The Bulldozer week/batch thread...

[AlleyViper]

CoreTemp or CPU-Z text dump will show VID.

OT: I haven't any FX to check so I'll ask: do BD's highest VIDs vary from last non turbo multiplier to turbo multipliers? I've seen a TXT dump from a FX4100 where 18,18.5 and 19X had exactly the same 1.4125 VID, while in a review CPU-Z indicated for this CPU different stock voltage readings for each of these multipliers. I didn't understand if turbo voltage is added to 18X VID as an offset, of if it's still directly determined by that table only in those processors that have different values (unlike my first example).

[BeepBeep2]

Yes, they do. However, 18x is a turbo multiplier on FX...it is possible that all the Turbo multipliers can have the same voltage I think.

I don't have a working BD setup right now to run a dump.

[AlleyViper]

But on that particular case (FX4100 from user on another support forum), 18X isn't a turbo multiplier. It's a 3.6GHz base CPU, this is why I'm a bit puzzled if turbo voltages are applied as a predetermined offset or directly via VID table (only when differences are existent). This question is out of pure curiosity, of course.

Btw, this was the CPU-Z reading:

# of P-States 7
P-State FID 0x16 - VID 0x0B - IDD 21 (19.00x - 1.412 V)
P-State FID 0x15 - VID 0x0B - IDD 21 (18.50x - 1.412 V)
P-State FID 0x14 - VID 0x0B - IDD 15 (18.00x - 1.412 V)
P-State FID 0x11 - VID 0x11 - IDD 13 (16.50x - 1.337 V)
P-State FID 0x9 - VID 0x1D - IDD 10 (12.50x - 1.188 V)
P-State FID 0x1 - VID 0x28 - IDD 7 (8.50x - 1.050 V)
P-State FID 0x10C - VID 0x2F - IDD 6 (7.00x - 0.962 V)

[tcclaviger]

Just wanted to add my chips to the mix, two FX8150s:

Tcclaviger...|AMD FX-8150.......|1209PGN......|????.....|4816 MHz @ 1.5V.................|Water......|4970 MHz @ 1.545V <-- 1.2875 VID
Tcclaviger...|AMD FX-8150.......|1210PGT......|0495....|5015 MHz @ 1.536V..............|Water......|5275 MHz @ 1.562V <-- 1.3 VID

Ambient room temps 24c, water temp @ idle 26.1, water @ full load 31c (CPU + 2 GPU in loop). Cool N Quiet is enabled for the 24/7 settings believe it or not lol.

Aside from that information, I can tell you the higher VID chip, the 1210PGT, runs drastically cooler than the 1209 did at the same settings, 7-10c infact. The 24/7 settings on the 1210PGT at 5015 1.536 are cooler than the 1209PGN at 4700 1.488V, 4c cooler infact!

Amazing at the wide spread of heat and overclocking ability in these chips, the 1209 would not make it into windows at 5ghz or above regardless of voltage/settings. The exact same settings as the 1209 were loaded for the 1210 and a full 1.0 multi added and it booted right into Win 7 stable :O


P-State FID 0x1A - VID 0x0B - IDD 18 (21.00x - 1.412 V)
P-State FID 0x17 - VID 0x0B - IDD 18 (19.50x - 1.412 V)
P-State FID 0x1A - VID 0x14 - IDD 12 (21.00x - 1.300 V)
P-State FID 0x11 - VID 0x1A - IDD 11 (16.50x - 1.225 V)
P-State FID 0xB - VID 0x25 - IDD 8 (13.50x - 1.087 V)
P-State FID 0x5 - VID 0x2F - IDD 7 (10.50x - 0.962 V)
P-State FID 0x10C - VID 0x37 - IDD 5 (7.00x - 0.862 V)

[tcclaviger]

Just wanted to comment on BeepBeep2's comments concerning the CVF, your spot on in my experience. While I throw a little caution to the wind about temps and voltages, it is a conscious decision and I know I am very much risking degradation of my CPU at the voltages I am running.

A note for you, not sure if this holds true for all BD, but both of mine it did, use a LOT of clamping force, I tightened till I was scared I was gona crack the MB, then twisted the thumb nuts two more full revolutions!! If I went lighter on the pressure, temps were insane, I think it has to do with flattening the bow on the XSPC Rasa and a huge die on the BD.

For an idea of the heat difference above 1.5vcore vs a Phenom II on a very high flow custom loop:

Phenom II 24/7 settings:
4475 mhz 1.608vcore
3008mhz IMC 1.35v
Load max temps (CPU Socket temps): 52c

New Bulldozer 24/7 settings:
5015 mhz 1.536vcore
2600mhz IMC 1.263v
Load max temps (CPU Socket temps): 71c
Load max Core Temps: 59c

and this chip is running way way cooler than my last one (replaced due to pathetic OC cap and insane heat output. Below is my old one.

Old Bulldozer 24/7 settings:
4816 mhz 1.5vcore
2580mhz IMC 1.31v
Load max temps (CPU Socket temps): 76c
Load max Core Temps: 66.1c

[FlanK3r]

very interesting...200 MHz difference between FX and only one week old.

