Jump to content
HWBOT Community Forums

Leaderboard

Popular Content

Showing content with the highest reputation since 09/28/20 in Posts

  1. If I personally had the skill to work on this it would be different already but I simply don't. We hired a coder who has been trying the last weeks and months to work himself into what we have here. It's not easy but we're getting there. I can absolutely understand everyones frustration but I can promise you that I'm trying everything to make sure HWBOT 2021 will be much better than the previous years.
    23 points
  2. 32M is a blast on rocket lake! Thought we could compile some tips to get everyone started. Feel free to add yours and I will quote you in the OP Gear 1 Mode necessary duh TRAS no longer limited to 28 go as low as you can. 16-17 should be ok to 4k mhz You need blast the SA voltage 1.55-1.65 all day VCCIO_M on the other hand doesn't need Z490 levels 1.35/1.45 has not limited in any way. wazza doesnt help much, maybe 1 second. unlike other gens you can pretty much just torque each timing as low as it will go and see a benefit RTL you can try Relaxing channel A and running them at the same timings IE 53/53 54/54 has netted some solid scores BCLK is king, run as high a bclk as you can. It will allow you to lower the multiplier of the CPU to a level that you can run at exactly 5005 for the entire run Binning for Gear 1 is a thing. The same setup I have cpus that wont boot 3900 and ones that run 4060 on air. a0 might be better than a2 dimms. I wouldnt call this a fact yet but it seems to be leaning this way. IMC needs cold to max out (obvious) take advantage of training lower and raising bclk in OS (usually 2-3bclk is fine) leaving cache freq very low on training may help (needs more science) Certain microcode version may be more efficient (needs more science) per core OC needed for highest clock. Try to change only core 02 as high as it will go. usually gains 50-100mhz depending on the cpu. Also makes it fail less after the bench completes and or when opening cpuz. Right click CPUz to select the high core. Updated as of 04/29/2021
    21 points
  3. Just freshly arrived on my doorstep, brand new retail Z590 OC Formula. Thanks daddy @NickShih for the board. Some shots of the motherboard. Bios 1.24 Easy 5066C15 at 1.95Vdimm with single rank dimms 4533C16 @ 1.65 with TridentZ Neo 3600c16 (very average B-Die sticks) The latest 1.27 Bios has optimized memory training profiles and tuned LN2 profile. HWBOT member Chispy made a bootable XP for Z590 OC Formula, you can grab it here: https://www.mediafire.com/file/8h55ndgfck10m4j/20CE0A7B83E5F08B-00-00.mrimg/file ( can be slow in download ) Use Macrium Reflect software for the above image. Use ASMedia ports (IDE) - Select XP compatible in Bios Set other SATA ports to IDE or Disable them In case you want to build your own "special XP", here are the ACHI files required to inject. Other software: Formula Drive 3.0.4 (click) OC Formula Motherboard Utility (click) Multi-purpose Bios 1.27A (click) Beta Bios 1.27D (improved Gear1) Latest Beta Bios 1.27H Enjoy! Z59OCF127A.rar Z59OCF127D.zip 1920580306_ASRockMotherboardUtility(v3.0.412).rar 22471576_FormulaDrive_3.0_407.4(portablelite).rar
    19 points
  4. Hello, after a few days of testing/trying here's the final ISO + TIBx for Z490 ISO: https://mega.nz/file/MJZB2CoA#uPgJY8HNq4Y2mCIuNLrk3ApwOsvo437pqvVpOLAn3hE TIBx for Acronis 2020: https://mega.nz/file/gRJ23T4T#iFOmOUa6AansaVL-TwFcFurP484KAw53DB0c6NWhln0 This ISO based on WINDOWS XP PROFESSIONAL SP3 X86 - INTEGRAL EDITION 2020.9.9: http://www.zone94.com/downloads/software/operating-systems/123-windows-xp-professional-sp3-x86-integral-edition includes lots of patches for modern hardware which improves perfomance of modern CPU's and lots of other. Additionaly integrated: - acpi.sys compatible with Z490 - Microsoft USB 3.0 xHCI driver (Backported Win8 driver) - universal USB driver. All ports works without any issues - Samsung NVMe driver (Backported Win7 driver). Driver works during XP install. So you can install XP directly to Samsung Nvme SSD. - PAE 128GB RAM Patch (May have stability issues) - allows XP x86 to use up to 128gb of ram instead of traditional 4gb. May have stability issues with Nvidia drivers versions 332 and higher. Didn't test it though. - Kernel-Mode Driver Framework 1.11 (Backported Win7 driver). Needs for Microsoft Universal USB drivers to work properly - UAS/UASP driver (Improves USB storage device performance) - Intel xHCI driver by Stavros taken from win-raid.com 1. I wasn't able to install this ISO from USB flashdrive no matter what. I've tried lots of different software and methods (WinSetupFromUSB , Easy2Boot, Firadisk + Firahelper, Yumi ) , no luck. Always got errors after F6 and F2 messages during installation like ntkrnlmp.exe, some errors in txtsetup.sif, corrupted files etc. Allthough the same ISO installs perfectly from CD/DVD drive without any issues. This might be somehow related to bios (i'm using 0088) or amount of ram (2x16gb in my case). So i suggest to use CD/DVD drive to istall ISO or direct install to drive from TIBx mount with Acronis for those who has no CD/DVD drive. 2. In case of installing from CD/DVD Drive, plug both HDD/SSD drive and CD/DVD Drive to Asmedia ASM1061 ports on motherboard. In case of APEX XII it's SATA_E1 and SATA_E2 ports, otherwise you'll get 7B BSOD during installation. It's related to INTEL AHCI drivers somehow not detected by Windows XP during installation (although they are integrated into ISO), but working fine after complete system install. You may switc SSD/HDD drive to Intel Sata ports after system install, it'll work fine. 3. As there're NVME Drivers for Samsung integrated into ISO, you may choose to install XP directly to nvme drive even if it's GPT. No problem here. 4. First boot into XP (after 3 reboots during install process) your device manager should look like this: You may install those missing 3 PCI Device and SM Bus with Snappy Driver Installer (SDI), it's Intel LPSS, SPI and SMBus drivers. 5. Final system device manage in my case: Good luck, hopefully this post will help and save your time.
