Jump to content
HWBOT Community Forums


Popular Content

Showing most liked content since 05/25/18 in all areas

  1. 14 points
    Best thread of 2018. Priceless
  2. 10 points
    The 10 bucks debate has been done a few times over the years. Issue is that some just want to pay to keep the site alive, others think they bought a 50% share and start asking for removal of benchmarks, start to interfere with moderation/competition setups,... I need to dig up the thread where you can donate freely to Fred's paypal addy to reduce costs... At the moment the BOT is a ship with major problems and the 3-4 man crew is trying to keep first the boat afloat. Secondly steer it away from the incoming iceberg and last but not least in the mean time still find energy to save some out of the water... Everybody knows what the issues are, biggest problem is how to resolve all these in a short period as possible. Fixing is one thing, though the algorithm and co have become so complex as Strunkenbold explained, it looks more like fix one thing break two... Frederik is trying to get most stuff sorted after his daily job is done and keep this in mind we are all volunteers, doing this in our free time and totally for free... Thx to Michael for helping me out once in a while...
  3. 7 points
    Fascinating... I can count on one hand the number of TEAMS here that bench for & fight for their TEAM, most other members here at Bot bench solely for them-SELF points/rankings and to hell with their team. Prove me wrong... And that is just fine for them. So Out of pure curiosity; What is the sudden interest in TEAM points/rankings. Hmmmmm. It has been mostly catering to crying individual benchers here anyway again nothing wrong with that, if thats their thing so be it. However why go after TEAMS points/rankings? Some individuals here can not compete on a one to one basis on the higher levels so they focus on and contribute to their TEAM the only way they can. And you want to again... Kill the concept of TEAM as it has been attempted before in the past. If the TEAM PP have been the same way all these years and no one had a complaint about it before why change it? I do understand if the BenchMark was completely removed from Bott.
  4. 6 points
    It's been this way forever. Leave it alone. Not happy with your team's standings?, act as a team. You bench everything. GL/HW points or not. You want to pick and choose only high points benches, you leave some team points on the table. Seems fair to me. You don't punish active benching. You promote it.
  5. 5 points
    The whole story is that people where in charge who thought they could establish overclocking as another esport mass phenomenon. This sadly failed. Technically and financially. Energy wasn't spend on the website. And so one problem after another appeared over the time and didn't got fixed. Everyone working close to the website was seeing this. But unfortunately those are in no position to decide anything. So the penny dropped long ago...
  6. 5 points
    You change it when you are not happy with your team standings and there's no other way out. This is a witchhunt. You tell me which team is going to take the biggest hit if those points are removed. I already know the answer. So do you.
  7. 5 points
    Leave TPP alone all this no points these days for a new person they might get what 0.2 points for a lower score but they also get a couple of Team Power Points which gives them some enthusiasm to continue at least they are getting something. When I first started I thought I will give this a go and was using Socket A which I had 2 solid state drives in Raid 0 it was just the way I had it setup and a Gainward 7800GS and was getting points and Gold Cups for PCM04 and 05 and I was hooked for SPI, wprime and such I sucked and was getting 0.1 points or thereabouts. If it hadnt of been for the points I was getting for 04/05 and the TPP I probably would of just walked away. You are supposed to be making it attractive for new members to join and have fun not just the people that are experienced and have the latest and greatest binned hardware. Giving no points weather it be user points or TPP doesnt make it look very good for new members. Not many worry about Team points anyway so why change it.
  8. 5 points
    You wanna see the real team standings, remove TPP from all inactive users after say, 2 years of inactivity.
  9. 5 points
    let me try my best ok? Random stuff I think he will start charging hwbot or us with a fee of $5k usd for his scores. Other random stuff about 8pack/der8auer that personally I am not aware about but sounds juicy (send info plox). Claiming WRs that dont get points? (omegalul) Whines about satisfaction of mod(s) banning his scores. Did I get it right? If yes, 10% of that 5k pls for my translation services. Thanks
  10. 5 points
    Is Ney back moderating? And maybe we're pissing on your hard work because you're a douchebag. Maybe.