[AlleyViper]

Is "P-State FID 0x1A - VID 0x14 - IDD 12 (21.00x - 1.300 V)" a multiplier misread? I thought there was supposed to be a stock 3.6GHz P-State in that place, and this one even follows your 1.3V for 1210PGT.

[BeepBeep2]

Really interesting things going on with the VID readings.
At first I thought maybe the 0x11 Pstate (16.5x) is the right VID...now I am just confused.

I apologize for saying that 18x was a Turbo multiplier, I don't know what was wrong with me >< I am very sleep deprived very often, been to numerous doctors and specialists trying to sort out my insomnia/overall tiredness.

[tcclaviger]

This morning I pulled the MB out to re-TIM the NB, as it was hitting shutdown temps for some reason at 90C :O I also installed a modified Asus back plate, since the XSPC Rasa has none for AMD mounting, that way I am crushing on steel and not my MB, this allowed a LOT of torque on the heat sink, with incredible results!!

I know it's not breaking any records, but for a mostly gaming computer, on a very sub-optimal OS, I think it's actually pretty good. RAM is Hynix Mushkin 993997 kit.

[BeepBeep2]

IBT, OCCT Linpack, LinX, are all super easy to pass on Bulldozer.

Prime95 Blend is super hard to pass...

[AlleyViper]

Thanks for looking, you don't even need to apologize. That P-States table is really confusing and that's the reason I brought it to the discussion; more so, FX4100s don't get much attention. There's also a big chance that CPU-Z might do some misreads, judging by post #229, where a FX8150 doesn't even show stock 18X, but 21X where it probably should be.

[I.nfraR.ed]

Finally got my RMA replacement and...it must be the worst FX-8150 in the world. It might be the board, but I doubt it.
It's way crappier than my crappy 8120, which could do over 5GHz and I thought all of them could...
CPU VID is 1.3V, CPU-NB VID is 1.215V. Idle vCore for this run = 1.5V, load = 1.56V (mobo has a vdroop mod).
Still CPU-NB is doing ok: 2700MHz with 1.33V, 2800 boots with 1.36, but unstable. Memory controller seems fine as well.
RAM is 2x4 Crucial Ballistix Tactical @ 1.65V.
Really hope it will do ok on LN2, otherwise it will be useless.


I.nfraR.ed....|AMD FX-8150.......|1209PGN.....|0056.....|............... .........................|Water......|4784 MHz @ 1.56V (23c)

[BeepBeep2]

Finally got my RMA replacement and...it must be the worst FX-8150 in the world. It might be the board, but I doubt it.
It's way crappier than my crappy 8120, which could do over 5GHz and I thought all of them could...
CPU VID is 1.3V, CPU-NB VID is 1.215V. Idle vCore for this run = 1.5V, load = 1.56V (mobo has a vdroop mod).
Still CPU-NB is doing ok: 2700MHz with 1.33V, 2800 boots with 1.36, but unstable. Memory controller seems fine as well.
RAM is 2x4 Crucial Ballistix Tactical @ 1.65V.
Really hope it will do ok on LN2, otherwise it will be useless.


I.nfraR.ed....|AMD FX-8150.......|1209PGN.....|0056.....|............... .........................|Water......|4784 MHz @ 1.56V (23c)

My first FX-8150 only did 4.9 on water.

Run your CPU at stock settings without LLC (on different board?), and stock cooler, see if it passes Prime95. I doubt it will > RMA

[FlanK3r]

Infra: dont fear me...because my new chip is 1210PGN ....only one week later and the same PGN. I want not

[TheBreezyBB]

I'm getting a new Fx-8150 very soon, I'll post the batch/week number when i get it. I hope i get a golden chip

[TheBreezyBB]

I got the same Chip as Infra

But this is the first validation from a quick OC not tested yet.

[FlanK3r]

push up it harder ...Now Im testing one week newest batch, 1210PGN (no OC, only try some setings in BIOS)...Im interesting for this batch...But I have a bit uncertain about quality of chip...1209PGN looks not much good.

[TheBreezyBB]

I've tried to run Prime95 for a bit, but the chip is way too hot for Air! it reaches 72c+ easily in a matter of seconds!! I'm not happy so far... I think i'll run it @ 4.4GHz/1.35v and see if it still running hot.

[FlanK3r]

So, my CPU VID is 1.3125V and CPU-NB 1.1875V....We will se, what it can do

[TheBreezyBB]

Mine is a high leakage chip it seems CPU VID 1.287... and i only have Thermaltake Frio no plan to upgrade anytime soon.

[TheBreezyBB]

If only this chip didn't run very hot, i would be running 4.8-5.0GHz easily 24/7


4.8


5.0

[FlanK3r]

this chip like not much voltage, think 1.45V will be limit for it, for more and higher clocks ( 5GHz+) will need some as awesome watter cooling setup or SS

[TheBreezyBB]

yeah, that's what i think. I'm running 4.5Ghz@ 1.356 idle and 1.368 100% load stable. which i'm pretty happy with right now.

[BeepBeep2]

That is a very good CPU Breezy.

[TheBreezyBB]

Thanks Beep! I need to go full water soon though could help me a lot