GHz and you never see any at 3.8GHz - 4GHz (bar those on water). It's a 790FX board...top of the range :D
What are all your voltage settings? Could you take a screen shot on AMD overdrive?
file:///C:/Users/Will/AppData/Local/Temp/moz-screenshot-2.png
I managed 3.85Ghz stable on my first 720 BE back in Jan last year. That was on a crappy £40 board with an old nForce chipset. That was 1.5V on the Vcore, 1.4V NB. I was using a TRUE Black.
I managed 3.85Ghz stable on my first 720 BE back in Jan last year. That was on a crappy £40 board with an old nForce chipset. That was 1.5V on the Vcore, 1.4V NB. I was using a TRUE Black.
This is what my overclocking screen looks like.
http://i.neoseeker.com/neo_image/162285/article/asus_m3a79-t_deluxe/bios0006.jpg
That's not my actual settings though.
Is that Northbridge (NB) or CPU-NB? I never increase my northbridge voltage unless it's going to go past 2500MHz. Increasing my CPU-NB voltage seems to have no affect.
What clock settings were you using (e.g.200 x 19)
Does the page in the BIOS do anything? I've been told it helps in overclocking...no idea what any of it means though.
http://i.neoseeker.com/neo_image/162303/article/asus_m3a79-t_deluxe/bios0015.jpg
Its because i had the RAM around 2600 i think. I really don't know buddy. Was a year ago lol. I recently had another 720 but was just for benching. Didn't see properly what was 24/7 stable.
Its because i had the RAM around 2600 i think. I really don't know buddy. Was a year ago lol. I recently had another 720 but was just for benching. Didn't see properly what was 24/7 stable.
Can you make any sense of those screenshots? Snigger see previous page (I did a ninja edit :p)
Why are you FSB overclocking on a BE?
FSB to 200.
Proccesor frequency multiplier to 19.
CPU NB HT @ 2600
CPU nothbridge frequency at 2600
Take the South bridge voltage to the 1.8V
Take the proccesor northbridge voltage to 1.55v
Take the processor voltage to 1.55v
Up the nothbridge voltage slightly.
My southbridge maxes out at 1.4v and I never take it back 1.28 (stock is 1.2).
Prime 95 is 15 minutes stable on blend but then goes BSOD...stable enough for any benchmark but not 100% stable though.
Take the SB to 1.4V then.
Take the SB to 1.4V then.
That's silly :lol: I don't want fireworks coming out for my box :p didn't u know that all SBs aren't the same?!
niether are cpu or mch or nb but people still ask other people for theit settings, funny that.
truth is will you can ask as many people as you like but nobody will know the right settings unless they play/test with your system. even somebody with the exact same system could only advise as their settings will be slightly different to what yours needs.
you want to know the secret to overclocking? change setting, test setting, bsod, repeat until stable.
nothing works better than trail and error.
niether are cpu or mch or nb but people still ask other people for theit settings, funny that.
truth is will you can ask as many people as you like but nobody will know the right settings unless they play/test with your system. even somebody with the exact same system could only advise as their settings will be slightly different to what yours needs.
you want to know the secret to overclocking? change setting, test setting, bsod, repeat until stable.
nothing works better than trail and error.
I want to know what settings do, like the ones in the screen shots I posted. There's no point me wasting my time with options that change the DDR voltage or something (that's obvious what it does but you know what I mean).
well post what options you have and we will google the answer for you :)
This is what my overclocking screen looks like.
http://i.neoseeker.com/neo_image/162285/article/asus_m3a79-t_deluxe/bios0006.jpg
That's not my actual settings though.
Is that Northbridge (NB) or CPU-NB? I never increase my northbridge voltage unless it's going to go past 2500MHz. Increasing my CPU-NB voltage seems to have no affect.
