Jump to content
HWBOT Community Forums

Dead Things

Members
  • Content Count

    225
  • Joined

  • Last visited

Everything posted by Dead Things

  1. Very nice! You should run Time Spy with this rig for the Country Cup!
  2. Feedback for deaf ears: rules re: mainboard tab and sensors tab are just good ol' fashioned retarded. Finally managed to fit it all in my window. Good luck to mods trying to make sure my competition wallpaper is compliant.
  3. Beauty - that worked! Thanks a bunch cbjaust!
  4. Looks like I'm missing something. Admittedly, I haven't been keeping up with HWBOT rules very well lately. Anybody know what I'm doing wrong here? http://hwbot.org/submission/3990096_ Thx!
  5. Nothing to change, Strunkenbold. Reported score is quite clearly the actual value.
  6. Nothing to change, Strunkenbold. Reported score is quite clearly the actual value.
  7. Ugh. Thanks man. Any other new rules I should know about before I try again? (For the record, this is exactly why competitive overclocking will never be "a thing.")
  8. Did this with an AMD Phenom II X2 555BE with the 3rd core enabled. Tried manually editing the submission after uploading it, but that didn't work either. Here it is: http://hwbot.org/submission/3732439_ edit - In fact, doesn't look like I'm able to edit any of my submissions that use a datafile upload to reflect the 3rd core being active. Same problem is happening with HWBOT Prime now. http://hwbot.org/submission/3732444_ And... ditto for x265: http://hwbot.org/submission/3732447_
  9. Thanks for putting this together TASOS! I'm looking forward to it! Just one question - would we be allowed to use slockets to run Socket 370 chips on Slot 1 non-Intel boards?
  10. For stage 2, what is the rule for unlocking cores? Phenom X2 550 BE, for example. Is it still a dual-core CPU with cores unlocked? Also, is "Please attach a picture of your overclocking rig in use" considered a polite suggestion again? Sounds similar to "Please provide a screenshot with official background, CPUZ CPU, mainboard, memory and SPD tab and GPUZ for 3D stages" - which is obviously not a polite suggestion.
  11. I thought about that too! But then, like mr.paco said, it's better to try to make the comp reasonably accessible and most people don't have workstation cards. Maybe we can have one stage for the more attainable workstation cards like the the Quadro FX 3700, for example. Maybe we could even allow cross-flashing from GeForce to Quadro for the comp. edit - FX 3700 is the 8800 GT in workstation's clothing
  12. That's a good point. I do like the notion of retaining some element of era-appropriateness, so how about opening up the GPU requirements to say 2-3 eligible generations per stage rather than forcing just the one? Would that help?
  13. Thanks for the corrections! And yes, all of your interpretations of what I meant to say have been correct. Sorry I couldn't be more clear in the first place! In my personal opinion, I would be in favour of ES chips being allowed for this. It is an OC competition, after all.
  14. Hi Mr. Paco, et al. Yes, I think you are right about that (^). Still I found I had trouble fathoming anything but a round-by-socket design when attempting to design stages. So I went ahead and did it that way. I hope you are able to find this helpful in some way: INTEL SOCKETS S775/S771 (Xeon-class processor required) - Stage 1 = 2 cores on 1 socket, SuperPi 32M - Stage 2 = 4 cores on 1 socket, 3D Mark 01 SE w/ GeForce 8000/9000 (single card) - Stage 3 = 4 cores on 2 sockets, PC Mark 05 - Stage 4 = 8 cores on 2 sockets, Cinebench R11.5 S1366 (Xeon-class processor required) - Stage 1 = UP Bloomfield/Gainestown, PiFast - Stage 2 = UP Gulftown/Westmere, 3D Mark Cloud Gate w/ GeForce 200 (up to 3x SLI) - Stage 3 = DP Gainestown, wPrime 32 - Stage 4 = DP Westmere, GPUPI 1B S1155/S2011 (Xeon-class processor required) - Stage 1 = UP Sandy/Sandy-E, Aquamark w/ GeForce 400 (up to 2x SLI) - Stage 2 = UP Ivy/Ivy-E, 3D Mark 01 SE w/ Radeon HD 7000 (up to 4x Xfire) - Stage 3 = DP Sandy-E, HWBOT x265 1080p - Stage 4 = DP Ivy-E, Geekbench3 Multi S1151/S2011-3 (Xeon-class processor required) - Stage 1 = UP SuperPi 32M - Stage 2 = UP Catzilla 1440p w/ AMD R5/R7/R9 200/300 (single card) - Stage 3 = DP 3D Mark Vantage w/ Nvidia GeForce 900 incl. Titan-M (up to 4x SLI) - Stage 4 = DP Cinebench R15 AMD SOCKETS S939/S940 (Opteron-class processor required) - Stage 1 = 1 core on 1 socket, PC Mark 04 - Stage 2 = 2 cores on 1 socket, Aquamark w/ Radeon HD 2000 (up to 2x Xfire) - Stage 3 = 2 cores on 2 socekts, Cinebench 2003 - Stage 4 = 4 cores on 2 sockets, 3D Mark 05 w/ GeForce 7000 (up to 4x SLI) S1277 (Opteron or FX-class processor required) - Stage 1 = DP Santa Rosa/Windsor, PC Mark Vantage - Stage 2 = DP Santa Rosa/Windsor, 3D Mark 06 w/ Radeon 5000 (up to 3x Xfire) - Stage 3 = 4P Barcelona/Shanghai, GPUPI 100M - Stage 4 = 4P Istanbul, wPrime 1024 G34 (Opteron-class processor required) - Stage 1 = DP Magny Cours, 3D Mark Sky Diver w/ GeForce 700 incl. Titan-K (up to 3x SLI) - Stage 2 = DP Istanbul/Abu Dhabi, Geekbench 4 Multi - Stage 3 = 4P Magny Cours, Y-Cruncher 10B - Stage 4 = 4P Istanbul/Abu Dhabi, HWBOT x265 4K I didn't bother going back to Socket 603/604 or earlier, because I figure that's OSIBS territory. I also didn't mess with 4P 2011+ because I figure that's IT pro/elite territory. I also tried to mix in various GPU combos with an eye towards 1) some limited degree of era-appropriateness and 2) maxing out some of the bigger scores that could benefit most from the MP Xeon/Opty chips. Hopefully at least some of this will be helpful to you. If not, that's cool too. I had fun designing it, so if this is as far as it goes, I'm still happy! edit - I should note that in my mind the UP/DP requirements can include populating just 1 socket on a DP board or 2 sockets on a MP board and that's totally legit.
