??? $229 oh so tempting here in my place a 920 is still under $290-300, OMG.:shocked:Quote:
Originally Posted by Rock&Roll
Printable View
??? $229 oh so tempting here in my place a 920 is still under $290-300, OMG.:shocked:Quote:
Originally Posted by Rock&Roll
I just picked up a 930 at microcenter tonight on that deal they are having. Now I just have to build up this machine so I can put it to work!
Thats the batch I have. Every cpu is going to be perform differently, I can't believe batch numbers mean so much to you guys.Quote:
Originally Posted by Rock&Roll
batch number are proved worthy in here. even it's also depend on the other hardware too.Quote:
Originally Posted by rezax58
So how about making your 3rd post one reporting how well the 3001A124 batch does? Inquiring minds want to know :welcome:Quote:
Originally Posted by rezax58
Man do u actually own a 920 DO and run it @ 4.4ghz with lower vcore or u just browsing threads...? :rolleyes:Quote:
Originally Posted by aoch88
Only few chips can run @ 4.4ghz with lower vcore fully stable. Mine needs 1.44v and a friend of mine needs 1.46v. All these with good wc'ing...
Hello All,
Just got my i7-930 (Batch # 3951A824) and started working on overclocking. I thought I'd share what I've seen so far..
Current MB: ASUS P6T SE (BIOS Version 0805)
Awaiting MB: ASUS P6X58D Premium - Coming Monday :)
Cooler: Corsair H50
Memory: 12GB (6x2GB) Kingston HyperX KHX1600C9D3K6/12GX
This system is going to be an VMWare ESX Server using Intel VT-d (which ASUS supports). I'm not looking for "crazy" overclocking. I'm shooting for insanely stable, with reasonable overclock.
First off, let me say that I was one of those folks that used Prime95 (64-bit) for overall stress testing (24 hours). I've left that behind for IntelBurnTest by AgentGOD (Thank your AgentGOD!)
Here is the battery of tests I run now:
- IntelBurnTest v2.4 (64-bit) / Custom / Maximum Memory I can set it to / 8 Threads in the case with the i7-930 / 5 Runs
- PCMark Vantage -- All Tests
- 3DMark Vantage -- All Tests (Performance)
- Sisoft Sandra -- (Proc Arith / Proc Multi-Media / Memory Bandwidth)
- SuperPI -- 1M and 2M Tests
- CINEBENCH (64-bit v11.5) -- All Tests
If it fails the above "battery"...back to the drawing board :shrug:
I first ran the "battery" with everything Stock. No issues.
In the BIOS I started setting all types of "manual" settings reading different i7-920 and i7-930 review sites--and I started having all types of trouble.
Long story short, I decided to let ASUS "Auto Handle" everything and see what that did for me.
So far, so good! All power states "Enabled" (including C-STATE), Turbo "Enabled" and everything is set to "Auto". The only thing that I've set is the BCLK to 165 and the PCIE Bus to 100.
During IntelBurnTest (when everything gets pushed the most) , here are some interesting data points:
Clock: 3.63GHz (Turbo Multiplier =22, Bclk = 165)
Vcore: 1.28750 (via ASUS TurboV)
QPI: 1.23750 (via ASUS TurboV)
DRAM: 1.56v (via ASUS TurboV)
Memory Speed: 1320MHz (8-8-8-20-59-2T) (via CPU-Z)
All other voltages (IOH, ICH, PLL, etc.): DEFAULT!!
Maximum Core Temp: 74degC (Room Temp: 74degF)
Idle Core Temps: ~31degC (@ 12*165 = 1.98GHz / Room Temp ~74degF))
Sorry for the long post...but I wanted to give a data point on how things are working here in new i7-930 land :-)
Once I move over to the P6X58D I will give an update. I am going to try pushing things a bit more with this "All Auto" technique and see where it breaks.
Lambusser
Hi guys, I have an i7 930 (3001A124) that I got from Newegg last week.
Seems to be a good batch although I'm not really sure what qualifies as good.
I'm running 4.2 GHz with HT on, on air IntelBurnTest xtreme/maximum stable with 1.296V vcore (under load, ~1.32 unloaded). QPI/VTT voltage is 1.255 V, using 200x21 with RAM at 1600 MHz 8-8-8-24 2T at 1.6V. Uncore at 3200 MHz. Temps are a bit warm, about 79C load with Prime95 after a few hours on Noctua NH-D14. My board is the Gigabyte GA-X58A-UD3R (which has terrible coil whine if turbo mode is enabled, btw).
