ADVenturePO

Yes, right but the problem is also in HWBOT rankings- which moderators should be changed. Most importantly the Global Ranking should be true - one gets scores for the amount GPU ranking truly to the amount of GPUs he does have in the machine. Second thing is that moderators should either get 5 categories : 1, 2, 3, 4 and more than 4 GPUs; or 1-10 GPUs categories. But be aware of the fact that the highest amount of cards connected to one motherboard is 17 - Octane Render 4.0 result shows this.

Dear Mr.Scott. Yes, because I saw this problem two months ago. And ever since I had reported few submissions like that few times, but nothing has changed.

Would be great if something would change. Yes, I see how it works, but I'm sure it shouldn't work like that if HWBOT is meant to be treated as the main OC platform. If I'm building hardware for 8 days, testing for another 7 and someone just ads cards into server board and get points in 4 cards category than this competition is worthless.

Hello OCers and moderators. @Leeghoofd @Massman @richba5tard Something is terribly wrong with GPUPI scores. Those submissions https://hwbot.org/submission/3991289_x_computers.eu_gpupi___1b_4_geforce_rtx_2080_ti_0sec_639ms/ https://hwbot.org/submission/4143713_pijonson_gpupi_v3.3___1b_4_geforce_rtx_2080_ti_0sec_549ms/ are marked as for 4 GPUs - and their not. And another problem is with points - I'm not getting any for my GPUPI 1B submissions. Why? There's an error when recalculating. Competing with 8-10 cards against 4 is somehow not great score. It's cheating. https://hwbot.org/submission/3540678_ This submission also violates cards amount and gets Global points for 4 GPUs. Delete them or disable its points in 4 GPU competitions. This really kills competition. Peter.

And now I reported this sub because it shouldn't get points for global 4x GPU score.

Dear OCers and admins. @Leeghoofd I did participate in challengers knowing that HWINFO64 will show precise frequency and part of my scores were having .4Mhz or even .8Mhz where for example 5203 was expected. As you may know hwmonitor is showing only full MHz. Thus even when you have 5203,9 in hwinfo64 you will have hwmonitor at 5203. Most of participants used hwmonitor. Due to automatic steering by the motherboard which exceeded xxx3MHz. I thought that if I'm not exceeding xxx4 than I'm alright. Also I get notifications from HWBOT at email but when I want to get at the topic the page is 404. I don't see it in HWBOT account either and so I can't reply to it than. Peter

Hi OCers! I need a bit of tips, not for benchmarks but in fact for maximum stability in a long period of time running under stress. What I have : i9-7960x, Sage WS 10G, 128GB RAM G.Skill Trident Z CL16, and 4 RTX 2080 TI. This config is stable, but system was hanging at 1.9 CPU input voltage - I've changed to 1.95. Haven't really move VCCIO 1.1 & VCCSA beyond 1.15 - low because this MB has PLX chips. I'm curious should I increase VCCIO and VCCSA or if it's stable should I leave it alone. I know that RAM is getting into the picture and 128 is a lot for this platform - without controllers and ECC. If You had any suggestions please share. Thanks Peter

Reconsider. It was used in Challenger so it is very legitimate benchmark. Omitting this whole version is for me a bit strange, as it has all functionalities: HPET and submission management. GPUPI 4 is not even at the table and You want to skip whole benchmarking effort. Especially mine PS: Such a shame. Very inconvenient for me, very convenient for others, haha. Well, it feels like a paper crown now. I don't get to play with big toys for a long periods all the time.

OMG. 6-6-6 ! I have 3 different Hyper X sets none can do it like that...

Another legit stock market CPU record beaten with ES... damn. Nice run though.

Hello, GPUPI 3.3 was the official benchmark in Challengers ... when will we have global points for 100MB CPU, 1B CPU and 1B GPU and 32B GPU? Just checking

A hahaha. XTU Well I'll throw here some of my experience: 1. You can be achieving great scores in other benchmarks with AIO or Air Cooling. If the temps are too low - XTU points are very low. In fact : running worse profile - which is clearly loosing in SuperPi, HexusPi, WPrime, HWBOT X265 and GPUPI - but is hotter, gives more points. 2. Sometimes it works with motherboard chipset drivers... or it don't. Sometimes installing Intel ME helps, sometimes doesn't. 3. One profile clearly wins in almost any benchmark - like 3rd or 4th place in Enthusiast ... and than XTU throws score like a half of other OCers points (???). Than you know you have to change or reinstall the chipset driver - why? Don't know. Does intel XTU read settings from UEFI correctly? Hahaha. No, It doesn't. Showing sometimes even other vcore mode than is stored in UEFI. I had Override Mode but it was showing Adaptive. Do i like to compete with XTU? No. What am I doing when running benchmark - I set fans to low. Do I loose a lot points because it likes higher temps - yes. Do I cherish XTU. Oh, hell no! The most stupid benchmark ever made. https://hwbot.org/xtu/analyze/4034384 Taichi XE -no OC? Yes, A80 Cryorig was running this CPU barebones with der8auer Direct die frame Skylake-X. 67*C too low. "I'm sorry.... your temps are too low. No points for you." Logic of this benchmarks hurts my brain.

