ScienceMark thread (with score list)

[Alec§taar]

CURRENT STANDINGS AS OF 08/07/2006:

Devious = 2181.97 (Intel Core2Duo E6600@3942mhz)
MikeLopez = 1700.24 (via 7x mult. X 428mhz FSB - Intel Core2Duo E6300@2996mhz)
Urlyin = 1537.93 (175 Dual Core Opteron @2800mhz)
Dj Dn = 1446.18 (Opteron 165 2700mhz)
Tigger69 = 1431.51 (Pentium D930@4385mhz)
Alec§taar = 1414.29 (via 11x mult. X 251mhz FSB: AMD ATHLON 64X2 4800+ CPU@2771mhz/RAM@198/200mhz)
_33 = 1395.99 (AMD ATHLON 64 3000+@2856MHZ)
Lekamies = 1371.23 (AMD ATLHON 64 3000+@2700mhz)
Steevo = 1360 (AMD 4000+ San Diego @ 3.0 1.65vcore)
JNT Raptor = 1334.32 (AMD ATHLON X2 3800+@2616mhz)
pt = 1256.23 (AMD ATLHON 64 3000+@2500mhz)
Ketxxx = 1140.97 (A64 3200+ @ 2.55GHz - Winchester)
Canuto = 1112.11 (Pentium D 930 @ 3300mhz)
cdawall = 1036.22 (AMD ATHLON 64 3000+@2440mhz)
Polaris573 = 1001.86 (P4 Northwood@3.2GHz)
Solaris17 = 886.72 (AMD ATHLON XP@2340MHZ)



* There ya are: Updated (w/ Canuto's updated/better score, 175 points better + Ketxxx's first new one!)...

APK

 

[pt]

Look like this thread was still alive

 

[Ketxxx]

CURRENT STANDINGS AS OF 08/07/2006:
Ketxxx = 1140.97 (A64 3200+ @ 2.55GHz - Winchester)




* There ya are: Updated (w/ Canuto's updated/better score, 175 points better + Ketxxx's first new one!)...

APK

take note of earlier post, 1140 was everything completely stock

I'm working it but I gotta admit, I don't really like SCM2.0. Its WAY too dependent on the CPU when there are so many other factors that are just as if not more important than the sheer GHz of the CPU in creating an efficient system. Such as memory subsystem, primary and alpha memory timings, HTT link and FSB speed, memory latency, the list goes on.

 

[Alec§taar]

take note of earlier post, 1140 was everything completely stock

So, you're saying your CPU spec from your signature is not how it is? I can change the result CPU type statement if you wish & state it is "stock timings" etc.??

Is this acceptable to you???

I'm working it

I would too, & DID for my score... but, I don't do anything I can't run stable, personally.

It's nice to get a HIGH benchmark score - but nicer knowing you set your system up in such a manner, it is "set it & forget it" easy from then on & NO MORE "messing around" due to finding instabilities in various apps!

In fact, my main one for this (above & beyond THIS one we use as a benchmark here), which acts as my "canary in a coalmine" against instabilities from overclocks...?

It is Folding@Home, actually! Its console mode/tty terminal/Character mode commandline driven model is VERY "sensitive" to overclock instability when used w/ it's "-forceasm" option...

I don't really like SCM2.0. Its WAY too dependent on the CPU when there are so many other factors that are just as if not more important than the sheer GHz of the CPU in creating an efficient system. Such as memory subsystem, primary and alpha memory timings, HTT link and FSB speed, memory latency, the list goes on.

Oh, it's a "heavy hitter" on the memory end... you'll find THAT out, nearly guaranteed!



* Which is WHY I was in favor of using it not only as a benchmark score, but as yet another "stability-test" as well... others here feel the same & have seconded that motion!

(Good program - 2 for the price of 1 (benchmark AND stability test))

APK

P.S.=> And, you're absolutely correct about there being more factors in play, as well as more you do on a PC than run benchmarks (this is certain) because where one person or CPU-type/RAM/mobo makeup can beat you in a particular test? You'll maybe lose in another like diskbound I/O, etc.... apk

 

[pt]

Tomorrow i get my new case so i'm gonna overclock it a bit more (my actual case has no front intake fan) and in September i will probabily put watercooling on CPU , hope to get to 3.0ghz

 

[Ketxxx]

So, you're saying your CPU spec from your signature is not how it is? I can change the result CPU type statement if you wish & state it is "stock timings" etc.??