What clock settings were you using (e.g.200 x 19)
Does the page in the BIOS do anything? I've been told it helps in overclocking...no idea what any of it means though.
http://i.neoseeker.com/neo_image/162303/article/asus_m3a79-t_deluxe/bios0015.jpg
These :)
top 5 are easy, you should know those.
vcore yep
cpu-nb is your fsb so as you up the fsb this will need upping too, at some point.
vdimm yea
nb/sb volts yea
vdda well this is what i found on that from google
basically what happens to the voltage when a load is applied on the CPU is the voltage actually drops. so if you set your voltage to say 1.45v in BIOS and the CPU VDDA is set to it's lowest setting, under idle conditions, the voltage should be around 1.45v. however, under load, the voltage will actually drop below your setting of 1.45v. how much depends on the make and model of your motherboard.
setting the CPU VDDA to it's highest setting will cause the CPU voltage to be (alot) more than what you set it to in BIOS. for instance, if i set my CPU VDDA voltage to 2.80v (Auto, 2.5, 2.6, 2.7, and 2.8v are the settings on my ATI chipset), under idle conditions, my CPU voltage will be about 0.04-0.05v more than what i set it to in the BIOS. while under load, it can go as high as 0.10v over my BIOS settings.
it does this to make sure that your CPU is always getting a good amount of power so that the power doesn't dip down and cause BSOD's. usually setting it to AUTO is fine as it will use the lowest settings while under idle and use the highest settings while under load. this makes it so that your not wasting too much power and/or over-volting your chip at idle while making sure under load, you're receiving enough power.
sounds like llc on my board but llc dont overvolt like this can from what i make of that.
auto xpress sounds like an auto oc option
ht tristate seems to be like an auto option for the ht multi so it can drop if its getting too high
the last 2 seem to be to do with the ram but i can find anything that says exaclty what they do.
please note these are from a quick google and maybe complete **** :D
Had my 920 at 4 ghz but the added volts heat not worth it for 24/7.....full time @3.8.
Got an i5 at 4ghz so i might play too :D
I could have 4 ghz HT off or 3.8 with 8 threads.....a no brainer for crunching.
My I7 laughs at 4ghz so i will play :lol:
most of the more modern cpu can do 4ghz no problem, if you tweak them abit, and thats why i thought this would be a good idea.
other than 3 tests we have (super pi, custom pc suite and cine bench thingie) what do you want to run?
Andrew Moore
18-01-10, 22:57
3dmark vantage gives a cpu score
as does PCmark.
You can also run tests in the sisoftware sandra suite,
Andy
other than 3 tests we have (super pi, custom pc suite and cine bench thingie) what do you want to run?
Boinc or Folding :D
PeterStoba
19-01-10, 09:59
most of the more modern cpu can do 4ghz no problem, if you tweak them abit, and thats why i thought this would be a good idea.
other than 3 tests we have (super pi, custom pc suite and cine bench thingie) what do you want to run?
wprime or maxxmem
imo the "mark" tests are all way too intel biasd, sandra could be interesting?
passmark i dont know.
crunching and folding, well not really a bench but if you want to f@h for aria its team 133107 :D
wprime, well i think its that or super pi, not sure we need both, maxxmem i dont know either.
This would be a piece of cake on my old P4 Extreme but I wouldn't be able to hear myself think. :D Good luck everything, I'm too scared/skint to try this. ;)
busman3000
19-01-10, 17:53
I doubt I could get my 720 BE to 4ghz without upping the volts by alot.
PeterStoba
19-01-10, 18:07
imo the "mark" tests are all way too intel biasd, sandra could be interesting?
passmark i dont know.
crunching and folding, well not really a bench but if you want to f@h for aria its team 133107 :D
wprime, well i think its that or super pi, not sure we need both, maxxmem i dont know either.
Maxxmem is more memory dependant, could be useful in a different thread.
passmark i dont know.
http://www.passmark.com/products/pt.htm
I've got my E8500 at 4.5GHz, and I got it there without too much trouble at all.
Though that's all down to luck... my overclocking skill is somewhat lacking. :D
if its stable at 4ghz would you bench it too so we have a dual core's results for the list too?
if its stable at 4ghz would you bench it too so we have a dual core's results for the list too?