Why are You all guys so happy about. This is clearly an advertisement run. Sorry, I'm building Workstations. I know what kind of stuff is on the market. Most of cards will do 2000Mhz when slighly OCed. Those did not. Why? Because he has some too small power suppllies. Why didn't he get any 32B records because cards were throttling a lot. This page... this organization is about overclocking. Achieving the best possible by increasing clocks. There are no motherboards for overclocking - SuperMicro doesnt have any, to hold 10 cards. Yet You all seem to be very cherish about a score like that. Sorry, H2o vs Ln2 is a clear winner. It's like having a competition in cars carnage and a guy takes heavy armored army vehicle and he wins. And You're happy? No OC with the CPU, no OC with cards. Pure server parts at stock. And throttling cards. No congrats. Well lets ask NVidia to take part with this competition with their best tesla based servers. Wow! Hurray! Nvidia won. Are You nuts? For me : H2o vs Ln2 is a clear winner here.

Don't remember - but it's early 2018 so it might not be. Since May 2018 all my i9 for 2066 are relidded with LM under IHS.

Hi OCers I just can't get along with my DDR4 RAM at Taichi XE and i7 7740x I have G.Skill Trident Z : 4600 18-22-22-42 (RGB) 4400 19-19-19-39 3600 15-15-15-35 All of them are capable of running 14-14-14-28 at 1.42 vDRAM , 1.22 VCCIO and 1.24 VCCSA 4600 and 4400 will run at 4133 19-19-19-39 at 1.45 vDRAM , 1.22 VCCIO and 1.24 VCCSA But I've tried every voltage from 1.5 till 2.0V ( about 0.03 step) and voltages from 1.26 VCCIO ->> 1.4 and voltages from 1.28 VCCSA->> 1.35 Nothing works. 3866 17-17-(17)-34 at any voltage point is no go either. CPU is at 5.2GHz 1.38V 4.8GHz Cache. It's rock solid with 3600 14-14-(14)-28 in any bechmark. 12-12-(12)-28 is impossible with any voltage .. Mostly at lower voltage RIG is resetting till failure and F1 or F2. At higher voltages I will get QC 78 for 3 second and then QC 91 - the RIG is fully blocked and have to be turned of with PSU switch. 68-91 is Chipset. Am I missing something or am I just one unlucky bastard? Some PLL tricks? Maybe You know some kind of solution/instruction "How to overclock DDR4 at x299 - extremely extreme version" I know it's a long post. I'm sorry. Thank you for your time. Peter

Thanks, Splave! A didn't get over 1.28 at VCCIO. That may be the solution...

Splave! How the heck did you make those memories go 12-11-11-28? Whoa?! I have 4400Mhz and 4600Mhz all I can do is 14-14-14-28.... What voltage did you give VCCSA, VCCIO and DRAM?

Don't close the subject on x79 - I have one... in my personal PC - and it's i7-4960x but it's only 4.7GHz with Noctua NHD14 at liquid metal (for silence). To participate i'd need to take it out and that very time consuming. I have no different x79 MB nor CPU. So I think I'll be able to participate with it till the end of this competition.

OMG. I just can't go shorter with timings. Slight more at VCCIO or VCCSA -BSOD. Shorter CL - BOOT block 3F. F*ck F*uck F*ck ;)))

"Fine by me but if X58 is allowed then X79 should be also imo." I'm in for that. ;)

Whoa 1GB RAM? How did You do it? :)

@LEEGHOOFD Thank You.

Hello, Is Intel Xeon X5690 on P6T considered as a server and thus not allowed? I'm asking because those chips were sold for workstations ... And I'm a bit confused. Sorry.

Brilliant! The temps are too low. The score is shitty. Wow. Intel Extreme Temperature Utility.

The better the score, the faster but cooler machine + the worse score it gets. Nice. Worthless benchmark. Truly.