Is this acceptable to you???



I would too, & DID for my score... but, I don't do anything I can't run stable, personally.

It's nice to get a HIGH benchmark score - but nicer knowing you set your system up in such a manner, it is "set it & forget it" easy from then on & NO MORE "messing around" due to finding instabilities in various apps!

In fact, my main one for this (above & beyond THIS one we use as a benchmark here), which acts as my "canary in a coalmine" against instabilities from overclocks...?

It is Folding@Home, actually! Its console mode/tty terminal/Character mode commandline driven model is VERY "sensitive" to overclock instability when used w/ it's "-forceasm" option...



Oh, it's a "heavy hitter" on the memory end... you'll find THAT out, nearly guaranteed!



* Which is WHY I was in favor of using it not only as a benchmark score, but as yet another "stability-test" as well... others here feel the same & have seconded that motion!

(Good program - 2 for the price of 1 (benchmark AND stability test))

APK

Indeed thats fine, I always revert to stock with anything new I do so I have a clear base. My OC results arent too bad so far coming in at 13xx. I also played with alpha timings and got squat in terms of a better result, and looking at the tests run, their not tests that particularly stress much of the system - just the CPU. If the test really factored the memory subsystem to the level and degree it should do then everybody would note a significant increase in their SCM2 score when OCd. I shot my bandwidth from a stock of 5.55GB\s to 6.88GB\s, and want to know how much the SCM2 score went up by? Nothing. Zip. Nada. Zilcho.

Thats why I dont like SCM2 I know I look at things in far too much detail, but from a benchmark thats supposed to be scientific, as such the system would be being used as efficently as it possibly can, then to see it isnt, SCM2 deserves all the slating it gets.

 

[Alec§taar]

Tigger69 - "I'm coming @ you man!" lol

CPU-z 1.35 CPU-mhz (2787 mhz):

&

CPU-z 1.35 RAM-mhz (199.8/200mhz):

* Check it out (got more outta it, but it is due to my running this test in the EARLY a.m. here, when temps are very low, imo @ least so far):

ScienceMark 2.0 new high score for me (1423.57):

BIOS DATA:

CPU HTT Frequency = 4x

DRAM CONFIGURATION:

CAS# Latency Tcl = 2
Min RAS# Active Time Tras = 5T
RAS# to CAS# Delay Trcl = 3T
Row Precharge Time Trp = 3T
Row Cycle Time Trc = 10T
Row Refresh Cycle Time Trk = 13T
Read to Write Trwt = 3T
Write Recovery Time = 3T
1t/2t = 1t

JUMPER FREE CONFIGURATION DATA:

O/C Profile = Manual
CPU Frequency = 231mhz
PCI Clock = 100mhz
DDR Voltage = AUTO (cannot do over 2.6v on my RAM, tried it already 2x)
Chipset Voltage = 1.6V (max)
HT Voltage = 1.25V (max)
CPU Mult = 12x
CPU Volt = 1.5V (max)

APK

P.S.=> I've noted that temps @ the time of the test seem to matter largely!

(I.E..-> When you take it, & what the surrounding "ambient temperature" is really does seem to matter)

E.G. -> When it was 100++F outside during the heatwave of summer 2006 earlier this month?

Heck - There was NO WAY I could pump out a score like that... none!

Now, it is cooler outside, around 75F?? Well, "suddenly" now? I CAN!... apk

 

[Alec§taar]

Indeed thats fine

Will do, I will note your score as "STOCK" mhz settings in BIOS...

If the test really factored the memory subsystem to the level and degree it should do then everybody would note a significant increase in their SCM2 score when OCd. I shot my bandwidth from a stock of 5.55GB\s to 6.88GB\s, and want to know how much the SCM2 score went up by? Nothing. Zip. Nada. Zilcho.