  11. 4 points
    who wants to try run coffelake on the M8i you can try the following flash bios: --> Bios 3801 (Mod Coffeelake Microcode) Nb : Flash bios using CH314a flasher do not try flashing through bios_flashback overalls work well (without any problems) : - IGP working - PCIEX x16 - Overclocking : Ratio Core, Cache and memory (all running normally) For 4core usage (eg: 8350K) do not need mod pin cpu. For 6core usage (eg 8600K and 8700K) need to do cpu mod pin (like cpu mod in MOCF) Here's the submission validation CPU-Z : use 8350K --> http://valid.x86.fr/hmpnfm use 8600K --> https://valid.x86.fr/aa73ih for other submissions can be viewed on my profile : HexaOC Powered By : Hexatekno.com
  12. 4 points
    Grabbing some more popcorn, loving the confusion, the arrogance and evil thriving within :p Must be lonely at the top...
  13. 4 points
    What happens if we react with HaHa, do you quit OC?
  14. 4 points
    I personal thanks the boss to answer so fast. They ask me to prove one more how I achieve these GPU scores and will be unblocked. They promise never to told anyone except the moderators about my tweak. But isn't der8auer one of them? What about Vince (hahaha) becoming director artistic at HWBOT? What should I do? I don't know, I don't know, need your help, I should trust them or not?!!! You can vote here by reacting to this; 1. React... SAD for YES, release the tweak as they will never used. 2. React THANKS for NO... don't trust der8auer or anyone ever. Thanks for your support.
  15. 3 points
    Yeesh people need to chill. Also as a side note, if benches no longer have normal points for a reason (buggy/cheated etc), makes sense for them to not have team points either, but if the points have been removed simply to 'balance' the main rankings (not too 2d/3d heavy etc) then I don't see a problem. I still blame Massman for pointing the ship in a certain direction then leaving it
  16. 3 points
    I wish I received 5k for beating him but I can tell you I would do it for just 5
  17. 3 points
    I found and fixed a major issue in the code that decides whether the submission is the best global submission for that user or not. Seems that fixed your submission, ksateaaa23.
  18. 2 points
    , IF, this change comes in it wont matter. we'll still get you and beat you, as small as we are ,😎
  19. 2 points
    It's already a very thin line between a perfect tweaked score and a cheated score. Maybe that's just my opinion but allowing points for bugged benchmark's will add complexity to a level we can't handle anymore. The whole bot and benchmarking in general is based off trust. And if the whole system can't be trusted anymore, it's over. By now it's already problematic but in the end the users decide. I just want to warn going this way.
  20. 2 points
  21. 2 points
    http://hwbot.org/submission/3866577_m1rror_pifast_celeron_g440_43sec_670ms/ - 2rd place thanks to all!
  22. 2 points
    Another thing I've found about pifast (Stelaras has confirmed it too) is that if you use a really really chopped up windows xp, you actually make it worse. A 150MB Win XP installation is perfect, no less though.
  23. 2 points
    @richba5tard Could please add that...
  24. 2 points
    Well no kidding, that's what he said; Anyway... I'm sure you can understand why seeing "HWBot Staff & Forums Administrator" next to der8auer's name would make slinky think otherwise. Hey, leave DJ out of this. Honestly, I've never had any reason to doubt leeg's honour. Besides, AFAIK Der8auer and KP don't have anything to do with hwbot moderation anyway. What part was unclear when I privately asked you "please don't troll that slinky thread, you're not helping"?
  25. 2 points
    I asked you not to share because "your tweaks" are not real tweaks, they bug the output Yosarianlives. A tweak is a setting you change, a new or old driver/benchmark version, run on affinity on different cores to get better scores, running another benchmark first and then run the other one, those are tweaks.... e.g. unplugging stuff so you get a 10% boost is not a tweak, that's a bug in the benchmark that you can trick... But opinions can differ, however in this case there are several heavily debated Cat subs from Slinky in the past and now the present... So one of the remaining mod(s) requires a clarification. Again no need to replicate this score lol, intrigued to see where the boost exactly comes from... ow and Der8auer and co left the ivory building years ago...
