AMD Thuban X6 OCing database

[lopri]

What is your NB voltage, I am interested to know. I might try OC'ing my NB just for the fun of it.

Anywhere between 1.225~1.250V.

 

[thelvman]

Yes vcore is too high. Can you manually set your vcore instead? What mobo are you using?

I'm using a gigabyte ga-890gpa
I can manually change my cvore. In the bios, I can switch voltage control to manual. It default v core to 1.475V. I can go up or down in increments of 0.025V. I was playing with these voltages I could not run stable at 3.5Ghz if I set my voltage below 1.425V. I just set voltage to auto now.
To overclock I set my bus speed to 250Mhz and changed my multiplier on other parameters to keep them at stock speeds (2000Mhz and 133.3Mhz)

 

[lopri]

If you manage NB speeds of 3 ghz or higher, do let us know. My C2 chips just haven't wanted to go there. They make a show of it, but . . .

I tried 3.25 GHz NB with DDR3-2000/8-9-8-1T running Linpack (same 4.0 GHz core frequency) and it locked up half way. (~7th loop?) Up to that point it was churning out 81+ GFlops.

 

[DrMrLordX]

Yeah I have the MSI 890fx board. Lots of bizarre things going on with this board. AMD overdrive reporting weird temps for TMPIN when I was at 3.8GHz(like in the -280c range....) when right now at idle im getting 28,35,35,0. It never reports anything but 0 for the 4th reading. I assume those are the power regulation chip temps but really im not sure. Also the voltage setting options confuse me. As far as I can tell cpu voltage does nothing but cpu vdd is what you have to use. Also cpu-z always reports my voltage in the over 1.4v range(usually 1.41v) even though its set to 1.375v in the bios and overdrive as well as msi control center are telling me 1.375.

That's . . . weird. I haven't had any problems with voltage manipulation on my 790fx-gd70. CPU vdd should allow you to tweak the chip's vid (I think), but it isn't what one would normally use when overvolting the cores. Something seems to be seriously fubared with your board's power regulation. I would seriously consider RMAing it, if you can stand being without a board for awhile. It's not supposed to do stuff like that.

I tried 3.25 GHz NB with DDR3-2000/8-9-8-1T running Linpack (same 4.0 GHz core frequency) and it locked up half way. (~7th loop?) Up to that point it was churning out 81+ GFlops.

Niiiice. Moar voltage! I routinely run a CPU-NB of 1.425v. Of course, my current chip never breaks 37C but still . . . yeah, I'm a little nuts.

 

[richierich1212]

Niiiice. Moar voltage! I routinely run a CPU-NB of 1.425v. Of course, my current chip never breaks 37C but still . . . yeah, I'm a little nuts.

Easy to replace a Sempy 140; 1090T, not so much

 

[DrMrLordX]

Easy to replace a Sempy 140; 1090T, not so much

Yarrr too true.

 

[richierich1212]

Yarrr too true.

Ready for some more pi battles? Have you tried y-cruncher yet?

 

[DrMrLordX]

Ready for some more pi battles? Have you tried y-cruncher yet?

Negative on both counts. The chip degraded a bit during my last foray into SuperPi32m madness. Either that, or the ambients running about 4c hotter than when I set this thing up are making the chip behave a little bit less well. Maybe if I crank up the AC . . .

I need a new chip. The x2 220:

http://www.eworldsale.com/amd-athlon-ii-x2-dual-core-processor-22028-ghzam3oem_8665_53326.html

may prove to be a better budget chip for Pi 32m. Really I'm curious as to whether or not the c3 memory controller on the non-Deneb chips can hit DDR3-1800 or better, and that's about the cheapest way to find out. I have had no luck finding cheap Sempron 145s around which is what I'd rather get than the 220. The cheapest 145 I can find is here:

http://www.dectrader.com/611196-001...lebase&utm_medium=shopping&utm_campaign=parts

manufacturer: HP indeed. Pfft.