    18 points
  5. Reserved for more Stuff/Guides. --- There quite a few significant changes, however and below is a list that compares traditional expectations vs actual behavior of RKL: Traditional Expectations Actual RKL Behavior Memory and controller can only run in synch Memory and controller can run 1:1 or 1:2 Both odd and even DRAM ratios work Only DRAM odd ratios work for higher frequencies 100:100MHz base DRAM Ratios work as well as 100:133MHz base DRAM Ratios 100:100MHz base DRAM Ratios are significantly worse than 100:133MHz base DRAM Ratios The cores share the same PLL and thus at any point of time, they share the same frequency The cores have their own PLL and thus at any point of time, they can run different frequencies AVX offset is referenced against the all core active ratio AVX offset is referenced against each individual core's ratio limit The voltage needed to run stress tests is more than sufficient for cache to switch ratios High Cache Frequencies require a much higher voltage to switch to that ratio than to run stress tests Even with ring-down bin enabled, Cache will pretty much run up to it's fused max ratio Due to the above, the VID for cache is so much higher than it's core counterpart that often the Cache VID would exceed core's and induce more down-binning since Cache would regulate its frequency to not exceed core's VID requirement. The Memory controller for Rocketlake is significantly different with new features and new limitations. Firstly, there are 2 “Gear” Modes the Memory controller can run in. 1:1 or Gear 1 means that the Controller and Memory clock is in synch. 1:2 or Gear 2 means that the Controller runs at 1/2 of Memory clock. Typically, the controller can maintain a 1:1 mode till about 3600MHz. After which it can no longer work reliably and needs to switch to Gear 2. Since the controller runs at a higher frequency in Gear 1 even if memory frequency is lower, the SA voltage required is thus higher for the same memory frequency. Our auto-rules try to maintain the 1:1 gear mode frequency until 3600. Related to this, since the controller runs at only 1/2 clock in Gear 2, the command rate by default in Gear 2 is 1T, and all sub timings such as Round Trip Latency and turn-around times are tighter in Gear 2. (btw, IO latency timing option is non-existent on RKL) This tries to make up for the latency associated with 1/2 clock. On top of this, there are 2 limitations in Gear 2 mode. This first one is that there are no odd ratios available in this mode. So for example there is no 41.33 ratio. The second limitation is that ratios based off BCLK:DRAM ratio of 100:100 are more unstable compared to ratios based off BCLK:DRAM ratio of 100:133. We have explained this in the help string and have set the non-recommended ratios in grey font. There are 2 VCCIO voltages for RKL. VCCIO Mem OC Voltage is the rail that would need to be raised during overclocking Memory. Typically in the range of 1.25~1.45v. CPU VCCIO voltage at default of 1.05v powers everything else. Another note is that the PCH is powered by a VCCIN of 1.8v which is used as it’s FIVR input voltage. - LN2 The dual 8 pin power connector on M13A has priority differences, the one towards the left is the primary and has to be plugged in. We experienced no Coldbug or Coldbootbug on RKL, so there is no RSVD switches needed As before on Z490 all you got to play with is vcore, sa, io/io memoc and pll term. As starting point set the LN2 Jumper on Mainboard, cool down to -80C Basic LN2 Settings are: Ratio 55 Cache 45 Vcore: 1.65 - 1.75+ VCCIO: 1.30 VCCIO OC: 1.50 VCCIO SA: 1.40 - 1.50 PLL Term: 1.60 - 1.70 Ring PLL Voltage: 1.05 / Auto Internal PLL Voltage: 1.05 / Auto PLL Bandwidth: 2 (higher blck maybe higher) -- Worktool Usage High Cache Frequencies require a much higher voltage to switch to that ratio than to run stress tests. So for example if you’re dynamically switching the Cache ratio in the OS, whether due to power saving or manual adjustments, to switch to high cache you need to set high VCore first to succeed in the switch. Once the ratio is switched over however, the voltage can be lowered significantly, for example you may only need to use 1.7v for cinebench but need to switch to the high ring ratio with 1.90v. The tool does it for you when you apply Fixed Ring Ratio, it well set high vcore before switching, then switch then lower vcore back to previous value. The cores perform slightly differently as the same frequencies, with the middle cores yielding best performance and the first or last being the slowest. If the processor’s favored cores are at the front or back, the single thread performance maybe slower than a processor with favored cores closer to the middle. 1. Put in your desired Ratio - 55 for example 2. Put in your desired Cache Ratio - 50 for example 3. Apply Core to switch CPU Ratio 4. Apply Ring to switch CPU Cache You can try "Maxout all Limits" after you set everything, this max out the Board/CPU Limits if something in Bios is forgot to set. -- XP Tool For XP and Ratio changes you need the attached XP Tools. -- Tools - Worktool https://www.dropbox.com/s/yo5n36kj3xmb6l5/OCTool1221.zip?dl=0 - TurboV Core 1.10.12 (590 and z490) https://www.dropbox.com/s/qfzmx3uvx7ib2xd/TurboV_Core_1.10.12.7z?dl=0 - MemTweakIt https://www.dropbox.com/s/kjowe9faml7e6xh/MemTweakIt.exe?dl=0 -- Legacy Tools - WinXP Ratio Tool https://www.dropbox.com/s/805ioagux8inbqz/xptools.zip?dl=0 - 590 ACPI Files (credits go out to OGS) https://www.dropbox.com/s/ixh046p92cc8pwf/winxp_acpi.zip?dl=0 -- BIOS 0903 BETA ROG-MAXIMUS-XIII-APEX-ASUS-0903.7z 0902 BETA https://drive.google.com/file/d/1osCYx94vY4vu2en_XFES2xDGOOhEsZLY/view?usp=sharing 0901 BETA ROG-MAXIMUS-XIII-APEX-ASUS-0901.7z