  26. 2 points
    Mr. Scott is right that this CPU is not a "970T" - it is "Thuban" based ("Zosma") without the "T" and that should be how it is named. We confuse this because the rest of that 4 core "Zosma" lineup also had a "T". The "T" however does not mean "T" for "Thuban", this is a misunderstanding. The "T" is for Turbo. "Thuban" had the first implementation of AMD Turbo Core which is why that was appended to the end of the model names. The "Zosma" 970 Black Edition does not feature Turbo Core, hence the absence of the "T". Regardless of the fact that it matches the clockspeed of the "Deneb" 970 Black Edition, L3 cache latency / IMC performance is different (~1866 MHz DDR3 limit of "Deneb"/"Propus"/"Regor" removed, and a minor IPC loss due to L3 cache configuration) and still differentiates it enough that it deserves its own page. Well, "Thuban" does have something to do with the CPU existing, but not in the way that the "T" stands for "Thuban" Just some fun facts.... "Zosma" is always the "Thuban" die. "Heka" is always "Deneb". "Propus" is actually its own die, with no L3, or the HyperTransport PHYs that "Deneb" / "Shanghai" include, though some "Propus" CPUs are "Deneb" inside. "Regor" is also its own die, though some "Regor" CPUs are "Propus" / "Deneb" inside. "Sargas" is always "Regor" inside. Some "Propus" chips are "Deneb" with disabled L3, however, this is easily differentiated physically by the 5-letter stepping code on the IHS : AACYC / CACAC / CACDC / CACYC AC ... **C** = Deneb, AADIC / AADHC / CADIC / CADHC / NADHC / NADIC ... **D** = Propus. Regardless of what die the supposed "Propus" CPU is in reality, CPUID and stepping (BL-C2 / BL-C3) will identify as "Propus", though performance is the same and they always share the same errata as their "Deneb" counterparts (RB-C2 / RB-C3). "Regor" is the same way, some "Regor" (DA-C2 / DA-C3) are actually 2-core dies physically while others are full "Propus" or full "Deneb" physically. Again, the actual "Regor" die shares the same errata and performance as "Propus" and "Deneb". CACAC / CACEC ... etc. = Deneb, NADIC ... etc. = Propus, NAEGC / NAEKC = Regor TL;DR Basically, 3rd letter of 5 letter stepping code = C, the die is Deneb. D = Propus, E = Regor. This goes for the whole 45nm Athlon II / Phenom II lineup regardless of SKU. Well, the "T" is still wrong. This chip doesn't have Turbo Core. You guys should fix it, and it can keep its own page. The CPU-Z identification is incorrect, as shown in the actual specification line: "AMD Phenom II X4 970 Processor" https://valid.x86.fr/show_oc.php?id=1913942 I studied AMD a bit much in 2008-2012. The last era where AMD didn't give a **** about wafer mask costs because their yields at 45nm were actually half decent. Cheers, Sam
  27. 2 points
    Make them signed NDA for the tweak LOL
  28. 2 points
    owkay this escalated quickly... as mentioned "the tweak" must be shared with the mods ( me or Strunkenbold) so we can replicate. Now if the tweak is what Splave suggests, that is bugging the benchmark and the output is not real. OC can be a simple thing, just clock up and bench,... why so much tweaking or bugging or whatever thing some guys pull off to justify a valid score....
  29. 2 points
    I thought most people usually just ragequit when asked to prove their tweaks?
  30. 2 points
    Its pretty obvious to me what Slinky is doing to get such a high score. you just have to run a big copywaza, then run GPUPI 32B and as soon as it finishes launch Catzilla.
  31. 2 points
  32. 2 points
    not even looking, not a the moment anyway, however if no explenation where the boost comes from, no more points for Cat scores... Is that plain understandable english?
