Nice gundamit. Nice to see your 146 doing better than your 144. But is an additional 100mhz worth an additional $60? Being the anal best-bang-for-buck guy that I am, I broke down what we have so far:
144 3.0ghz o/c / 146 3.1ghz o/c (so far) = 3.3% increase
144 $160 retail / 146 $220 retail = 37.5% price premium
You be the judge. But it's going to take alot more for me to feel bad about choosing the 144.
xenolith -> oc = 0.000 %
It's depending of the die (stepping)
But 146 = 10x max
The last thing anyone should be feeling is "bad" about buying either. Rather thay should be thanking their lucky stars and Vandi for giving us this opportunity to play with the best bang for buck CPU this year. And we did it before what Charlie described as the "unwashed masses" get wind of whats going on. Hopefully Vandi keeps us in mind the next time something like this comes along even though I'm sure it was a big headache and time consuming putting these orders together.Originally Posted by xenolith
Nice gundamit. Nice to see your 146 doing better than your 144. But is an additional 100mhz worth an additional $60? Being the anal best-bang-for-buck guy that I am, I broke down what we have so far:
144 3.0ghz o/c / 146 3.1ghz o/c (so far) = 3.3% increase
144 $160 retail / 146 $220 retail = 37.5% price premium
You be the judge. But it's going to take alot more for me to feel bad about choosing the 144.
Hmmm ... might not even be 100mhz difference my two CPUs. I do like the 146 because of more chances to hit the sweet spots on memory. But being a cheap bastard I am in awe of the 144!
Sourcing parts for a mildly over clocked abacus.
The volts keep streaking up while the clock comes up just a little. Nearly topped out now.
I think I'll get back to my old BH-5 to see if I can get <24:20 in 32M.
Sourcing parts for a mildly over clocked abacus.
Nice clocks, what kind of temp do you get under load at the 1.552v?
That's why I was going to get the 146. Having the extra (10x) multi was very attractive. But by what I've seen so far, high HTT doesn't appear to be any problem with the right setup.xenolith -> oc = 0.000 %
It's depending of the die (stepping)
But 146 = 10x max
146 had 10x multi which is much better than 9x. Not so sure about me sweet spots... I need a /13 divider
Nice clocks BTW gundamit.
Phenom II 940 BE / ASUS M4A79 / HD5770 Crossfire
3770mhz CPU 2600mhz NB | DDR1040 5-5-5-15 | 900/1250
The ss on link #352 with the 144 at 3078.9 @1.472v with 2gb ram compares great with the 146 at 3140.8@1.552v with 1gb ram.The volts keep streaking up while the clock comes up just a little. Nearly topped out now.
I think I'll get back to my old BH-5 to see if I can get <24:20 in 32M.
144 pi32m=24m24.391
146 pi32m=24m32.937
I'm still under 50c but that might get better when the block settles.
Sourcing parts for a mildly over clocked abacus.
Yeah but to be fair to the TCCD they aren't hitting thier sweet spot and I didn't spend time playing with the timing. Love the patriots though ... my memory ... not so much the team.
Sourcing parts for a mildly over clocked abacus.
still seems a tad low to me
---
@gundamit,
I don't want to be negative... But either you must have had CPU-Z and ITE SG both open during that run above or I don't know what else. But 24:32 is really bad for 3140Mhz an 1GB RAM, even at 7-3-4-2.5-1T...
EDIT: I would change that RAM to BH-5 for further testing, you know. Those times can make people suspicious, especially when looking at such high clocks on air/water. I just noticed that you have Dynamic Idle Cycle Counter on Disabled... That will kill 32M more than TRC=12,
Last edited by bachus_anonym; 10-02-2005 at 10:38 AM.
[QUOTE=eva2000]still seems a tad low to me
LOL, I'm comparing him to himself. You are one of the all time greats with memory. I'm sure you could take a $150 set of 2gb ram and beat a $300 set.
gundamit needs to tighten Idle Cyle to 16clks and enable Dynamic counter and see what happens i guess
this is 12x245mhz 2.5-3-2-5 with 2x1GB Patriot PC3200LLK
can't wait to get my opteron 144 and 146s to put some ram through their paces as well
---
@eva2000,
yeah, I edited my above post before I saw yours... I just get suspicious when I see such bad times at very high clocks
Very powerfull your memos, god luck whit the 315 @ 3-3-3-5-7-14-1t (more volts ?)
When do you put your opty in subzero temps ?
hug'zz
Portugal@Home status: running
that's great MM....are you able to do any SPI with 310MHz
Gee guys. Here I was all set to go to my Mushkin BH-5 and suddenly I get tips on tightening my timing on the TCCD. I did mention I didn't do anything with the timings right? In any case I appreciate the input and give it a try and will report back in.
Sourcing parts for a mildly over clocked abacus.
WOW
I always did love quality silicone.
----------TOWER OF POWER----------
Dell 2405 FPW 24' LCD
Amd 64 +3200 @ 2.929ghz @ 1.5vcore
Msi Neo2 Plat Does ~385htt
Ocz Pc4800 Plats @ 292 2.5-3-3-7 @ 2.85v
Ocz Powerstream 520 watt
Ati 9800pro @ 470/360
2 X WD 36gb Raptors Raid 0
WD 16Mb Sata2 250gb
All on air, loud and very bright.
*cough*silicon*cough*
no, you mean boobies!
MM, 315mhz would be amazing. get to work
Got a fan over those memory sticks? No? Well get to it before you kill them
gundamit which bios are you using? I heard you have to use 704-BT(A) to get TCCD working, is this correct?
DFI LanParty UT nF4 Ultra-D R.A02 | AMD Opteron 146 0530APMW | 2048MB Team Xtreme PC4000 Cronus Micron 5B-F | XFX 6800GT PCI-E | OCZ PowerStream 520W
time to drop the multiplier
Per your suggestions I've changed the timings and dropped some time from the 32M run at 3.14ghz. Thanks.
Sourcing parts for a mildly over clocked abacus.
Posting Permissions
Reply With Quote
Bookmarks