Maybe I should try y-cruncher to distract myself from wanting new hardware when the cup runneth empty.

edit: to avoid wasting space in this thread which is not about a Sempron 140, I'll simply post my best y-cruncher 32m pi calculation result here:

45.362 seconds total time (with output + verify)

That was at 3.85 ghz core, 2772 mhz NB, DDR3-1642 6-7-5-15 1T and the AC bumped down to 65F

I only got away with that 'cuz the wife and kids were sleeping snug in their beds, so they did not notice me messing with the AC. yar har har.

Anyway, back on topic . . . I'm sure a Thuban chip would do better than that! Especially one of those really keen 95w 1055Ts.

 

[jvroig]

manufacturer: HP indeed. Pfft.

And they say there's no competition beyond Intel/AMD

I tried 3.25 GHz NB with DDR3-2000/8-9-8-1T running Linpack (same 4.0 GHz core frequency) and it locked up half way. (~7th loop?) Up to that point it was churning out 81+ GFlops.

They really seemed to have improved a lot from C3 to E0. My C3 can't go beyond 2.75GHz for the CPU-NB, and I can only reach that when I run my X4 as an X2, otherwise I can only top out at 2.706GHz. And from what I read, 2.6-2.8 is generally the best you can expect from C3 chips. It's damn surprising that Thubans seem to get over this, despite having 6 cores.

 

[busydude]

They really seemed to have improved a lot from C3 to E0. My C3 can't go beyond 2.75GHz for the CPU-NB, and I can only reach that when I run my X4 as an X2, otherwise I can only top out at 2.706GHz. And from what I read, 2.6-2.8 is generally the best you can expect from C3 chips. It's damn surprising that Thubans seem to get over this, despite having 6 cores. jvroig is online now Report Post

Exactly. I also could do 2.8 GHz with my C3 chip and the system is very stable at that speeds. I'll have to wait for the ambients to drop down a bit, this week was the hottest I have seen in the midwest.

 

[richierich1212]

Well some of the Phenom II X2's have weaker NBs, which is why they were sold as X2.

 

[Jovec]

What's the consensus on the maximum safe CPU-NB voltage? My 1090t will do 2.6GHz on the CPU-NB at 1.225v but needs 1.3v for 2.8GHz.

 

[DrMrLordX]

1.3v has always been pretty safe for the other 45nm K10.5 chips, so I would assume that the same would apply for Thuban. Anything over that and the potential for danger may exist.

 

[Idontcare]

1.3V is all? I thought folks were taking their 45nm AMD chips up to 1.5V routinely when OC'ing...maybe that is only on water?

 

[MarkLuvsCS]

1.3V is all? I thought folks were taking their 45nm AMD chips up to 1.5V routinely when OC'ing...maybe that is only on water?

I believe they were referring to the CPU-NB voltage stock 1.2 or so i believe...

I definitely plan on putting my 1090T to 1.45/1.5v for some OCing when i get in in a few days.

 

[DrMrLordX]

Correct, that is in reference to CPU-NB.

I have run CPU-NB as high as 1.51v, but not for extended durations. It's hard to tell what those voltage levels do to the chip since CPU-NB seldom makes it run much hotter (in my experience).

 

[DrMrLordX]

Just an FYI, TigerDirect now has the OEM 95W 1055T for $190 shipped:

http://www.tigerdirect.com/applicati...IN#detailspecs

It says 125w in the specs, but if you check the CPU part number, you get this:

http://www.cpu-world.com/CPUs/K10/AMD-Phenom%20II X6 1055T - HDT55TWFK6DGR (HDT55TWFGRBOX).html

which is the 95w part.

So, if anyone wants one, that seems to be the place to get one.

 

[richierich1212]

Just an FYI, TigerDirect now has the OEM 95W 1055T for $190 shipped:

http://www.tigerdirect.com/applicati...IN#detailspecs

It says 125w in the specs, but if you check the CPU part number, you get this:

http://www.cpu-world.com/CPUs/K10/AMD-Phenom%20II X6 1055T - HDT55TWFK6DGR (HDT55TWFGRBOX).html

which is the 95w part.

So, if anyone wants one, that seems to be the place to get one.

