The Bulldozer week/batch thread...

[Daveburt714]

Kevy: Are you just shooting for low power?
My Thubans crapped out ~56c, but most FX chips are fine all the way up to ~70c...
If your load temps are 48 you've got lots of headroom left.

Does anyone have any results from newer 8120's?
The last thing I need is another Zambezi with PD being so close, but I'm pretty bored, and MicroCenter has the 8120's for $140...

All my FX chips are older with lowish Vids, and they all run pretty hot in the 4.6-4.7 range on ambient.
Their good chips, I just can't help but wonder if they've gotten any better with time...

[I.nfraR.ed]

I don't know about very new batches, but for me it's going worse with every newer chip.
The first 8150 actually was pretty good, but died shortly after I put it on LN2, the next 8120 was worse and the 8150 from RMA replacement is absolute crap.
Too bad the first one died before I even manage to do something with it under LN2...it looked promising.

Now I don't even have the will and passion to test the other 2 chips.

[FlanK3r]

I tried 4 chips...3 from all are old batches. Obe good, but very hot (still I have it in workstation), second good and low power consumption (7300 LN2), third average at air, "good" with LN2 (7500+, but died in Superpi 7500 MHz after restartin PC) and last one forth is 2012 batch. Average at air, LN2 7300+. I will take last one in next month, we will see...After this no more sense for Zambezi and will waiitng for Vishera

[demonkevy666]

Kevy: Are you just shooting for low power?
My Thubans crapped out ~56c, but most FX chips are fine all the way up to ~70c...
If your load temps are 48 you've got lots of headroom left.

Does anyone have any results from newer 8120's?
The last thing I need is another Zambezi with PD being so close, but I'm pretty bored, and MicroCenter has the 8120's for $140...

All my FX chips are older with lowish Vids, and they all run pretty hot in the 4.6-4.7 range on ambient.
Their good chips, I just can't help but wonder if they've gotten any better with time...

I just wanted lower load temps on stock settings.
My room isn't very big.

[Daveburt714]

I had a little extra cash and some spare gear to put together a Craigslist rig, so I went and picked up a new 8120 @ MicroCenter for $140...
What can I say, I was bored!

I was curious about the newer chips (all mine are batch 11xx), so I thought I'd post some results incase anyone else was interested.
First, heres the newest batch they had on hand, not very new but it was the latest week they had:


The first thing I noticed about this chip was that it ran MUCH cooler than the older ones...
I realize this is only one sample and YMMV, but it seems like they've been working on the leakage (temps/power) more than the clocks.
It's NOT a bad chip, and clocks decently on ambient cooling but it still doesn't match my older ones other than running lower temps.
I should have pulled out the Kill-a-Watt and tested power consumption. I get the feeling it'd be better (may test later), but once I saw it couldn't quite keep up with my older samples clock wise I pulled it and went back to the 8150 (my best ambient w/c clocker)...

Anyway... Here's what I got with this chip, 28 minutes (P95 Custom/Max Mem).
I was going for a 30 minute run, it crashed @ 28 but I was also surfing while blending, so IMHO this should be stable for daily use.
Max temp only hit 58c, compared to 68c on the 8150 (die sensor), so theres definetly some improvement on the process...

[FlanK3r]

what max Cinebench Dave? Can u test it?Thx

[BeepBeep2]

I had a little extra cash and some spare gear to put together a Craigslist rig, so I went and picked up a new 8120 @ MicroCenter for $140...
What can I say, I was bored!

I was curious about the newer chips (all mine are batch 11xx), so I thought I'd post some results incase anyone else was interested.
First, heres the newest batch they had on hand, not very new but it was the latest week they had:


The first thing I noticed about this chip was that it ran MUCH cooler than the older ones...
I realize this is only one sample and YMMV, but it seems like they've been working on the leakage (temps/power) more than the clocks.
It's NOT a bad chip, and clocks decently on ambient cooling but it still doesn't match my older ones other than running lower temps.
I should have pulled out the Kill-a-Watt and tested power consumption. I get the feeling it'd be better (may test later), but once I saw it couldn't quite keep up with my older samples clock wise I pulled it and went back to the 8150 (my best ambient w/c clocker)...

Anyway... Here's what I got with this chip, 28 minutes (P95 Custom/Max Mem).
I was going for a 30 minute run, it crashed @ 28 but I was also surfing while blending, so IMHO this should be stable for daily use.
Max temp only hit 58c, compared to 68c on the 8150 (die sensor), so theres definetly some improvement on the process...

Dave, what VID?
High VID launch chips ran mighty cool too

Ones like yours and mine, not so much.

loud_silence sent me a 1210PGT (I think PGT, maybe PGN) here, haven't tested it on anything but LN2 yet, but from what I hear it is promising.

[Daveburt714]

what max Cinebench Dave? Can u test it?Thx

I don't think it would have done all that well, didn't test though...
I took the v's all the way up to 1.525 (since it ran so cool) and couldn't get it to blend for more than a couple minutes over 4.7ghz.
It may have done OK on Cinebench, but I didn't test it. I'm sure I'll put it back in eventually, I'll test it for you then.

Beep:
The VID on it isn't really that high (1.3125), that's the reason I was surprised it runs so cool.
OTOH, it doesn't scale very well with voltage either (past 4.6). Didn't really give it enough time testing...
I'll probably put it back in this weekend and give it a little more time, I need to get more TIM though.