RKL OCTVB Preview Bios ROG-MAXIMUS-XIII-APEX-ASUS-0020.rar Guide: rkl_octvb.docx 0801 https://drive.google.com/file/d/1TixpPAyQ1oIom_FNmWOv5qI4ex91hOeX/view?usp=sharing 0705 https://www.dropbox.com/s/qz2dw2r2lxic8iv/ROG-MAXIMUS-XIII-APEX-ASUS-0705.CAP?dl=0 0706 https://drive.google.com/file/d/1C5Bt7CmfgLPt7u3HbyN5MV-nzmIQv5yV/view?usp=sharing -- BIOS FOR SUPERPI 0070 (beta) https://drive.google.com/file/d/1SRhK0Jk0XqZkOjJ2X97VP2UgnOSUzNap/view?usp=sharing This is an old BIOS, so it is recommended to use it only when running SUPERPI. Because the latest BIOS has better performance 0062 (beta) https://www.dropbox.com/s/jul8s6col0ownsp/ROG-MAXIMUS-XIII-APEX-ASUS-0062.rar?dl=0 Some refinements to gear 1 , in particular cmd rate 1t mode , plc check if helps!! I use 1.45v io mem and 1.6v SA for 3866 1t just to rule out voltages -- BIOS FLASHBACK PORT Red marked is the BIOS FLASHBACK PORT, Rename your File to M13A.CAP before flashing. -- SPI HEADER FLASH Set the Jumper(blue) left for Slot 1 or right for Slot 2 .ROM BIOS FOR FLASH This Bios is withouth the UEFI Header so you can flash it only via SPI Programmer. https://www.dropbox.com/s/s53p7opjshbfeij/ROG-MAXIMUS-XIII-APEX-ASUS-0801.rom?dl=0
    17 points
  6. Hey everyone, it's time for another freebie! I was going through my hardware recently and came across the KPC Venom container. This is such an iconic and legendary CPU pot but I just don't use it anymore. It's still one of the best all around pots and is especially good for 'fullpot' LN2 or Dry Ice. Rather than let it collect dust or sell it, I'd like to just give it away to someone who will use it and love it the way I did. The container has been lapped and it's in excellent condition. Other than the base sanding, it's in OEM condition with no mods done. I am including a KPC thermal probe and all original hardware except the back plate. Unfortunately the back plate has been used to death and butchered up for other projects, so it wont be included with the freebie. Shipping: I will pay USA shipping, but international people will need to pay the entire shipping bill, and it could be over $50 USD Who's eligible: Everyone in good standing at Hwbot! However, I really do want to see it put to use, so only enter if this is something you would use. How to enter: Just make a post in this thread. Winner selection process: The freebie will be open for about 4 weeks. I have asked @Leeghoofd to conduct the drawing on my behalf from the entries. I will 'close' the event and then turn it over to him to randomly select a winner.
    17 points
  7. The past is the past, noone wants you to forgive nor forget, I would just like all of us to move on.... Nothing me nor the new management or whatever you want to call them are affiliated with. PJ's dream was big and this resulted in some great events like Gamers Assembly in Poitiers and the ones at Computex. Some events were labelled overrated and maybe in the eyes of many not even worth to be present at all.Nevertheless it allowed people from e.g. Australia, Brazil and Indonesia to participate at the Grand OC Final in Berlin. An experience for them, that otherwise would never had been possible. PJ had great ideas, but I guess the site couldn't keep up the pace. For your info the last years the Bot was running break even, meaning Frederik could just cover the server costs. This thx to the great overclocking friendly companies like G.Skill and Galax. This lack of extra budget is probably the explenation that not much development or bug fixing happened. I don't need to tell you that in 2020 Frederik nearly payed most of the server costs himself, as most events got cancelled. Kuddos to Frederik for doing that. Without him I would not be typing at this forum right now. Also people still believe what they want to believe, but there are or were no big contracts with any CPU brand for the last 2 years. If they have real proof than just here-say, let them play their cards... Bussiness plan? What bussiness plan are you referring too.... I heard I got payed big cash too, suuuuure. I nearly payed all of my trips to Taiwan just to hang out with the crazy OC people. All my free time goes into this place, I don't even have time to bench anything myself. We need to bring the HWBOT back on the rails and improve it. This will take probably at least till the end of Q1 or even Q2. Improvements are required before we can engage ourselves to host big vendor competitions again or look for sponsorship. Most important of this long reply is that Roman and his associate Eike are giving this site a second chance. I'm more than gratefull for this opportunity as we are making big progress behind the scenes. To be honest I was more than done with the self destructive attitude from some people here, that it made me nearly quit this scene. The same bad vibes that made icons like Ney and Websmile leave the building. However I came out stronger thx to Roman and co. The latter is one of the reasons I will moderate all crap talk like there was before in this thread. Let it be a warning for the members that bring their real life frustration to our place of fun. I won't call names, but if you think this is the place to ventilate your personal issues or just want to blow things up for shits ang gigglez, you are dealing with the wrong old guy. I will remove it once, 2nd time you get an official warning and strike three you are out. There's no room for negativity Alex, life is already hard enough, just relax bro. The people you are upset with are not calling any shots anymore. Let it go... Me, Roman and Eike want this site to properly work again and allow OCers to have fun again and especially preserve the great Database. Fluck the past, we are moving forward bro!!!!
