(I am aware that certain Benchmarks mostly do NOT show whole cross comparison Real_Life performance between different sistems! .. well some do it , like PassMark´s Perf.Test, quite good for example..); * marked means "idiot proof" for cross comparison between different sistems ... (How to benchmark - the guide to) ...






- CPU - I keep it here @1,35GHz : (my sistem performs like Amd-AthlonXP1600+ [common setup] PC w/DDR in average, but heats much less ..)





........CPU benching......

bench

........SiSoft Sandra.......

- myTemps&Volts , idle at early summer, here CPU = on-Die !

- Ram135fsb bench´ , but much better for "real life" test

, is to use instead of that theoretical max mem bandwith test, another bench´ DL , the newest WinRAR 3.3 with its Built_in mem. sub-system test (starts using Alt+B) for truly comparable* results with others ; mine turboTweaked setup gets index of 377 points for example (who says that DDR is faster!) . ..

- CpuMM135fsb bench´ & also CPUmath135fsb, bench´ ,

... or use better bench´ DL*: index of 119 points with CPUMark99 (ZD Inc.), (or score 3398 with MadOnion´s PCMark2002 ) or these days probably best : like "Sandra" nicely reports 2 me (maybe best using all same latest Lite version of it from SiSoft while doing "real life" Cpu bench´*) its estimated Performance Rating (PR) index ... that I get it for true "Sandra" 2001pro comparison* with others...

- Other : [ & also CPU relative Performance chart ]

- Cpu150fsb bench´ ,

- MoBo150fsb temps & voltage .

- Max L2 for Cpu, yeah it is Tualatin Celeron!

- MaxCPUident-wow!, but in another setup does it even more!

 

............Hard Disk......

bench

.......SiSoft Sandra & HD Tach.......

IMHO, that Ata66-133/Sata stuff etc. it is mostly a marketing trick... What matters is just spinning speed, access time & data density on platters & like "Sandra" 2001pro nicely reports 2 me [but this bench depends too much on UDMA mode!] (using All THIS same linked version DL while doing "real life" HD bench´[using all same ATA mode]*) its Drive index or another not so reliable bench´ PCMark2002 by MadOnion/Futuremark, which gives (but depends of Cpu used to much) also a score ; I get 626 points there...

- 1.QuantF+AS@ata33 . . (truly comparable* with others if using UDMA33 mode !)

- 2.QuantF+AS@ata33

 

..........& VideoCard ......

bench

.......Aquarium Screensaver.......

This is a real life (this older version is truly comparable* with others ! IF screen setting are the same:1024/768/16-bit colors only !) bench´ DL (not for hardcore gamers!) for average users in FPS/s to see their videoCard performance depending only on its GPU & Ram mem-subsystem how it really performs regardless the latest "screen view enhancement technologies" incorporated onto latest GPUs for gamers & runs entirely of the card when loaded! Minimal requirements: 1GHz ? 100Fsb Cpu, 100/100MHz videocard Gpu & mem clock with 8Mb Ram, DirectX 8.1 recommended; I get 28 Fps/s in average there with my completely outdated videocard ...

 

..an experiment with temperatures..

bench Amb+Aux = CPU on-Die !

.......with different fan speeds......

...people just worry about temperatures too much...

- 12vFanTemps

- 7vFanTemps

- 5vFanTemps

- NoFanTemps.. OUCH ! 82°C !!!

¤ - MBProbe CPU temp = in_socket/near_slot1 here ..



The temperature inside CPU core can vary between few degrees & even 25°C or more, depending of how good the cooling is & where the sensor is placed, but usually is around 10°C higher than the one measured from the bottom of the HSF near IHS (core). If the sensor is mounted in socket, the difference can reach also 15°C or more! Read about
revealing CPU temps , before arguing about temps on Net ! See the results I ´ve got for info time ago! (these days temps are a bit higher than before since I´ve silenced my setup even more lately ...)


Test done with my Tuallie: (idle at 19° ambient, PC case closed!, only old PSUs th.c.fan continuesly running)


tempTest tablerotfan


Just see how the temps & their differences increase with lowering airFlow/fans Rpm... temps in °C ...

Yes, I pushed my working CPU max to 88°C & it survived! (MoBo thermal throttling engaged for safety..)

Recently, since I have changed my PSU, some fans, added one & enhanced further air flow thru the case (additional pics will follow), the temps are now lower than shown; I get with all fans now @ only 5V all the time temps almost the same as before having fans running at 7V shown on upper slide, so machine is now practically inaudible (if needed I still have an option to switch to 7V, but no more to 12V, because really not needed!) . AirFlow thru case is now so well done [even if estimated only around 10
CFM], that there is practically no difference in temps between PC case closed or opened (maybe half degree max) and with slightly positive air pressure inside @ idle, to prevent excessive dust accumulation inside the case thru time ! [The smaller fans have estimated only few CFMs]

So my temps are now @ idle & Amb./room/ one 19°C like this now : Cpu diode 47,5°, socket sensor 41,5°, HSbottom-IHS 40,5°, old Cpu on MoBo under Slot1 HS 40°, MoBo /case/ hottest spot 28° ; other temps : Video Card HS 46°, NB HS 53° ; MosFets HS 42°, HD top-case 34° ; out-take PSUs air temp. 30°.
 
This is with ALL my fans running @ 5V, only 2-fan Psu back out-take one @ 4V (5V max @ 50° air temp inside Psu, thermo-controlled ! + 20/25%, starts rising very slowly after 25° ambient) /& lately also Psu inner fan even when PC is off @ 3V-800Rpm/. The intake air temp inside case for Cpu fan for eventual °C/W HS measurement is 26,5°C. RPMs of fans are: front 8x8cm intake 1200, Cpu 5x5 one 2400, NB 4x4 1400, thinner Psu intake 8x8 one /normally constantly @ 5V !/ 1350, Psu 8x8 out-take only 1000 (max.th.c.1200), HD 5x5 fan /constantly @ 5V/ under 1900, newly mounted Cpu voltage regulators one 1950. RPMs may vary up to +/-10 % with air & HS´s temperature changes close to fans, since they are with grease & not oil lubrificated bearings now !... If I switch fans to 7V (really not needed practically except maybe in hot summer days! RPMs are then accordingly 1600, 3100, 1900, 1400, 1300/1600thC+ , 2600), temps drop around 2° & if I run "heavy tasks", they rise again for few °C ... except ... ...BurnMMX.exe run on high priority level with P switch (from BurnBp6 suite), which heats CPU extremely bumping its temp up from idle for 10°C !!! measured from onDie ... & remember: onDie diode temps do NOT follow linearly ambient temps change! The differences are SMALLER than with regular probes !

 

The REST : Old Celly 400 ......

bench

.......pushed a little bit more.......

..also AirCooled..with LOUD fans .. 2,4Vcore & this BIG heatsink - not much less than 1kg monster , performing almost like WaterCooling (and which I Btw, wanted to modd & mount it on planned to buy Abit ST6 MoBo, before becoming one of the earlier members of a group at MadOnions /now Futuremark/ forum at March 2002, who actually made it possible to stably run Tuallies on older BX MoBos) ..

... Celly400 almost 100% stable @ 672MHz , TL stable @ 660MHz ....... but nice & silent @ 600-default Vcore!!! (50% OC) with a bit bigger HSF than stock one !!! & @ 75°C core temp still worked normally !

- 1.C400@682MHz , with heavier load crashes ...

- 2.C400@703MHz , an than Crash follows in few minutes..!