I'm sorry, are you talking to me? :o
yea m8, sry :o i dont quote as much as i should :o
Sure, I'll do so some benchmarks at 4GHz for you tomorrow.
I'd do it now, but I can't be bothered rebooting, etc. I have an essay to do. :p
no worries m8, not nailed down what test we should be running yet as we dont want too much of an intel bias but as yet we aint had many from the green team posting what we should be running :(
Andrew Moore
20-01-10, 01:33
Wprime?
Andy
i think it was stoba who mentioned wprime earlier in the thread and i dont see the point in running it and super pi but as yet we have had no discussion on the pros or cons or either??
Theres already superpi being run so seems pointless running another just for people to run at 4ghz.
wprime? no idea what it is tbh, sounds like a stress testing thing more than a benchmarking thing??
PeterStoba
20-01-10, 07:52
Theres already superpi being run so seems pointless running another just for people to run at 4ghz.
wprime? no idea what it is tbh, sounds like a stress testing thing more than a benchmarking thing??
It's basically a multi threaded super pi that scales well with cores.
I think we'd be better off with SiSoft or something.
A bit of natural subzero cooling, check out his outside overclocking thread.
i think it was stoba who mentioned wprime earlier in the thread and i dont see the point in running it and super pi but as yet we have had no discussion on the pros or cons or either??
Well... Since wprime is multi threaded it would mean different categories per number of cores, it also doesn't react to ram timings and nb speeds as drastically as say super pi will ;)
You could always do more than one benchmark but leaving them all single threaded like memmark pifast spi etc( or even make it a requirement to run just 1 thread on wprime) it will enable everyone to compete in just 2 seperate catergories. AMD and Intel :popcorn:
A bit of natural subzero cooling, check out his outside overclocking thread.
He didn't get 4GHZ in that attempt I don't think?
And even then, he can't really claim 4GHZ when it's not stable, and since he's went above the 1.55V he's killed his warranty (Strictly speaking, a CPU that doesn't go above the voltage set by AMD/Intel, regardless of overclocked or not will be allowed to be RMA'ed).
Doing tests at 4GHZ to join in this would do more harm than good.
He didn't get 4GHZ in that attempt I don't think?
And even then, he can't really claim 4GHZ when it's not stable, and since he's went above the 1.55V he's killed his warranty (Strictly speaking, a CPU that doesn't go above the voltage set by AMD/Intel, regardless of overclocked or not will be allowed to be RMA'ed).
Doing tests at 4GHZ to join in this would do more harm than good.
But your already at 4ghz :rolleyes: I have never killed any hardware through benching and I don't know anyone who has without doing something really daft. Usually being condensation on the board :drool: or dribble.
Theres already superpi being run so seems pointless running another just for people to run at 4ghz.
totally agree with you there, so we need some that dont have their own thread here yet, (not saying those we use wont get their own later if people want to show how they do in it higher tho :D) so thats some we wont use :)
Well... Since wprime is multi threaded it would mean different categories per number of cores, it also doesn't react to ram timings and nb speeds as drastically as say super pi will ;)
You could always do more than one benchmark but leaving them all single threaded like memmark pifast spi etc( or even make it a requirement to run just 1 thread on wprime) it will enable everyone to compete in just 2 seperate catergories. AMD and Intel :popcorn:
this is the kicker, i would like to try and keep it simple with 1 overall table and then 1 for each make for each test, for this i want to try and find test that cover all the bases so that a semp can compeate with a phenom in some runs (single thread test) even if it gets murdered in another (multi threaded one) with the same for all dual tri core cpu too.
maybe we just run the sandra stuff and rename the thread? but i would like to keep it to 4 tests (maybe 5) other wise it will be a mare to organise/maintain.
But your already at 4ghz :rolleyes: I have never killed any hardware through benching and I don't know anyone who has without doing something really daft. Usually being condensation on the board :drool: or dribble.
My 4GHZ uses 1.5V for a kick off, has been tested for 20 runs of IBT maximum and passes stable.