    16 points
  8. Hey there I hope everyone is doing good so far and you had a decent start into 2021. As you've probably seen there is some movement right now with our coder understanding more and more how things work on the bot. Albrecht is in good contact with him and I'm sure we're getting there. Right now our priority is still to fix stuff which is not working and work on the most important updates. There are few things I want to address quickly: Benchmark Updates: Albrecht created a power point for us to see what kind of changes the community wants to see regarding benchmarks. Personally I'm totally open and we can just follow whatever the community wants to see. You guys eventually have to play on the field so you can also give feedback on it. Finances: Financially we currently have costs of about 4000€ per month which we have to cover somehow. I'm not trying to earn any money of this project and just want to keep it alive and improve things if possible. About 900-1000 € per month is the hosting cost right now to pay domains and keep the servers running. The rest is to pay the coder and the taxes/health insures and all this stuff we have to pay for him. I'm trying to be as transparent as possible. Besides that there are no additional costs we have to cover at this point. There are multiple ways we can cover the ~4000€. At a certain point we will try to form partnerships again with e.g. G.Skill, ASUS, Intel or whoever wants to join. Right now as long as the site is not 100% fixed we will just try to get some parts back by hosting competitions but also due to corona it's very difficult to see how things evolve with Computex and so on. As I mentioned in my initial Post from earlier 2020 I'm going to set up a Patreon page for HWBot within this week and hope that some can help us keeping this project alive. Thinking about how much we all spend on hardware and LN2 per month I hope we can find some guys to donate monthly in a range of like 10-25€. I hope we can cover some good part of our cost this way and stay independent. Nothing of this will be mandatory and you will not get benefits for donating. Well you can have fun on here which is a cool thing I guess Annual Ranking - Overclocker League 2021/2022/2023...: Albrecht told me there is a good interesting in keeping an annual ranking which is also fine to me. Eike and me will change the way this ranking works to make it more interactive, easier to understand for newcomers. More details on this once we have tested our idea. Tweaking/Cheating: This is probably the only part where we're taking a decision some of you might hate but this is also a major part for constant trouble in our community. As long as I've been part of OC this always lead to drama, confusion and anger. The general issue is that we have to decide which part is considered a cheat and which part is considered a tweak. Objectively speaking both things are pretty much the same. In an ideal world all overclockers would tweak their hardware and press start on benchmarks instead of finding every loophole to somehow receive a higher score even though the hardware is still performing on the same level. We all know this is never going to happen because overclockers are overclockers. If there is a grey zone or a loophole it will always be used (I did the same in the past) Right now we have such complex rules and have to use tools such as benchmate because we all made our own lifes harder and harder every single year. Instead of adding a new rule every few weeks to limit the damage we will take a different approach: Only tweaks which are known to moderators will be allowed. If there is a new tweak which has not been discussed with moderators and it is used it will automatically be considered cheating followed by a temporary or permanent ban. Everything which measurably changes the performance will be considered a tweak. It doesn't matter if it's 0.5% or 20%. What you have to keep in mind is that we want to make this hobby more appealing to new people and if you need to know 300 tweaks for each benchmark in order to be competitive it won't improve things! I will keep you updated on the Patreon page and ofc whenever the new rule will be activated. Thanks! Roman
    16 points
  9. The bot is not stopping, again it has got nothing to do with non profitable biss transactions. Roman's primary goal was to secure the database. Mission accomplished by him and Eike. Next goal is to improve the lackluster aged coding, which will be the toughest thing to achieve, mainly thanks to the overaged Java version and not receiving proper access to the coding From there on improvements and simplification to the points algorythm, interface,... but that will be 2021 projects without a doubt... I honestly wished we already had a 2020 front end and points for everyone, bye bye complex esports related algorythms that hardly anyone can understand, but 8 months down the road we are about to shift into first gear....
    14 points
  10. Thought I'd let it go wild. A benchBIOS made by Hardware Hackers (me in particular). Been known by some for about 10 years. The thing is, I didn't add any features, it was all there done by Asus, I've just unlocked them and made some bench-related changes. List of changes: Based on latest 1024 beta Set by CMOS default: 1) No logo screen 2) Floppy disabled 3) IDE timeout = 0 4) Overclocking mode = manual 5) MPS 1.4 6) Timings set manually 7) Spread spectrum disabled 😎 Onboard devices = disabled 9) Speech POST reporter = disabled 10) USB timeout = 10sec 11) USB Highspeed = enable Moved following menus: 1) Hyperthreading moved to jumperfree (it's the only option needed in CPU configuration so moving in jumperfree saves some time) 2) Moved PCI Latency to chipset settings (it's the only option in PnP that is useful for changing for PCMark for example, so moved to a more popular section) 3) Moved memory timings to jumperfree except burst length (from previous investigations this menu doesn't do anything and I don't like useless things). 4) New memory divider menu has replaced the old one 5) DRAM Command Rate placed in Jumperfree under main timings 6) Created an "Advaned memory settings" submenu in Jumperfree and moved subtimings there (Burst length, Idle Timer, Refresh Rate and ECC capability too) 7) Placed Microcode updation instead of "Max CPUID level 3" (who needed that one anyway?) in CPU configuration. 😎 Placed AGP 4x Override under "Graphic Adapter priority" in chipset settings. 9) Removed AI tuner menu So Jumperfree currently has frequency, voltage and timings on one page and that makes it close to AI tweaker. Keyboard input for FSB frequency - it was such a waste of time to set the required value by pressing + or -. Unhidden menus: 1) All strap & memory dividers available despite the FSB frequency and labled clearly (by Asus, not me) 2) Command Rate 3) Dual Channel Organization 4) Write to Read Command Delay 5) Write Recovery Time 6) Write-Read Command Spacing 7) Read-Write Command Spacing 😎 Read Delay 9) Dynamic Read Delay Adjustment 10) Microcode Updation 11) AGP 4x Override P4C800-E mod.zip
    13 points
  11. Apparently CPU-z wasn't bin to win enough as is. So next are we gonna ban maxmem for memory frequency valids and require all mem channels instead of just 1?
    13 points
  12. Hello my name is Amy, and honestly I never find the right words to express my feelings or what to write here, but let's have a go at it: I adore this community very much, I especially like the strong arms of @Alex@ro and @Splave and that sexy music preference of @Luumi . For the rest @Leeghoofdremains a douche for turning me down and where's my ex Websmile hiding at?
    12 points
  13. Hey everyone, I was cleaning out some of my GPU boxes and I found this cool little tool that I don't need anymore. I added a few key voltage displays and mounted the whole thing on a piece of anodized aluminum. They are exceptionally rare!! I think only a handful were made for the Computex event and it never went into production. I received it from Gigabyte free of charge so I'd like to pass it on for free to anyone who wants it. I will pay USA shipping but international guys will need to pay the shipping bill...should be about $33 for most locations. It's open to everyone! If you want a chance to win this, just make any post in this thread. After a few weeks I will use a random name picker.