Probably correct - POINT Mr. Ketxxx!

Thats why I dont like SCM2 I know I look at things in far too much detail, but from a benchmark thats supposed to be scientific, as such the system would be being used as efficently as it possibly can, then to see it isnt, SCM2 deserves all the slating it gets.

I like it because it either makes it THRU a test, or breaks-down, indicative to me @ least that I may be "putting on the straws that break the camel's back" etc. (Folding @ Home's another one that gets "nuts" IF & when I 'push too hard' etc.)... they're what I said for me @ least:

My "canaries in a coalmine" & IF they 'mess up' &/or don't finish? Step back a notch!

APK

 

[Ketxxx]

Not bad. Your RAM is really holding you back tho. With a nice 1:1 ratio and DDR500+ your score would be screaming along at 1700pts+ probably.

 

[Alec§taar]

MikeLopez = 1700.24 (via 7x mult. X 428mhz FSB - Intel Core2Duo E6300@2996mhz)
Urlyin = 1537.93 (175 Dual Core Opteron @2800mhz)
Dj Dn = 1446.18 (Opteron 165 2700mhz)
Tigger69 = 1431.51 (Pentium D930@4385mhz)
Alec§taar = 1423.57 (via 11x mult. X 253mhz FSB: AMD ATHLON 64X2 4800+ CPU@2771mhz/RAM@199.8/200mhz)
_33 = 1395.99 (AMD ATHLON 64 3000+@2856MHZ)
Lekamies = 1371.23 (AMD ATLHON 64 3000+@2700mhz)
Steevo = 1360 (AMD 4000+ San Diego @ 3.0 1.65vcore)
JNT Raptor = 1334.32 (AMD ATHLON X2 3800+@2616mhz)
pt = 1256.23 (AMD ATLHON 64 3000+@2500mhz)
Ketxxx = 1140.97 (A64 3200+ @ stock mhz - Winchester)
Canuto = 1112.11 (Pentium D 930 @ 3300mhz)
cdawall = 1036.22 (AMD ATHLON 64 3000+@2440mhz)
Polaris573 = 1001.86 (P4 Northwood@3.2GHz)
Solaris17 = 886.72 (AMD ATHLON XP@2340MHZ)



* There ya are: Updated (w/ MY updated/better score, 175 points better + Ketxxx's first new one @ his STOCK cpu mhz speed!)...

APK

 

[Alec§taar]

Not bad.

Yup, she's "getting there"... so far @ least - it passed thru this, but my REAL "test of fire" is Folding@Home consolemode app w/ it's "-forceasm" switch... touchy as HELL w/ overclocks!

Your RAM is really holding you back tho. With a nice 1:1 ratio and DDR500+ your score would be screaming along at 1700pts+ probably.

Oh, I wholeheartedly agree & take your word for it + sort of have to:

Fact is, you're one of the guys here (w/ POGE & Tatty One) that showed me how to blow past 2.592mhz (farthest I took it) & look @ my system now: Punching CLOSE to 2800mhz!

(AND, great RAM speed returns -> 199.1/200mhz DDR-400 rated speed... & I think the tip on the TRas you gave me is helping here...)

* Trust me - I am NOT one to complain @ this point!

APK

P.S.=> It makes me wonder what this machine would/could do, with better RAM... you know? apk

 

[Ketxxx]

just give me your X2 and we can see how far she can go

 

[Alec§taar]

just give me your X2 and we can see how far she can go

Heh, I ought to "crank up" my air-conditioner, & stick a duct around it RIGHT INTO MY PC, lol, & set it to do 65F temperature or lower, & THEN SEE WHAT IT CAN DO!

(LOL!)



* That & better RAM? Yea... it'd be faster!