But I'm not justifying my 4GHZ.
Point being, Will's 4GHZ is unstable, running hot if he's in doors when running, it'll do more harm than good.
But, it's his equipment, not mine.
has he claimed it is stable?
i mean my sig show mine @4.5ghz but i aint ever said it was stable above 4.4ghz for eg.
has he claimed it is stable?
i mean my sig show mine @4.5ghz but i aint ever said it was stable above 4.4ghz for eg.
He hasn't, and as far as I know I never said he claimed it stable, or at least I never meant to.
Interpretation is different to everyone I suppose.
Interpretation is different to everyone I suppose.
Exactly, personally i never bother priming for more than an hour or so absolute max as all i use it for is temps. I have better things to do than run that for hours on end.
For me "stable" means it doesn't crash doing what i want it to do, at the end of the day i'm not trying to put a man on the moon or running some battle ship so insisting on 8h prime stable is pointless imho (or even 24h for some really anal people on certain forums i could mention).
Exactly, personally i never bother priming for more than an hour or so absolute max as all i use it for is temps. I have better things to do than run that for hours on end.
For me "stable" means it doesn't crash doing what i want it to do, at the end of the day i'm not trying to put a man on the moon or running some battle ship so insisting on 8h prime stable is pointless imho (or even 24h for some really anal people on certain forums i could mention).
I personally say 20 runs of IBT maximum is enough to test stability.
If it passes, it's stable.
That'll take less than an hour dependant on RAM.
im the same, i have better things to waste me power on.
if ibt passes 5 runs on max only the heat will take it down, now that could be cpu, mobo or psu as they all get warm fully loaded for prolonged periods but what do you run in normal use that places that ammount of demand on your system?
hardest my pc works is when i get a guilt pang and i run f@h for a couple of days and even that doesnt get my system as hot after a day than linx can in 10 minutes.
I personally say 20 runs of IBT maximum is enough to test stability.
If it passes, it's stable.
That'll take less than an hour dependant on RAM.
^^ The best way of testing, 5 or 10 passes isnt enough as 'almost' stable usually locks up around 15 passes ;)
I have never used prime for more than about 30mins as it is a waste of space. how/why anyone can run that for over 8 hours is beyond me.
Why has my outdoor overclocking been bought into this :p
http://img697.imageshack.us/img697/1074/3dmarkvantage4ghz155vco.png
That's from the thread...3Dmark vantage score
Impressive, I'm going to assume that was outside?
Will wanna make that larger?? i cant view it on my 90" projector lol
Will wanna make that larger?? i cant view it on my 90" projector lol
Tis tiny for me :D
Will wanna make that larger?? i cant view it on my 90" projector lol
stop moaning :lol:
look here Martin see here
http://forums.aria.co.uk/showthread.php?t=13946
there
you'll see the same 3Dmark screenshot :)
stop moaning :lol:
look here Martin see here
http://forums.aria.co.uk/showthread.php?t=13946
there
you'll see the same 3Dmark screenshot :)
These old eyes of mine arnt what they used to be ;)
theelusiveyoda
22-01-10, 16:56
can i join in i managed to get my q9450 to 4.5ghz on water but not stable at that but at 4.2ghz all is fine with a fsb of 530mhz. i will get some screenshots up in a min when i've finished installed windows 7 on my laptop.
thats good going from a 9450 that m8 :D
all you need is 4ghz, nothing higher but as soon as i/we get which tests we shall be running i will put another thread up with dl links and results tables.
so what do you think we should run?
I take it this is for cpu tests only?
yea m8, looking for test that are as cpu bound as pos tbh.
So are we benching?, are we?, are we? :P
I3R0K3N7FEET
02-02-10, 02:38
so what was the point in this thread? :P hitting 4ghz? would be good if it was an AMD 4ghz challenge thread :P makes more sense really
if we can get a couple of tests nailed down to show just cpu power with both single and multi thread results i will start a thread for submissions.
we have has a few named but i have been short on time to look into those as of yet to see if they do fit the bill :(
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.