Jump to content
HWBOT Community Forums

buildzoid

Members
  • Posts

    488
  • Joined

  • Last visited

  • Days Won

    2

Everything posted by buildzoid

  1. Oh man that sounds like exactly my pot and mem situation though my mem does 1488MHz but yeah I have to mount the Raptor 4 with 2 screws and tilted a little because my 2GB GDDR5 gigabyte card has a rectangular mount. Also about Icestorm. It's legal to set desktop resolution to like 800x600? because that makes the benchmark run at super low res and gets a huge FPS boost.
  2. If I have a B350 board with no external clockgen it's legal on win10 right? cuz I got this when doing some testing: and can't really be bothered setting up win7 right now.
  3. finding a GDDR5 GT 730 cheap is proving harder than I thought it would be.
  4. Ok so I guess I should go hunt down all the i7 owning rookies and novices in my team. Stage 2 allows Ryzen CPUs correct? I like the no 1080Ti/TXP rule. 1 because I don't nvidia. 2 because they are extremely expensive. Also Nvidia just launched an even faster TXP2 so the 1080Ti might end up being too slow and who ever buys a new 1200$ card is gonna win. Also about this rule: For each stage, overclockers from a same team will have to use either different CPUs, chipsets or GPUs to count in the Average scoring / Team. I.e. For Stage 1, the average top 3 scores for Team A is calculated as: It mentions different chipsets so would Z170 + 7700K and Z270 + 7700K count as different enough?
  5. I have 2400 16-16-16-39 Micron kit that does 3000 11-14-14-28-1T on 1.7V and 2800 10-14-13-26-1T on 1.6V. It also does 2666 10-13-13-28-1T on 1.45V with ease. However it is pretty old RAM so your kit might clock very very different.
  6. Ok so I finally got a B-die kit and a 7700K and in my first 6 hours with kit on the board I got 4000 12-13-13-28-1T working on 1.97V 32M stable. However the moment I set 4000 12-12-12-28-1T on any voltage from 1.88V to 2.05V I get a 3F POST code. Am I correct in assuming that my 7700K's IMC is the problem?
  7. 7970 should scale to 1.4V and beyond on water cooling.
  8. Yeah I bought a license. And I did do the File >> Save thing and the app just does nothing when I press save. I guess it might be my win 10 install but everything else works fine. 3Dmark has no problem saving files.
  9. so how do you get Geek Bench 3 to save a validation file? because every time I pressed save or export nothing happened.
  10. just put it under watercooling. Stock generally means the reference blower cooler.
  11. Z270N seems to lock cache to be 200MHz less than core. I have cache set to 50x but get 4800 cache with 5G core and 4600 cache with 4.8G core. The benchmark is really inconsistent. My 6663 run was pure luck if I fire up the system on the mem settings I used for the 6663 run it scores ~6.5K but sometimes it randomly jumps above 6.6K.
  12. I can 1T at 3600! Train the RAM at the highest speed that POSTs 1T then Enable Fast Boot(which disable RAM training during POST) and raise the RAM speed to what you actually want. I don't have anything stronger than some D-die that can barely do 3733 15-19-19-28 but I can get 1T working on them above 3600 with the method I described.
  13. Great so I can go back to rerunning timespy for a while but at 0.02 less BCLK or what ever it is that will make 3Dmark happy.
  14. wait so do I need to redo my 102.05BCLK run? Because I have 5G CPU-z but 3Dmark is showing 5004MHz
  15. 4.5 low volts doesn't really do much for 5Ghz capability. I had a 4.5 1.13V chip that needed 1.45V for 4.9.
  16. Have you tried 3500? I have 3530 1T running just fine on a terrible D-die kit. But 3600 1T is impossible for me.
  17. So this is a thing: One intresting thing to note is that the second CB run takes more time IRL time to finish than the first so I think it is very likely that the CPU is still at 4GHz but the Windows clock broke which leads to higher scores.
  18. CBB on Sandy bridge is normal. Most of those chips CB around -80C.
  19. gain an extra 10MHz on you 12-11-11-28 Bdie OC! Though to be honest I've been toying with the idea of trying to add some extra caps to a stick of RAM.
  20. BIOSTAR apparently managed to validate a G.skill 3600MHz 4x4GB kit on their Racing GT7 board. If BIOSTAR can get 3600 working I would hope that we can run 3200-3400 at decent timings.
  21. Are you powering the PVCC pin? Without 12/5V going to it the controller doesn't have gate drive power.
  22. What are you feeding the UP6210's VCC with? It needs at least 5V to work.
  23. The CPU 8pin connector can do a lot more than 150W. The GPU 8pin is only 3GND and 3 12V and that can do over 150W. So CPU 8pin is at least 200W and if you go by the spec. Which is actually really underated because the official spec for the connector is that each pair of 12V and GND pin is good for 13A(this is more than the 10A rating of the 18AWG cables PSUs use). I've shoved close to 400W through CPU 8 pins no problem and based on their electrical spec I would advise not pushing more than 450W through an 8 pin since at that point the cables might melt. As for the VRMs blowing up. Without knowing what boards blew up that doesn't really tell us anything. Plenty of AM3+ boards also blow up. If none of the boards currently announced can handle maxed out Zen CPUs there's always the E-power option but I really doubt that will be necessary
  24. The 8 core is a good way to get competitive in the 3Dmark rankings. Sure you'll still lose to the 6950Xs but you'll be keeping up with the 6900K and 5960X and IMO that's enough if you want don't want 1st places and more aiming for top 10/20.
×
×
  • Create New...