APK

 

[Alec§taar]

CPU-z 1.35 CPU-mhz (2787 mhz):

&

CPU-z 1.35 RAM-mhz (199.8/200mhz):

* Check it out (got more outta it, but it is due to my running this test in the EARLY a.m. here, when temps are very low, imo @ least so far):

ScienceMark 2.0 new high score for me (1423.57):

BIOS DATA:

CPU HTT Frequency = 4x

DRAM CONFIGURATION:

CAS# Latency Tcl = 2
Min RAS# Active Time Tras = 7T
RAS# to CAS# Delay Trcl = 3T
Row Precharge Time Trp = 3T
Row Cycle Time Trc = 10T
Row Refresh Cycle Time Trk = 13T
Read to Write Trwt = 3T
Write Recovery Time = 3T
1t/2t = 1t

JUMPER FREE CONFIGURATION DATA:

O/C Profile = Manual
CPU Frequency = 253mhz
PCI Clock = 100mhz
DDR Voltage = AUTO (cannot do over 2.6v on my RAM, tried it already 2x)
Chipset Voltage = 1.6V (max)
HT Voltage = 1.25V (max)
CPU Mult = 11x
CPU Volt = 1.5V (max)

APK

P.S.=> Ketxxx, now that I have you here? Take a peek @ the stuff under "BIOS DATA" & if you see ANYTHING ELSE you think needs an adjustment?? Please, DO let me know... wanna try to beat Tigger69's Pentium D score in ScienceMark, lol... apk

 

[Alec§taar]

Ketxxx, yet another increase: Please check "BIOS DATA" & give me a tip if is possible

CPU-z 1.35 CPU-mhz data (2797mhz):

&

CPU-z 1.35 RAM-mhz data (199.8/200mhz):

ScienceMark 2.0 score:

BIOS OVERCLOCKING DATA (for reference):

CPU HTT Frequency = 4x

DRAM CONFIGURATION:

CAS# Latency Tcl = 2
Min RAS# Active Time Tras = 7T
RAS# to CAS# Delay Trcl = 3T
Row Precharge Time Trp = 3T
Row Cycle Time Trc = 10T
Row Refresh Cycle Time Trk = 13T
Read to Write Trwt = 3T
Write Recovery Time = 3T
1t/2t = 1t

JUMPER FREE CONFIGURATION DATA:

O/C Profile = Manual
CPU Frequency = 253mhz
PCI Clock = 100mhz
DDR Voltage = AUTO (cannot do over 2.6v on my RAM, tried it already 2x)
Chipset Voltage = 1.6V (max)
HT Voltage = 1.25V (max)
CPU Mult = 11x
CPU Volt = 1.5V (max)

APK

P.S.=> Ketxxx, (editing) GOT YOUR TIPS IN MY O/C THREAD, thanks.. wanna try to beat Tigger69's Pentium D 1431 score in ScienceMark, lol... apk

 

[Alec§taar]

CURRENT STANDINGS AS OF 08/07/2006:

Devious = 2181.97 (Intel Core2Duo E6600@3942mhz)
MikeLopez = 1700.24 (via 7x mult. X 428mhz FSB - Intel Core2Duo E6300@2996mhz)
Urlyin = 1537.93 (175 Dual Core Opteron @2800mhz)
Dj Dn = 1446.18 (Opteron 165 2700mhz)
Tigger69 = 1431.51 (Pentium D930@4385mhz)
Alec§taar = 1426.29 (via 11x mult. X 254mhz FSB: AMD ATHLON 64X2 4800+ CPU@2771mhz/RAM@199.8/200mhz)
_33 = 1395.99 (AMD ATHLON 64 3000+@2856MHZ)
Lekamies = 1371.23 (AMD ATLHON 64 3000+@2700mhz)
Steevo = 1360 (AMD 4000+ San Diego @ 3.0 1.65vcore)
JNT Raptor = 1334.32 (AMD ATHLON X2 3800+@2616mhz)
pt = 1256.23 (AMD ATLHON 64 3000+@2500mhz)
Ketxxx = 1140.97 (A64 3200+ @ 2.55GHz - Winchester)
Canuto = 1112.11 (Pentium D 930 @ 3300mhz)
cdawall = 1036.22 (AMD ATHLON 64 3000+@2440mhz)
Polaris573 = 1001.86 (P4 Northwood@3.2GHz)
Solaris17 = 886.72 (AMD ATHLON XP@2340MHZ)



* There ya are: Updated (w/ MY updated/better score, 13 points better than before)...