    12 points
  14. Hola everyone , as you already know this brand new architecture of AMD RDNA 2 has come in the form of RX 6xxx video cards. I can really say this video cards are fast and can trade blows with the rtx 3090 / rtx 3080 and rtx 3070 as i have tested myself. " Monster in all Firestrike benchmarks - FS/FSE/FSU " and nvidia cannot touch it there. In the wake of this new architecture some things have change in the amd drivers and in the rx 6xxx Bios too. As it stands right now amd has an artificial lock on the max clocks on the core and memory as you already might heard or know. This is the downside of this new video cards but hopefully not for long as there will be software to change this , right now is in Beta testing 👍 Thanks to some in depth investigation , work and mods found by @Splave and @elmor you can now tweak this video cards using the Elmor's evc2 controller to get you higher power limit in order to maintain a more constant core clocks as the clocks on this video cards are always dynamic and not static , hence with the mod you can sustain for longer core clocks *example 2750Mhz fluctuates and bounce between 2600Mhz~2750Mhz during benchmarks because of power limit , but if moded it will sustain those core clocks closer to 2700/2750Mhz thru the benchmark , hence giving you a better score overall. Most of the time higher clocks do not mean a better score ( You need to find the higher average sustain clocks during the run of the benchmark , look at FM/UL systeminfo monitoring tool report once you have sub this to your online account at FM/UL , or just use GPU-Z monitoring and logging tool , that will give an idea of the real clocks you are able to sustain , stable , highest average , not fluctuating much clocks is what you should be looking for and not the highest peak clocks ) , there has to be a balance between clocks , power limit and voltage , so keep this in mind *example 2850Mhz core clock can score lower than 2750Mhz same with the memory , 2150Mhz memory clocks can score lower than 2130Mhz *. Find the balance , sweet spot and tweak your clocks accordingly to the score on your benchmarks ( adjust as needed , play with it , that's where the fun is 😜 ) . *** Tip - When in the overclocking section of the driver and you punch your max core clocks you will also need to adjust the minimum core clocks around 100~200Mhz lower than the max core clocks ( example 2700Mhz max / 2500Mhz min ) , adjust voltages as well for the minimum clocks.*** Here you can find a great overclocking guide and tips when overclocking this video cards , big thumbs up to @Splave and @elmor for it. https://www.elmorlabs.com/index.php/forum/topic/6900xt-power-limit-tricks/#postid-338 *** Good news , there is light at the end of the Tunnel *** 👍 , as i just found out the AMD - Red Bios Editor guys are actively working on a fully, working Bios editor software to change everything possible in the Bios eg. ( clocks , power limits , voltages , etc... ) , at Igor's Lab , here - read post # 434 - https://www.igorslab.de/community/threads/red-bios-editor-und-morepowertool-bios-einträge-anpassen-optimieren-und-noch-stabiler-übertakten-navi-unlimited.2415/page-22#post-102478 As it stands right now they are still looking for volunteers to test this software. ( Let's see once this software is fully working who will be the first one to break a core clock of 4000Mhz , throwing that challenge out there as to what might be possible in the near future 😛 and who will be the first one ... About morepowertool on Igor's lab - download here the latest version - https://www.igorslab.de/red-bios-editor-bios-eintraege-anpassen-optimieren-und-noch-stabiler-uebertakten-navi-unlimited/ The software seems to work sometimes but many times you will find you get stuck at 500/500Mhz clocks due to amd protections , if you don't refresh/re-install the driver it will start acting very weird and sometimes bsod wish leads to a broken OS 😬 , it happened to me and i went thru 3 windows installations on one day , ( If that happens to you it is needed to re-install drivers and start all over again as it did happen to me too many times while testing mpt on my 6900xt. ) If you get unstable settings on mpt and cannot boot to the OS ( black screen ) you will need to enter in safe boot into the OS and revert the changes there in mpt and uninstall the driver , then and only then you will recover from bad settings. ( I tested this for many days myself ). Morepowertool is a great software to use but you got to take it easy and work your way up to find for yourself stable settings. I was able to make a profile of an mpt file for my 6900xt working at 1.35v with 550 power limit , it did work only once, after a crash during a benchmark run the system reboot itself and i was greed it with the infamous black screen and unbootable system. Never again i have managed to replicate those settings as AMD driver seems like it takes control of it 99.99% of the time. If you get a reboot while running high clocks in the middle of the benchmark , run lower core clocks and lower v.core , revert back to stable settings as this has happen to @Splave and myself ( The driver freaks out sometimes and goes into some kind of protection mode and make the system reboot ) I want to give credit to @Dancop and @OGS for confirming the low cold bug and cold boot bug of this RX 6xxx cards , every card is different so it will behave differently on sub-zero temps ( For those that remember it's the same CB-CBB issue like the old Radeon's HD 6970-HD 6950 TeraScale VLIW4 Architecture all over again ). My own 6900xt has a very low cold boot bug around -46~50c. Other cards have more or less this same problem with cold boot bug give or take 10~20c. Cold bug as reported by Dancop and OGS is around -100c give or take. *Weird mpt behavior and settings* - This were the settings that worked for me only once with mpt and i have not been able to replicate this ( not sure if it was working 100% either ) - I used the AMD driver overclocking software with this settings for overclocking the graphics card , v.core 1.20v / 550 power limit 2850Mhz core clock. As it stands right now the most common overclocking software used for AMD cards do not work yet , ie ( msi ab , OverdriveNTool , etc... ) . Some of the motherboard vendors have working overclocking software designed specifically for their own brands that somehow works but not fully. It is highly recommended to use the built in overclocking software in the driver as it is the best to use at this moment. Bare in mind this is a complete new architecture and there is only 3 drivers available for this cards , hence i expect for things to get much better once more drivers are out from AMD and we have Bios flashing software available as well as a more stable morepowertool. The Red Bios Rebellion Team at Igor's lab is working hard to get this software working as soon as possible. If you have something to add , share experiences , tips , tweaks and tricks i encourage you please post it here on this thread. It is a challenge to find all this alone hence sharing is the only way to help each other. Enjoy it !
    11 points
  15. It's me again, beta 11 adds ability to set AGP/PCI frequency as well. It has most of the functionality I want. It even has a better granularity for the FSB frequency than clockgen. What's left is to add ability to save and load OC profiles, complete the CPU detection/info and add some options like "minimize to tray". NForce2Xtreme-v1.1-beta11.zip
    10 points
  16. The Corsair compo could have been so much more, it has a strong competitive field of our community members, but nearly no newcomers. You should have seen the reactions on the Discord channels when you posted the placeholder. The newcomers, that is the entire reasoning behind the this whole thread correct? The ommitance of the CPUZ tabs when using BM, to make benchmarking easier and more accessible? My entire point of my previous rant post is is that the level of restrictions laid down by the rules and the required verification screenshot are not the sole cause of the participation level in compos or HWBOT in general. Think about how freakingly expensive Hardware is and even harder to get hold of these days. That's another reason to cherish the legacy benchers, they are a vast value and need to be taken care off. To be competitve you need to indulge yourself in different OSses, tweaks,... like explained before. OC has a hard learning curve and some just don't have the dedication or motivation to get it done... even though OC tools make it soooo much easier versus 10 years ago. Take XTU, so easy to do, but from the 4K subs we get a month, how many continue to come back and start to bench other benchmarks? Many newcomers also think that a one time big investment in the latest and greatest will make them shine, so they will get a sponsorship deal and that they can keep up their ranking by posting a score once in a while. I guess we all know better. It's an attitude that BM nor we can't change... If I was rude, my apologies, I'm not a keyboard guy, I speak from the heart.