No more 15.2% bing cash back :thumbsdown:

 

[DrMrLordX]

No more 15.2% bing cash back :thumbsdown:

Bummer. Still not a bad price, though ~$30 off would be nicer.

 

[Lazlo Panaflex]

Anyone undervolting their 1055/1090? I've been messing around with the voltages, since this board has stupid auto settings. The CPU runs hotter than hell. Gigabyte must be on dope if they think 1.475 is "normal".
FYI, I've been Prime stable @ 3.1, 1.184v for about an hour now. Turbo is enabled (16x222).

vcore: 1.184 (set to 1.35v in BIOS)
NB: 1.1v
vDIMM: 1.82v, 5-5-5-13 2T
HT Link/NB Freq: 2220
Temps: ~45c (load)

 

[LoneNinja]

Anyone undervolting their 1055/1090? I've been messing around with the voltages, since this board has stupid auto settings. The CPU runs hotter than hell. Gigabyte must be on dope if they think 1.475 is "normal".
FYI, I've been Prime stable @ 3.1, 1.184v for about an hour now. Turbo is enabled (16x222).

vcore: 1.184 (set to 1.35v in BIOS)
NB: 1.1v
vDIMM: 1.82v, 5-5-5-13 2T
HT Link/NB Freq: 2220
Temps: ~45c (load)

Gigabyte thinks the stock voltage of my Athlon II X4 630 is 1.475 as well. lol

Nice undervolting by the way.

 

[lopri]

I think the blame is on AMD in large part. My 1090T's default vCore is 1.275V, but when Turbo kicks in it goes to 1.425V which is way more than needed. My 1090T actually can do 3.6 GHz at its default (1.275V). Apparently AMD decided to be on the safe side, and all the board manufacturers receive the base BIOS from AMD so that's where the individual board's BIOS build on. (Though I am not sure why it affects Athlon II X4)

It's highly recommended to manually adjust the voltages for Thubans, whether overclocking is involved or not. Most can run with less voltages than what "auto" sets for them.

 

[Lazlo Panaflex]

Thanks Mr. Ninja That's interesting, didn't realize that the base BIOS is from AMD. Gigabyte really needs to release a more Thuban-friendly update. It's just silly that the default voltage for Turbo/CNQ is 1.475.

2+ hours Prime stable! (small FFT's)

 

[Chiropteran]

It's highly recommended to manually adjust the voltages for Thubans, whether overclocking is involved or not. Most can run with less voltages than what "auto" sets for them.

Anyone know how to reduce the voltage on a crosshair IV formula board? I feel like a noob, while I have done a bit of OCing in the past my previous computer was a Dell and now that I am back to using something I could overclock it seems like the BIOS settings and options are a lot more complicated than I remember. I have choices for offset voltage, what does that mean? I don't really see how to reduce it either, seems like I can only increase from the base, I must be missing something.

I tried using the AMD Overdrive software, which works for the most part, but when I reduce voltage it just immediately goes back up to what it was previously.

 

[Flipped Gazelle]

Built a system for my brother last week; this is all to the best of my recollection:

* CPU Model: AMD Phenom II X6 1090T BE
* CPU Stepping: CCBBE CB 1022 BPMW
* CPU Frequency: 4.1 Ghz
* CPU vCore: 1.46V
* CPU Multiplier: 17
* CPU Turbo: Disabled
* CPU NB Speed: 2700 MHz
* HT Bus Speed: 2200 MHz
* RAM Speed: DDR3-1500
* RAM Timings: 9-9-9-25-2T
* RAM Configuration: 2 x 2GB
* RAM vDIMM: 1.8V
* Motherboard: Asus Crosshair IV
* Chipset/Socket: 890FX / AM3
* Cooling: Corsair H50, Vantec Stealth & Antec Tri-cool 120mm fans push/pull
* Temps: 25c idle/51c load (78f ambient)
* Operating System: Win7 x64
* 32/64-Bit: 64
* Stable/Suicide/Untested: Stable OCCT overnight
* PSU: Rosewill Xtreme 750w