I was able to get what I think was stable at 4.5 GHz with about 1.46V vcore, however, as soon as I fired up Prime95 the temperatures shot into the 90+ C range so I stopped.
Was able to get into windows at 4.6 GHz with 1.55V and didn't try anything higher than that because the temps were way too high. Prime95 caused BSOD after 5 seconds at this speed.
For anyone that uses the X58A-UD3R, please beware of using Auto voltages for QPI/VTT (uncore). When I set my bclk to get 4.5 GHz, the board would set my QPI/VTT to 1.55V!! Took me a while to notice it, the first sign was that my temps were super high.
i7-930 @ 4635,75mhz on air http://valid.canardpc.com/show_oc.php?id=1077169
http://img101.imageshack.us/img101/6946/snap30m.th.gif http://img412.imageshack.us/img412/6947/snap31.th.gif
Hi All, may be somebody can help me. Multiplayer x22 come up to x23 all the time, its bios bug? http://valid.canardpc.com/show_oc.php?id=1075190
just got a i7 930 batch 3001a124 as well from microcenter in boston.
Will let you know my results soon.
The point is to compare against 930 so are you telling me 930 can do 4.4Ghz with that kind of Vcore or you're just guessing? If 930 uses 1.5V or more (unless proven wrong) then what's the point giving up our 920 D0's?Quote:
Originally Posted by JoeBar
i7 930 3001A541 EX58-UD4 F9
Air
http://happy.ap.teacup.com/yomoyama3/img/1267890010.jpg
As for 930, 920 and the character are different. 930 might be a sourpuss.
Do you like the Intel 930 cpu?I think i may get me one in a few weeks.
ok ok ok... im a bit confused. why cant i get my 930 multiplier above 21? it seems locked. ***noob question*** sorry
Pretty sure the multiplier on the 930 is upwards locked with a maximum of 21.
It's turbo mode. It allows one core to run at a 22x multiplier when enabled. The remaining cores remain at 21x.Quote:
Originally Posted by xann27
I'm with you on this one, batch no's mean little to me. I've seen too much variance in processor's from the same batch to go out of my way to get that "Golden" batch. Some people pay $300-$400 for 920's from a certain batch, that's just stupid.Quote:
Originally Posted by rezax58
I agree. I don't think all chips from the particular batch would perform just as good especially paying a lot more for it is the last thing I would do.Quote:
Originally Posted by rouge
Picked up a 3001A126. Testing to begin soon.
well a 920 new batch is also not so good. it's seem like they love a high voltage too. and sometimes the 20x multi that it's use was not good enough compare to the 21x multi. mine is the example if i run 20x200 it's not stable even on the 1.425v but on the 21x200 i can even run a less error even with just a 1.38125v. so i hope that the 930 will perform much better than this one.
I wish I did not throw the box away, but this is a new batch 920, just don't remember which, and poorly air cooled, but did not need that much V core. but the one I did the week before needed 1.42 to do 4.3Ghz.. lol so they seem to vary a lot
http://i18.photobucket.com/albums/b1.../Capture-8.png
I'm still more interested to see some good 930's in action :D
I would be highly surprised if the current 920 and 930 perform any differently over a large sample size, there is not much reason for them to. It's not like they have any chips to bin as 920's because they fail as 930's.
3001A126. 4Ghz on stock volts...Barely. Crashes at 193bclk. Time to start upping the volts.
http://i48.photobucket.com/albums/f2...TOCK_VOLTS.png
Testing BCLK, I ended up around 219 as a max. So i doubt I'll be limited there. (Roughly on par with what my 920 was able to do)
Currently just ran two successful passes of 10 in LinX at 4440MHz. 1.5Vcore, and +50mV VTT are all I've done so far with voltages. PWM Frequencies at 490 for VTT and 1210 for CPU gave me stability where I was previously lacking at 370 and 940 respectfully.
I'm pretty much near the limit of how much the CPU will take on the core clocks. It's getting pretty hungry for voltage. I figure 1.5 is about as far as I should go for 24/7 usage on H20. Load temps are right around 70C
I'll try to push the chip a little farther before bringing my RAM up to speed and start 24hr testing.
http://i48.photobucket.com/albums/f2...t_490_1200.png