    9 points
  17. If 12900k has 16 cores, treat it as a 16-core cpu. Period. If Intel decides to make 8 of them stronger and 8 of them weaker, have HT on one and not on the other, thats then Intel's decision. From what you hear it even performs like a "true" 16-core 5950x. So what's the big deal. Also 12900k is monolithic. For example the 5950x isn't even monolithic and has chiplets. Can I disable chiplets and make it join the 8- or 4-core category? No I can't. So if you have a 12900k you can join two categories and essentially gain double the points in some way, that's just weird and complicated. Allowing disabling part of the chip so it can joins a different category so it looks stronger is contradicting previously made statement regarding disabling cores for frequency validation. ppl can disable all they want it is still an 16-core chip, and therefore it should get treated that way eventhough some cores are stronger than others, some are more efficient than others. It's the overall performance that counts. Regardless of the amount of threads right? When Intel comes out with with quad HT, what do we do then? It's still X amount of cores but 4-times the amount of threads. The way I look at it is the individual cores just got more powerful/efficient. But plz don't make it more complicated with dividing CPUs into multiple categories. Please correct me if I'm wrong but we do we differ between a 6-core with HT and one without?
    9 points
  18. Corsair Stage 1: AIDA64 Read Bandwidth Corsair Stage 1 package in download (file at the bottom of the post) or https://www.mediafire.com/file/sdk7xw7y3njdbtu/Stage1.rar/file Official Corsair Compo background I still need to update some links & change stage titles, but the admin page is down But you can already start with prepping stuff... CorsairStage1.rar
    9 points
  19. After some works, finally get 700 FSB validation https://hwbot.org/submission/4703644 Platform can do 680 FSB Pi 1M all by air cooling. PL value reading is wrong, it's 11 actually, not 1
    9 points
  20. In house bencher, someone either paid by a company or works for a company or given unlimited access etc. In some regards I feel this maybe the only way for us to really grow, I also can't help but to feel guilty sometimes. They can still claim world records but it gives the people a chance to accel in rankings etc.
    9 points
  21. This weekend we got some heavy maintenance planned on the database of HWBot. This could lead to you being unable to submit ( 500 error ) Sorry for any inconvenience caused, but it has to happen some time What are we doing: Removal of a lot of bot created accounts Releasing new benchmarks from Beta stage Alphabetically sorting the benchmarks Adjusting the global point system More functionality for the admins ... So just bench and hang onto them scores, no need to sub them now, The Bot will be back soon with improvements under the hood...
    9 points
  22. Thank you all for your support! Please close the sale -Return of the Jedi-
    9 points
  23. Hi team, we have this incredible platform with hwbot that we mostly love and sometimes hate. Train of thought: Recently I uploaded a result with Benchmate commented by buildzoid who pointed out (and rightfully so as it is compliant with the rules) that one of my three cpu-z windows didn’t show the memory tab as I made this annoying mistake and opened up the same cpu-z tab twice. That I‘m sure happened to all of us once in the heat of the moment trying to conserve that precious score in a screenshot. https://hwbot.org/submission/4810041_ w Ah annoying! Well I re-ran it before it even got picked up by an admin the next day and put up the proper screenshot. No big deal right? However it got me thinking to the point that I wanted to write a thread to check if any of you feel a similar way. So in my case described above all the relevant info about memory were present in the screenshot of our one stop shop: Benchmate. Despite the fact that we have this great tool in 2021 that has been constantly evolving over the past years which streamlines the benching experience, makes it all simple and safe we still have to open half a dozen cpu-z windows that in most cases don’t offer any relevant additional info when used in conjunction with Benchmate. From a historical standpoint we’ve been using cpuz forever to monitor hardware info. However this is not history class, it’s overclocking so I don’t count an argument that is just like „we do it because we have always done it“. Let’s evaluate: Cpu-z has barely evolved at all. Same layout/gui as ever. Here is the deal: I’d be cool with it if there were like a combined window/tab with all important info in one, call it „benchmark view“ or whatever. But if it brings nothing new to the table we might just get rid of the necessity to include it in many screenshots/validations where Benchmate is used that in contrary manages to include all info in one tidy little window, simple, easy and compliant with judging requirements. Listen I love hwbot and XOC and it’s not a secret that it has issues attracting new users and interaction over a long period of time. It is kinda hard to see things from a new user‘s perspective after doing this for a long time but I remember at the beginning of my hwbot endeavors all the different benchmarks and their rules - it was all a bit overwhelming and intimidating eventhough I rly wanted to do it I guess at least I‘m still here right. So working in a global hardware/software company I get the impression that hwbot needs to get a whole lot more accessible to increase the adoption rate. As we can see from the recent Corsair comp: great prices are not enough to make ppl play by our rules. Holding back highscores can’t be the solution either. Ppl these days want it easy without annoying obstacles, that’s why zoom was king over Webex and MS teams at the start of covid, that’s why ppl use windows over Linux, that’s why ppl use Apple over PC… you’re mileage may vary but I hope you get the point. So for hwbot to stay relevant we need to attract more user’s, that might not be as hardcore. These days I’m certain the guys in charge of hwbot rely on their great connections within the industry to get us competitions with prices. We can make their lives a whole lot easier and our overclocking a whole lot more rewarding (comps with prices) a whole lot more often if hwbot managed to get more user to play the game of overclocking. Hence I encourage you to evaluate how we can make the experience here a great one. Knocking off a new user because he forgot or didn’t know what extra windows to open which I bet happens all the time is not motivating. Or imagine you have been sitting there for hours and finally got the score you've been chasing there on the screen but CPU-Z bugs out and won't open for you to take the correct screenshot, man that sucks! The goal should be to streamline the benching process and I feel like we have the right foundation today. If it means to get rid of irrelevant bulk of tools/rules and even benchmarks that are unpopular, to make hwbot slimmer and more accessible heck then let’s stop being stubborn think outside the box and take necessary steps.
