havli

It seems CPU-Z detects wrong SDRAM clock on VIA KT133 chipset. CPU-Z states FSB + 33 MHz. 132 + 33 = 165 on screenshot below. However the real mem clock is FSB + PCI (Aida 64 shows this correctly)... In this example calculation is: PCI = 132/3 = 44 MHz FSB = 132 MHz Mem = 132 + 44 = 176 MHz So the question is - can I submit the real clock (Aida 64) or stick to the (incorrect) CPU-Z value?

Wrong... Celeron Mendocino (PPGA 370) - PII Deschutes based, 250nm, 128KB L2, no SSE. This is the one I use, see the screenshot. Celeron Coppermine (FCPGA 370) - PIII Coppermine based, 180nm, 128KB L2, SSE. Not allowed here. Celeron Tualatin (FCPGA2 370) - PIII Tualatin based, 130nm, 256KB L2, SSE. Not allowed here. //edit: Sorry, wrong again. http://www.cpu-world.com/CPUs/Pentium-II/Intel-Mobile%20Pentium%20II%20366%20-%20KP80524KX366256.html Cache location and speed doesn't change architecture of CPU. Another example is PII Xeon - it has external full-speed L2 cache.... and still is member of PII family. http://www.cpu-world.com/CPUs/Pentium-II/Intel-Pentium%20II%20Xeon%20450%20-%2080523KX4501M%20%28BX80523KX4501M%29.html

Covington is PII Deschutes without L2 cache. Mendocino is based od the same design, but instead od external half-speed 512KB L2 cache there is 128KB on-die full-speed. Celeron has exactly the same architecture (P6) and instructions (X86, X87, MMX) as the "full PII" has. All Covington and Mendocino Celerons are on the list of CPUs suitable for this compettion. I even asked explicitly whether is ok to use Celeron and the answer from Massman was "yes". No grey area here. Cache size doesnt really matter. How about mobile PII Dixon with 256KB on-die full-speed L2 cache? Thats almost the same CPU as Celeron Mendocino... only cache is twice as big. All these CPUs belong to the Pentium II family, like it or not. Anyway, in wprime cache has very little impact on score.

No it isn't. However PII-based Celeron is on the list of allowed CPUs http://hwbot.org/hardware/processors#key=intel_pentium_2 (approved by Massman) So I don't see a problem here.

Well, Celeron Mendocino is part of the PII family. It is allowed to use it in this stage - http://forum.hwbot.org/showpost.php?p=292635&postcount=32

Yes, I know. I think the best benchmark for voodos is the 3Dmark 2000. Most of them can run it at default settings - V2 SLI and up. 3Dmark 99 is bit too much CPU limited. It would be nice to have some competition with voodoos in the future.

Actually Voodoo4/5 can run GT1... however it requires some sw hacks via 3D analyze... http://www.3dfxzone.it/enboard/topic.asp?TOPIC_ID=1129&whichpage=64 As for the hardware - I have everything except the GeForce GT610. Maybe I'll buy one.

The Pentium II stage says : Is it ok to use PII based Celeron (Mendocino)?

I can't find any log file in c:\Program Files\HWBOT\Aquamark3. The "logs" folder is empty. The only log.txt file is in documents and settings (included in the zip file I linked before).

I just tried the lastest version (2.53) and it still doesn't work. Aquamark runs fine, score is visible in the wrapper... but can't be saved. The "Save to desktop" button doesn't work. This time the testing system was: 2x Opteron 252 FireGL X1 256MB (similar to R9700 pro) MSI K8T Master2 (K8T800pro) 1GB DDR Win XP SP3 CZ - no mods, no updates, clean install Logs and everything else generated by the wrapper is here: http://hw-museum.cz/pom/Aquamark3.zip

Yes, I know this rule... altrough some benchmarks score better with more memory. Anyway this is not the issue. There is also lower frequency of GPU/RAM on mine card. This is official model made by Ati - there are two variants of Radeon DDR - 166MHz and 183MHz. It is same as Radeon 8500 LE / 8500, GF3 / Ti200 / Ti500, Voodoo3 1000/2000/3000/3500... and many more. All these cards have separate categories.

Please add this VGA It is similar to this 7200 DDR, however mine has only 32MB and 166MHz GPU/RAM. I think it deserves separate category. http://hwbot.org/hardware/videocard/radeon_7200_ddr/ Thx.

I'm afraid this bug is still here... at least for some VGAs. Radeon 9800 XXL (OEM variant of 9800 pro) + catalyst 9.3 + win XP SP3. Aquamark runs fine but the score in the wrapper remains at 0. Auto/manual mode doesn't matter, score is still 0. Older version 2.27 works normal. http://hwbot.org/submission/2456097_havli_aquamark_radeon_9800_xxl_60679_marks http://abload.de/img/aqm_error9df1g.png http://abload.de/img/aqm_error2oae9y.png

Please add this MB: http://valid.canardpc.com/nwpe1d Thx

GF6800 unlocked to 16/6

Please add this MB: Gigabyte GA-686NX http://www.gigabyte.com/products/product-page.aspx?pid=1464#sp Thanks

Hi, how did you manage to get a Pentium Pro validation? I have the same CPU here on Gigabyte 440FX board and CPU-Z 1.65.1 refuses to validate it. Normal online validation fails with error code 99. Offline validation doesn't work either - "Dump corrupted - Virtual Softwares not Allowed". Is there a way to validate this thing? Thx.

Please add this MB: http://valid.canardpc.com/show_oc.php?id=2817984 It comes from SGI 550 Workstation, however name of the board itself isn't very user-friendly http://www.hardware-reselling.de/ProductLinks/SGI550.pdf btw. OC is just wrong clock detection by cpu-z... Thanks

Actually it is more like Velocity 100 than V3 - second TMU is disabled in OGL/glide.... just like the V100. No problem, there wouldn't be much competition in that category anyway.

I have a new VGA here. It is Voodoo3 2000 8MB integrated on MSI MS-6168 board. Not sure what name should be used... maybe Voodoo3 2000 IGP 8MB. There are also 16MB 2000 IGP and 3000 IGP variants, but these are very rare and I don't have them (yet ) Thanks

Please add this MB: Fujitsu Siemens D1194 http://valid.canardpc.com/show_oc.php?id=2802323 Thanks

small error in HW masters ranking

Challenge accepted...

Yeah, I used 1.6V on 775 Prescotts and no problems there. That P4C800 was running Northwoods at 1,9V absolutely fine... maybe just a bad luck with Prescott. One of the mosfets blew up instantly after applying voltage change in BIOS.

Nice score with DDR platform. I was testing Prescott 3.0 on my P4C800... and the board died. 1.5V vcore was apparently too much. I hope replacing VRM will fix it.