APK

P.S.=> I am going to try to beat Tigger69's 1431... apk

 

[Ketxxx]

Let me know if those tweaks work short of those there really isnt anymore to do apart from squeeze the MHz out that CPU, another 4MHz and I think u'll have him

 

[Canuto]

Today i'm gonna try 3.6Ghz i'll post them later.

 

[Alec§taar]

Let me know if those tweaks work short of those there really isnt anymore to do apart from squeeze the MHz out that CPU, another 4MHz and I think u'll have him

Well, they seemed to have actually "slowed it up" some, I got a 1409 score on ScienceMark 2.0 using them & no biggie (this stuff is ALL "try, try again" experimentation anyhow @ this point imo @ least)...

Well, yet again I redid them!

I got a BETTER score still than any yet (1428++), by using settings "in-between" what you outlined as guidelines!

(See, I don't expect you to be a "magician" or "Jedi Knight" using the force to get the EXACT numbers for it remotely - probably impossible, but you DO give me GOOD solid guidelines)!

CPU-z 1.35 CPU-mhz (2798mhz):

&

CPU-z 1.35 RAM-mhz (199.8/200mhz):

ScienceMark 2.0 Score @ those settings above, & BIOS settings below:

BIOS OVERCLOCKING DATA (for reference):

CPU HTT Frequency = 4x

DRAM CONFIGURATION:

CAS# Latency Tcl = 2
Min RAS# Active Time Tras = 8T (up from 7T)
RAS# to CAS# Delay Trcl = 3T
Row Precharge Time Trp = 3T
Row Cycle Time Trc = 12T (up from 10T)
Row Refresh Cycle Time Trk = 13T
Read to Write Trwt = 2T (down from 3T)
Write Recovery Time = 2T (down from 3T)
1t/2t = 1t

JUMPER FREE CONFIGURATION DATA:

O/C Profile = Manual
CPU Frequency = 254mhz
PCI Clock = 101mhz (upped from 100mhz default)
DDR Voltage = AUTO (cannot do over 2.6v on my RAM, tried it already 2x)
Chipset Voltage = 1.6V (max)
HT Voltage = 1.25V (max)
CPU Mult = 11x
CPU Volt = 1.5V (max)

TECHPOWERUP FORUMS SCIENCEMARK 2.0 BENCHMARK CURRENT STANDINGS AS OF 08/08/2006:

Devious = 2181.97 (Intel Core2Duo E6600@3942mhz)
MikeLopez = 1700.24 (via 7x mult. X 428mhz FSB - Intel Core2Duo E6300@2996mhz)
Urlyin = 1537.93 (175 Dual Core Opteron @2800mhz)
Dj Dn = 1446.18 (Opteron 165 2700mhz)
Tigger69 = 1431.51 (Pentium D930@4385mhz)
Alec§taar = 1428.17 (via 11x mult. X 254mhz FSB: AMD ATHLON 64X2 4800+ CPU@2797mhz & RAM@199.8/200mhz)
_33 = 1395.99 (AMD ATHLON 64 3000+@2856MHZ)
Lekamies = 1371.23 (AMD ATLHON 64 3000+@2700mhz)
Steevo = 1360 (AMD 4000+ San Diego @ 3.0 1.65vcore)
JNT Raptor = 1334.32 (AMD ATHLON X2 3800+@2616mhz)
pt = 1256.23 (AMD ATLHON 64 3000+@2500mhz)
Ketxxx = 1140.97 (A64 3200+ @ 2.55GHz - Winchester)
Canuto = 1112.11 (Pentium D 930 @ 3300mhz)
cdawall = 1036.22 (AMD ATHLON 64 3000+@2440mhz)
Polaris573 = 1001.86 (P4 Northwood@3.2GHz)
Solaris17 = 886.72 (AMD ATHLON XP@2340MHZ)



* There ya are: Updated (w/ MY updated/better score, 2 points better than before)...

APK

P.S.=> Can't QUITE get to Tigger69's 1431, yet (wait until winter lol) & I am happy w/ my progress here in the ScienceMark 2.0 benchmark & stability tests here on these forums:

1390 -> 1401 -> 1414 -> 1423 -> 1428++...