    8 points
  24. If you call it a 8 core its is fighting against 5800x and 11900k IMO that doesnt seem fair at all. If you were to pickup the phone and say "hello MR.Intel, how many cores does a 12900k have?" they would say it has 16. In my opinion its a 16 core processor, this is due to the fact that it has 16 cores. We need to not dilute the rankings by catering to hopeful companies to claim PR victories.
    8 points
  25. 8 points
  26. Thanks to everyone who joined the freebie! This offering has run it's course and tonight I did a quick random name pick to determine the winner. The winner is: gkmltd I don't believe we've ever run into each other but expect a PM from me shortly. Stay tuned for a new freebie I have in the works...something quite a bit better this time :D
    8 points
  27. Didn't realise this was an issue, it's been allowed for as long as I can remember, it's allowed for other benches so why wouldn't it be for this? Gives people a better chance of needing less binning, instead of 1 core holding back 7 good ones for example. If people really do care then having two categories is reasonable, but don't see any logic getting rid of the current one.
    8 points
  28. HWBOT goes Patreon. Those that regularly visit the HWBOT community forum or for those that are on the LN2 Haven Discord channel are already aware that Der8auer has announced that HWBOT now has its own Patreon Page. HWBOT is going through some major changes and 2021 will be a transition year to a non buggy, more responsive and easier to understand website, entirely dedicated to our passion; Overclocking. This ofcourse goes along with extra development costs on top of the monthly Amazon server hosting fee. Therefore you can voluntarily become a HWBOT Patron, just to give us that extra push to move forward. You can find more info on how to become a Patron at the HWBOT Patreon page To reward those that support HWBOT by being a Patron, it has already been decided to host an exclusive Patron Member Competition with prizes. Competition specifics on the date and such have not been decided yet, but we are aiming around October/November. You only need to be a Basic Rate Patron Supporter for at least 6 months to be eligible to participate in this exclusive competition. To avoid any confusion, donating does not provide you extra HWBOT Boints, additional decision rights, shares nor other extras at the HWBOT site. Remember that HWBOT will remain a free to use website. So it is entirely up to you, to decide to step into this Patreon Project and to help us cover the extra development costs we are currently facing. Now if you can't spare the money or you just simply don't want to donate, you all will still be welcome to submit your overclocking scores and to engage on the Community Forum. Those that donate, thank you for your support, it is highly appreciated!
    8 points
  29. I remember back in the day must have been like 2008 or 2009 era trying to find the most obscure hw to vmod and put on chilled water to get my gold cup and 5 points even for no submission or 2-3 user categories. I really enjoyed that time. It was so pure. I hope to see us maybe get back to a mix of that and the new. Also some fresh new benches getting points would be great too. I'm confident in leeg and the new leadership will get this place back to greatness. Thanks for the discussion men.
    8 points
  30. I'm fine with points for in house benchers. Main reason - such results will be listed on main page, giving these results necessary attention of the community. Probably we should have some competitions hosted by these guys here - would be great to get a prize from HiCookie / Shamino etc. All in all, we usually get detailed guides from these guys and that is their priceless contribution.
    8 points
  31. Tbh, over the course of this thread and a bit of thinking I've changed my mind, I honestly think we should just keep going as we already are - so 16C. *Shrugs*
    7 points
  32. Hey guys, let's talk and post results of the hynix djr ic here. You already know that these broke memory frequency WR on Z490, well things only got better on Z590. With the better IMC from RocketLake now these sticks can be used for competitive benching as well. I had a shot with them on Asus Maximus XIII Apex and things are interesting. Let me put it this way. On air a decent Hynix DJR Kit is most of the times better than a very good B-die kit. On cold if your Cpu can handle both high core clock and memory clock at same time, a decent Hynix DJR kit is similar to a great B-die kit. To tell the whole story, Hynix DJR is way easier to handle, no need for maxmem, if it enters windows it usually runs without hickups. I was really excited about the Hynix DJR performance and really hoped that they will send the Samsung B-die at the thrash bin after so many years of painfull benching, sadly they cannot top but they are a very very viable alternative. You can find 3 profiles in Apex Bios which work and they give a good starting point if you want to try Hynix DJR on your brand new Apex and Rocket Lake Results on Intel 11900KF and Asus Maximus XII Apex: Hynix DJR cooled down to -150: DDR4-5866 17-27-27, memory score at 5 ghz on older microcode bios ==> 12058 Which is better than the memory score of Samsung B-die DDR4-5333 15-15-15 (12034) in same conditions. Hynix DJR on air can do between DDR4-5600-5700 with C19 and the performance is equivalent to a Samsung B-die DDR4-5000-5066 C14. Cold findings with Hynix DJR: - they scale good till -50, at -150 there is a small gain but not much. - trcd and trp remain same air vs cold, no gain - trfc which really helps these ( 20 values tighter in TRFC gives 100 points boost in geek3 memscore) sadly is the same on cold - secondary and tertiary go very low, on ln2 tighter than air but no performance gains. There are no twr or trtp or whatever issues Samsung B-die has. Here is my best result with HyperX DDR4-4600 C19 based on Hynix DJR on ln2: Similar performance to Samsung B-die at DDR4-5333 15-15-15 Feel free to post your results and experiences as well, maybe we can find some better binns or better sorting methods.
    7 points
  33. Treat it as it is, 16C is still 16C doesnt matter slower core, faster core, or different clocked between core, core count is still same. If 12900K/KF lose to 5950X/10960X, leave it as it is, because actual multithreading performance is slower, they still have single core performance advantages anyway. No need more complicated rules, now is already complicated enough.
    7 points
  34. From HWBot perspective I'm fine with either way, but I think we are at a point where things are already so complex that we're shooting ourselves in the foot on a daily basis. We have so many different plattforms and benchmarks that it's simply not possible to have one unique way of validation. I'm also not a friend of changing things retrospectively. For example we could start "BM only" from Alder Lake and also ban WinXP with Alder Lake but not change older generations. Yes, yes some will hate me when I talk about ban for the good old WinXP. However I'm with CENS, that things are very complex for new people and if they see that a 20 year old OS is required which might be older than themselves - could talk about how we handle this in future. But again - I'm very open for changes as long as the majority of the community shares the view (same goes for the WinXP example).