  15. I fully agree - and this is how Jumper envisioned this thing to begin with in the OP. Would give us all a chance to focus in on a platform for a while and eek out as much performance from it as we can. The more I toy with ideas in my tiny brain, the more I keep coming back to this round-by-socket formula.
  16. That's cool - they were just some ideas to help TASOS out. Please feel free to refine them, replace them, or submit your own. Personally - I'm up for pretty much anything. I think TASOS has a good idea of what he wants to do with OSIBS this year and I'm excited to find out what he decides.
  17. I'll put some thought to it this weekend. As a collector, it's sometimes hard for me to remain cognizant of what is reasonable to expect others to have or acquire for comps like this. Maybe I'm the only one here with Dempseys and nobody else is even remotely interested in going out and acquiring those old, hot, slow chips. (But historically fascinating seeing as how Woodcrest was released mere weeks later!)
  18. You make a point, TASOS. It is a legendary piece of hardware, though - and if not in OSIBS, then when do we get a chance to bench it? (Then again, same could be said for Socket 4 I guess - if not OSIBS, when? Never?) It's your call though... edit - changed my mind about the following: Backup proposal: Remove Geforce 256 from Stage 2, move Geforce2/Radeon 7000 down to Stage 2, move Radeon 9000 down to Stage 4 and add GeForce 6000/Radeon X800 to Stage 5. Howabout we just swap in Voodoo2 SLI for GeForce 256 instead?
  19. It occurs to me - although I'd personally love a chance to bench Socket 4, it may be more difficult to acquire than the others. So I might propose this as a potential replacement round: Round 2b: Slot 1 Stage 1: PII "Celeron" PiFast (single-slot only) Stage 2: PII wPrime 32M full-out (multi-slot OK) Stage 3: 3DMark 2000 w/ PII & GeForce 256
  20. There have already been some pretty good proposals in this thread, so is there a certain framework you are working with for each round, TASOS? Or do you want round proposals from scratch? Well, in any case, here are some more round proposals for your consideration... Round 1: 486 - Stage 1: SuperPi 1M 5x86 + X5 - Stage 2: SuperPi 1M 486-only - Stage 3: 3DMark99 Max Round 2: Socket 4 - Stage 1: PiFast - Stage 2: SuperPi 32M - Stage 3: Aquamark Round 3: Socket 7 - Stage 1: Non-Intel or AMD CPU frequency - Stage 2: Intel-only PCMark 04 - Stage 3: wPrime 1024M full-out (K6-2/K6-3 & multi-socket OK) Round 4: Socket 423 - Stage 1: RIMM memory frequency - Stage 2: 3DMark 01SE w/ GeForce2 or Radeon 7000 - Stage 3: HWBOT Prime Round 5: Socket 754 - Stage 1: "Athlon" PCMark 05 - Stage 2: "Turion/Sempron" Cinebench R15 - Stage 3: 3DMark 05 w/ Radeon 9000
  21. Sorry for chiming in so late on this. Wondered if there might be merit in looking at specific cores in some of these cases? Dempsey is a far cry from Harpertown on S771, for example, so each might be deserving of its own stage - not to mention the myriad of cores in between. Barcelona/Istanbul on S1207, MC/AD on G34, Sledge/Italy on S940 - also good examples where one realistically cannot compete against the other, hence maybe deserving of a unique stage. I wouldn't necessarily wanna get locked into only ever benching one set of chips for each platform but instead having a chance to investigate a number of different generations on each. I mean not always - but some of the time. Thoughts?
  22. I fully support the notion of limiting a possible 486 stage to exclude AMD DX5, otherwise that's the chip everyone will use. Kinda boring. But whatever - I'd be just as happy to bench mine again. Love the idea of S4 and some of the other vintage cores being suggested like Klamath and Tbird. I really like where this is going. I think websmile is starting to get a bit anxious about finalizing things, so I'll forgo any further suggestions. Looking forward to it!
  23. Why not? This is akin to being willfully ignorant of how people have actually been building computers for home use. For simplicity's sake, let's only look at the case of S1366 Xeons. Before SB-E was a thing, if you wanted 6C/12T without having to sell your firstborn, you went Xeon. In the real world, limiting S1366 comps to 970/980X/990X is actually favouring industry players compared to allowing Westmere Xeons because those chips were so much more expensive than their Xeon equivalents - even back then! I'm sorry, but this is really just a very narrow-minded viewpoint - a one-rule-fits-all approach. I thought with the allowance of S939 Opterons in the Team Cup that HWBOT was starting to realize this.
  24. Yes. Yes I do. Was just hoping to use something as innocuous as an X5680 on a RIIIE. Oh well.
×
×
  • Create New...