my Q6600 B3 ES and Q6600 G0 retail both did 500FSB on P35 but on 965P/975x limited to around 400-420FSB (965P) and 375-400FSB (975X)Originally Posted by George_o/c
Or maybe, that happened because first batches of Q6600 (like B3 stepping) had issues ?What I imply with that, is that maybe it's happening because of the CPUs just like Fr3ak said ...
BTW, nice chip Brian y.
---
On the DX58SO with 920ES under Ultra 120 Extreme:Hell, I couldn't get over 150 with the intel board..
- Just, call me dami1 -
- We do not born expert, but we can become one -
Not clocking as high as everyone wants and having a wall at exactly 222 MHz on several boards from different manufacturers being reported from a lot of people is different I would say. Probably that limit was not put there by intention, maybe it is just a coincidence.
In the past there was a range, where the FSB was maxed out, just like eva said, but having a wall at exactly 222 MHz implies that something has to go terribly wrong after that ^_^
オタク
"Perfection is a state you should always try to attain, yet one you can never reach." - me =)
i've asked gigabyte engineers
hopefully they can shed some light WTF is going on with BCLK limit
Probably they don't even know. When it came to strap settings, they didn't know either, because it wasn't documented, from what I know. Probably only Intel can shed some light =/
オタク
"Perfection is a state you should always try to attain, yet one you can never reach." - me =)
Omg and everyone was saying they would not oc close to 5.5ghz
Yes same problem with engineers a orhers marks ..i've asked gigabyte engineers
hopefully they can shed some light WTF is going on with BCLK limit
Maybe we hope the better BCLK with the low QPI (same : 3.2 or 4GT/s or others) but the performance maybe less good (if it's possible)
This is exactly the reason I didnt jump on board at launch. Figured I would wait for new cpu stepping, or new mobo revisions before buying... always a few kinks to work out.
Hey guys !
Why didn't you create a topic dedicated to our results, which will gather informations in a large database about ocing i7 ? I think it will be better
I think you're looking at that
idk, not participating, just watching you guys
Here's what i get on water...
BCLK is connected to QPI link speed and lowest divider is 36x, so if you use 220BCLK your QPI speed is 7920GT/s. Around there is the limit at the moment. We need lower dividers, cause QPI link is not limiting performance if it will stay around 6500-7500GT/s. QPI divider 28x or 32x would be cool.
If we would have divider 28x we could hit ~280BCLK. Of course there might be other limiting factors, but maybe X68/ P55 chipset will bring us more overclocking potential. X58 can't handle more QPI link speed at the moment.
I'm afraid that we will have to wait until 2009, when LGA 1160/1156 will be released. (Mainstream, Lynnfield etc.)
Is this the answer you wanted to hear
You are as good as your samples are!
Yep, agreed. All the boards can reach 222, so it appears that this limit is due to the QPI rather than to the chipset.
Possessors of 965 don't worried about that because QPI is not important, no influence on performance it seems. But 920/940's owners will be much more deceived...
gigabyte have a slow mode which runs QPI at 100MHzBCLK is connected to QPI link speed and lowest divider is 36x, so if you use 220BCLK your QPI speed is 7920GT/s. Around there is the limit at the moment. We need lower dividers, cause QPI link is not limiting performance if it will stay around 6500-7500GT/s. QPI divider 28x or 32x would be cool.
If we would have divider 28x we could hit ~280BCLK. Of course there might be other limiting factors, but maybe X68/ P55 chipset will bring us more overclocking potential. X58 can't handle more QPI link speed at the moment.
I'm afraid that we will have to wait until 2009, when LGA 1160/1156 will be released. (Mainstream, Lynnfield etc.)
Is this the answer you wanted to hear
it doesnt help BCLK
Last edited by dinos22; 11-26-2008 at 03:03 PM.
Not just Gigabyte. Biostar, EVGA and .. um I forgot, MSI (?) have that slow mode feature too, but you are right, it does not help BCLK. Some boards allow you to change QPI and Uncore multipliers in BIOS for non XE CPUs, but in fact it doesn't do anything :P Didn't check on all boards, though.
Last edited by Fr3ak; 11-26-2008 at 02:23 PM.
オタク
"Perfection is a state you should always try to attain, yet one you can never reach." - me =)
placebo features on my 300€ board? no wai
Still no sighting of DrWho's cherry picked i7's? Anyone know when he'll have results up?
Notice that I added "for non XE CPUs", Hornet331. Works fine on a 1000€ i7-965 XE
オタク
"Perfection is a state you should always try to attain, yet one you can never reach." - me =)
It might be a week or two or more. Whenever they get together, and whenever they decide they want to post about it, or *if* they want to post about it.
What are you talking mate
QPI is Quick path interconnect speed. It is marked as GT/s Gigatransfers per second. You see this value in CPU-Z front page "QPI link" At stock speed it is saying ~3200MHz, when it should say 6400GT/s(2x3200GT/s). When you raise BCLK this value will raise too and with lowest QPI multiplier (36x) it will go up to ~8000GT/s before it locks. Now if we would have lower multipliers than 36x we could get our BCLK higher before that ~8000GT/s limit.
If you remember AMD 64 days and 1X to 5X HT multipliers, you will see what this issue is. Back then you had to lower HT multiplier to 4X or 3X to get your HT speed up to 200 to 250.
After all this processor is designed for servers and these really high QPI speeds were not what they had in mind. Maybe this will be fixed with new chipsets or 1160 processors. If QPI multipliers on X58/core i7 are limited to 36x on hardware level, there is nothing we can do. Just LN2 to CPU and Chipset and we might see some scaling
Are we on the same page now
You are as good as your samples are!
heheh have you tested the GB board
in bios when you select Slow mode it actually says 100MHz instead of 4800 or whatever it is normally lol
i dont know if it actually sets it at that but thats what the bios states
My first attempt,
Turn on,
Bios,
Change settings,
restart,
Windows,
CPU-Z Validation,
It was all done in less than 5 minutes.
Who did say Nehalem would not overlock at all?
An idiot said that.
Who did say Nehalem would not overlock at all?
I think The Inq had something about it a long time ago.An idiot said that.
Yep, and Fud thus my post.
Posting Permissions
Reply With Quote
Bookmarks