    7 points
  35. I actually wasn't trolling and my point couldn't be more relevant to this discussion. AMD is doing this already but the cores just match. Makes it even more relevant when you clock the cores separately like we will be with alder lake. None of the top 5950x scores using matching clocks, I know because I have them. Close the thread if the decision is already made. It seems to be giving the false sense of value to other opinions than the already chosen decision. The people that run the show should decide anyways as there is always a spilt in what should be done. ❤️
    7 points
  36. 352 GB of RAM in one thread. Is that a new record?
    7 points
  37. New version of the tweaker. Started from scratch and ported all the functionality from the previous app. This is a native one and doesn't need .NET framework, which results in much faster start up. There are several improvements and new settings, FSB slider doesn't do anything yet. Source: https://github.com/irusanov/nForce2-xtreme-tweaker-vcl NForce2Xtreme-v1.1-beta3.zip
    7 points
  38. Some experience from me playing around with both 2x8GB and 2x16GB the last few weeks: It's hard to find IMC that does good 2x8GB on air (2000+). On LN2 should not be a problem for most cpus. But it's much harder to find IMC that does good with 2x16GB when cold! For example my ES, which has the best IMC so far, can loop PYP at 2x16GB C12-11 very tight at 2050MHz all on air. When full pot IMC tops out at just 2000. The same CPU can do 2x8GB 2200++ C12 on LN2 full pot. You have to pay attention to timings. Lower is definitely not always faster. And then balance performance vs stability. For example, with my timings TRFC 180 is worse than 220 for both PYP and 32M. Same goes for some subtimings. Some timings give slightly faster times but ends in crash 9/10 times. For me that is not worth it when another setting passes 10/10 times. For me A2 is better than A0/A1 for 2x8GB. But I haven't played so much with A0/A1 because my A2 sets are fantastic. I know A0/A1 likes slightly different timings, so maybe I didn't tweak that enough... For 2x16GB I only have A2. Best 2x16GB kit so far did around 1970MHz C12-11 32M, 2050MHz PYP. Best 2x8GB does 4300+ C12-12 32M and 4400+ PYP. 32M with benchmate is really fast on win7. I use the same OS as I do for 3D, so nothing special really. Don't have to do waza either. A good win 7 run can compete with XP for low clock challanges, but it varies more than XP. For full out XP will give you higher clocks, so XP is still better there. On ASUS make sure Round Trip Latency is enabled on mem training settings. And use latest SPI bios and maximus tweak mode 2. For 32M with 2x16GB on XP higher maxmem than 600 is better. When benching cold I would recommend 2x8GB, because unless you have a killer IMC it will have trouble handling 2x16 when cold. At least when you bench full pot. You will most likely loose MHz compared to air, unlike 2x8GB where you can see massive gains.
    7 points
  39. Maybe some of you noticed that the latest versions of UL's benchmark software is missing the Sky Diver benchmark. We have 2 options to get it installed. Online installer or Offline installer fix by Antinomy Getting it to work is pretty easy as Antinomy has gathered the required files and zipped them into one package. (click link) Al you need to do is to unzip the package in the DLC Install path: C:\ProgramData\UL\3DMark\chops\dlc\ (this is the default folder during the install, see above picture) When you check the Options tab, you will notice that Sigh Diver is now installed To Bench it you will need to click on Filters (Top right on the benchmark Tab), the below window will open, move the slider of Include unsupported tests This will unlock Sky Diver, Cloud Gate and Ice Storm series Big thanks to our Dbase Guru Antinomy for providing an offline fix!!!
    7 points
  40. 50/48 4000mhz 12-11-11-18-180 SR 5min04.922 12-11-11-28-280 DR 5min04.422
    7 points
  41. Sorry to keep you waiting so long, need some time to process my VCRs. We all know, EP45 UD3R/UD3P are good boards to Core 2 Quard CPU for 500 FSB stability, but few OCer would like to use EP45/EP45T UD3 series for 700 FSB goal, they felt not so good, so as I did. Since Mr. GRIFF made 708 FSB by EP45 UD3P, I begin to pay more attention to these UD3 boards. After severals weeks, I found that UD3 boards were really good. Now, The most important point I want to say is, When CPU and NB are both with LN2 cooling, at least 50% of E0 stepping C2D E8X00 CPU can do 700 FSB CPU-Z validation, by EP45/EP45T UD3 series. If you can't make this, the first thing you suspect should be NB, or BIOS tweak, not CPU. The important BIOS options on UD3 boards for raw high FSB (in fact, they're almost the same as on the other P45 boards): Vcore: 1.40V to 1.60V (no need LLC) Vtt: 1.40V to 1.60V Vpll: 1.35V (air) to 1.59V (LN2) Vnb: 1.46V to 1.70V (get a IHS on NB chip! make sure NB cooling is good!) Vmem: your mem choice CPU Vref: 0.61x to 0.62x NB Vref: 0.578x to 0.591x CPU Clock Drive: 800mV to 1000mV CPU Clock Skew: 50ps to 100ps Performance Enhance: "Standard" Mem strap: 1:1 CPU features: all to "disable" Other BIOS options remain "auto" Mem for DDR2 UD board: Micron D9GMH Mem for DDR3 UD board: Elpida Hyper VGA: PCI card for 720+ FSB
    7 points
  42. Just finished a meeting with the programmer, new benchmark global point system will be setup this weekend, first on the test uat.hwbot.org, afterwards on the main site.
    7 points
  43. Please remove MaxxMem entirely, it's super buggy and inconsistent. Other than that, everything seems "fine".
    7 points
  44. This one maybe bit more interesting?
    7 points
  45. And that's what Allen 2-5 spots up ? you guys have to work and compete at the highest level. Each time you have to prove yourselves. This brings a lot of pressure with it and requires total devotion on eg these launch events. People just think give me that 5.9 CPU and I'll do the same... some might just do that, however many will fail... I feel reverting back to the old point algorythm is a better solution for most... but that's for 2021
    7 points
  46. The only reason i didn’t freeze a kit so far is that i coudn’t remove damn heatspreaders. I consider that i paid 300$ for that service and not the item itself :)).
    7 points
  47. Hopefully everything will settle down and fix , let's think positive. Alby my good old friend please do not leave as you are essential for this hobby and you are a great person. Give a shout to Frederick and Roman to get all this stuff fix and i think is time for them to join the community conversation to see how we can help. Kind Regards: Angelo
    7 points
  48. Truly hope things get fixed soon so you don't have to leave it behind, you've done an excellent job keeping it afloat with very little to work with from the sound of it.
    7 points
×
×
  • Create New...