IMO @ least - I'm not going to catch Tigger69 today, it's heating up outside by this point, but I am happy though & I think I am "peaking-out" as you say...

Tigger69's @ 1431, & now? I am ONLY 3 points behind, & I can live w/ that... lol! apk

 

[Alec§taar]

TECHPOWERUP FORUMS SCIENCEMARK 2.0 BENCHMARK CURRENT STANDINGS AS OF 08/08/2006:

Devious = 2181.97 (Intel Core2Duo E6600@3942mhz)
MikeLopez = 1700.24 (via 7x mult. X 428mhz FSB - Intel Core2Duo E6300@2996mhz)
Urlyin = 1537.93 (175 Dual Core Opteron @2800mhz)
Dj Dn = 1446.18 (Opteron 165 2700mhz)
Tigger69 = 1431.51 (Pentium D930@4385mhz)
Alec§taar = 1428.17 (via 11x mult. X 254mhz FSB: AMD ATHLON 64X2 4800+ CPU@2797mhz & RAM@199.8/200mhz)
_33 = 1395.99 (AMD ATHLON 64 3000+@2856MHZ)
Lekamies = 1371.23 (AMD ATLHON 64 3000+@2700mhz)
Steevo = 1360 (AMD 4000+ San Diego @ 3.0 1.65vcore)
JNT Raptor = 1334.32 (AMD ATHLON X2 3800+@2616mhz)
pt = 1256.23 (AMD ATLHON 64 3000+@2500mhz)
Ketxxx = 1140.97 (A64 3200+ @ 2.55GHz - Winchester)
Canuto = 1112.11 (Pentium D 930 @ 3300mhz)
cdawall = 1036.22 (AMD ATHLON 64 3000+@2440mhz)
Polaris573 = 1001.86 (P4 Northwood@3.2GHz)
Solaris17 = 886.72 (AMD ATHLON XP@2340MHZ)



* There ya are: Updated (w/ MY updated/better score, 2 points better than before)...

APK

P.S.=> Can't QUITE get to Tigger69's 1431, yet (wait until winter lol) & I am happy w/ my progress here: 1390 -> 1401 -> 1414 -> 1423 -> 1428++... apk

 

[Ketxxx]

Right you found absolute best timings it seems now, so from here, eek that FSB just 2-4MHz and you will have him

ed- oh yea i bought that venice off tatty_one today too, cheques in the mail so ill probably stab at SCM2 when i get it just for the hell of it. my aim will be 1500pts.

 

[Alec§taar]

Right you found absolute best timings it seems now, so from here, eek that FSB just 2-4MHz and you will have him

Yes... but, that is the TRICK now, isn't it? That LAST 3mhz, ugh...

(LOL! Talk about straws "breaking the camel's back" & all that!)



* Hey, I'll try again tomorrow, early in the A.M. when it's like 70-75F or so again (most likely, you never know in this "Summer Heatwave 2006" we're all seeing worldwide).

APK

P.S.=> I could run my A/C right into the side of my rig, lol, & get it REALLY cool, but that'd be lame imo & unrealistic conditions (fake), well... until WINTER that is! Then, I'll have the ambient temps to do it I would wager, w/ ease! apk

 

[Ketxxx]

its 3MHz, be brave, just do it! if not from BIOS do it from windows with a64info. it allows fsb adjustments.

 

[Alec§taar]

its 3MHz, be brave, just do it! if not from BIOS do it from windows with a64info. it allows fsb adjustments.

Is 255 FSB safe to try via BIOS you think?



APK

P.S.=> If so, I will try it in the A.M., tomorrow, so I get the COLDEST possible ambient temps... they DO help, & my tests during this heatwave seem to verify this (I couldn't do this, no way, in the 100F temps I got my 1390-1401 in)...

Above all - as per usual? Thanks for the "hand-holding" here... lol! apk

 

[Ketxxx]

255 in bios should be perfectly safe, im just thinking more if your RAM doesnt like the few MHz more you will have to reset your bios vs just hitting the reset button on the case.

np on the "hand holding" either, allclockers need a guiding hand sometimes