  33. 1 point
    Yup, no more silly name changes. If you really hate this new policy, you can complain to the Australians who were messing with this functionality . Otherwise ... :celebration:
  34. 1 point
    When you remove a benchmark , like lets say UCBench or PcMark ... or whatever , you do it for a reason. Because the benchmark is vulnerable not bullet proof or for any other reason. Example Team A , benches legally with default settings and gets as top spot , rank 100 and team B benches with altered batch size and gets ranking 1 ... is that consided fair ? You wanna keep TPP for such cases ? When a benchmark is out of hwbot point system ... its out , period. Keep the position rankings for history purposes , but deleted all kind of points. I agree with you.
  35. 1 point
    7x revisions Mr Scott, same as throwing out 7x anchors to slow the Warp9 dreadnought advance
  36. 1 point
    Damn You're right now the 19,7 is my least score in main 15 Global.... damn...;) Thanks GeorgeStorn
  37. 1 point
    Is there anything you suck at? Slaying PSC and 980Ti in the same weekend?? Awesome!
  38. 1 point
    This is an autocascade , that mean use an Single compressor to cobdensate and evaporate a refrigerant mix ..hydrocarbors. It is very hard to condensate proper an r1150 gas in autocascade to give you an -100 evaporating temp boost.... that mean in slhx you need to hold 2 times heat that you target on evaporating temp on the evaporator head. For -90 at 300w you need 600w at -35 -40 on the hx...witch i'm dubt that compressor can. I think 200w at -85 is more realistisch. For 300w at -90 you need Dedicated r404a r507 Stage to condensate proper r1150 and able to hold load 300w+ we speak already about 2 stage cascade. Sorry for input but..need to be more realistic wgen we speak so easy about those watt Heat 😁
  39. 1 point
    Okay point taken. Corrected the core to Zosma and renamed the category to Phenom II X4 970 BE (Zosma). Also renamed the Deneb category to Phenom II X4 970 BE (Deneb), to not confuse the people. As we have now two ugly categories more in the db I hope everyone is happy now.
  40. 1 point
  41. 1 point
    The easiest way to run pifast in real time priority, is to run the hexusFastR shortcut from the rar I've attached No need for PriFinity, Lasso, or any of those tools. Other than that, PiFast pretty much likes what SuperPi does. Priority, affinity, lighten the OS a bit, RAM timings, sometimes also maxmem and a simple copywaza (depending on the platform). Put a folder on desktop that has 700MB files inside, or 1GB, whatever you like and just copy it once to C, then run it multiple times until you get "the best" score. @Leeghoofd is right about the win7 thing too, I've noticed that sometimes, pifast is faster on windows 7 on some platforms. For all the older processors I bench though, I use Windows XP of course. hexusFastR.rar
  42. 1 point
    for 5k i know some people would beat his scores and get a flight ticket and beat him in person lol, you are literally delusional, nobody lives from hwbot scores.Instead of investing in crazy hardware setups get yourself a ticket to Computex and see the reality
  43. 1 point
  44. 1 point
    Any ideas how a ranking becomes screwed? - edit: When thinkering on UAT with submit/edit/delete functionality I noticed the rankings don't fully update or deadlock when you modify your score after submission. I've improved this and patched it to production.
  45. 1 point
    @Jumper118 That's definitely a thing on the C6H, slots 2 and 4 are much more stable due to the way the mem traces have been optimised, wouldn't be surprised if it's also carried over to the C7H. Inversely my giga x370 board is pretty indifferent about which slots are populated.
  46. 1 point
    you are not allowed to change the standard settings of the benchmark, only things to adjust (besides GPu clocks :p) Allowed optimisations LOD Change tessellation level in graphics driver
  47. 1 point
    I had this problem with some card , I worm the core with blower 2 minutes , after this it work good for some card(7950/gtx 260/gtx670)
  48. 1 point
    If you're in it for the competition and fun, ignore the points. They don't mean anything. Don't play for the rating, play because you like to.
  49. 1 point
    Yes. It just means he ran the game tests one at a time. Totally legal.
  50. 1 point