I take it your also using 32bit xp? the 64bit OSes take more to get stable. I haven't opened up my box with the i7 system yet because I'm waiting on my GTZ waterblock.
Printable View
Yes. It must be a revision or something similar, (1.25, 1.27,5) the difference of 3V is minor but worth :D
We need to know how accurate his motherboard Vdrop works. If it gets in Windows with 1.20V and the Bios shows 1.25 then in load times the voltage must be also different.
HI~
Testing 3838B079 code ...
http://cfs14.tistory.com/original/2/.../49387a25a184c
http://cfs14.tistory.com/original/10.../49387a25b07f6
nice volts and temps nam.
How is this one?
http://chuckbam.com/i7_P6T-D/_i7Box.JPG
Yeah this really sucks...
Thank you can you please elaborate on how to update the chipset drivers though? I run the utility but there seems to be no change can you explain how to update? I created the same directory but it doesn't seems to do anything or maybe I just don't understand exactly what you are talking about...
I could not resist :)
Batch Number:
http://dpweqa.blu.livefilestore.com/...h%20Number.JPG
http://dpweqa.blu.livefilestore.com/...%20Threads.JPG
Is there anything wrong with having a "A" code FPO/Batch number?
I just got my 965. Still waiting for my other parts. The number is 3836A383. I am hoping for the best.
I'll have my results up as soon as I get my TRUE.
But for now I have it at 3.4GHz with 1.153v (multimeter).
My batch # is 3838A646 and C0/C1 stepping
Ya, it does nothing without the -OVERALL. Intel is playing it safe.
Create a shorcut. Go to browse and find your utility and after with a space add -OVERALL
For Example, c:/Intel/setup.exe -OVERALL
Or use Run. put the path of the utility and place -OVERALL after.
You are using MS 2006 inf. Intel has the correct ones for better/ proper x58 setup. I think that may fix your file transfer problem. If you still can't figure it out, post the name and path of you utility.exe and I will send you a shortcut. If you update and still have slow speeds, I may be able to help you with that too.
I am pretty sure I have the correct INF the one on my disc is the correct one and I did the -OVERALL and it went through the process but there was no change, which makes me think I had the correct version installed but I'm not sure how to tell.. Here is what I did I put the Inf utility in C:/Intel/INF_allOS_9.1.0.1007_PV.exe -OVERALL then double clicked and it ran.. It just immediately said successfully installed..
I am trying different hard drives right now to see if the problem could be the hard drives....
With -OVERALL it will update them anyway. Go into device manager and look at the date/ manufacturer of one of your x58 devices. If it Intel, you are probably in good shape. If it is MS 2006, you need updating.
Is your update file named INF_allOS_9.1.0.1007_PV.exe ? I have seen the same utility named different.
Put this on your desktop and click it.
http://chuckbam.com/Posts/INF_allOS_9.1.0.1007_PV.exe
Yeah man thanks I have narrowed the problem down to these 1TB Seagate Drives with 32mb cache.. Anytime I try and transfer to these drives from anything else my transfer rates are all f'd up so now I need to know if it's a Seagate problem or a 32mb cache problem.... I would bet on the X58 and the 32MB cache being the problem... Looks like we need one more Inf update somebody call intel and tell them Nvidia called and they want their board back....
Anyway I can transfer from WD Velociraptor to WD velociraptor no problem but any transfer to the Seagate and it's like 5mb/sec versus like 80-100 normally so booo...
I notice that a lot of you are going ape with the overclocks (4.0ghz+) but have sacrificed hyper threading. Anybody have an idea how far most people will be able to stably push an i7 920 with hyperthreading on (using air preferably)?
Batch: 3837B080
Super Pi 1M - 4010Mhz "211x19" - 1.152v - Windows Vista 64 bit:
http://img518.imageshack.us/img518/2...8463jn4.th.jpg
Super Pi 2M - 4010Mhz "211x19" - 1.152v - Windows Vista 64 bit:
http://img179.imageshack.us/img179/2...7339xb9.th.jpg
Super Pi 4M - 4010Mhz "211x19" - 1.152v - Windows Vista 64 bit:
http://img234.imageshack.us/img234/6...5647ei4.th.jpg
Super Pi 8M - 4010Mhz "211x19" - 1.152v - Windows Vista 64 bit:
http://img522.imageshack.us/img522/7...5965yx5.th.jpg
Super Pi 16M - 4010Mhz "211x19" - 1.152v - Windows Vista 64 bit:
http://img522.imageshack.us/img522/4...4845kf7.th.jpg
Super Pi 32M - 4010Mhz "211x19" - 1.152v - Windows Vista 64 bit:
http://img515.imageshack.us/img515/1...7866pc1.th.jpg
Very good!:up:
Batch: 3838B079
http://cfs11.tistory.com/original/18.../49391576b0ed7
4G 1M
http://cfs11.tistory.com/original/25.../493915763c90b
4G 32M
Have you tried 4.2Ghz?
Answer: If yes
How much voltage to get 4.2ghz stable?
Thanks,
An update: I can run Super PI 32M with 1.5 4.0GHZ too but it freezes when it starts priming.
All retails as of since the release have been the one mentioned by Hornet at the above post.
http://s31.photobucket.com/albums/c3...boy/Core%20i7/
My Core i7, I can't wait to get a motherboard for it! :cool::yepp:
I'm having issues with Prime95.
I've had my i920 OCed to 4.2 and played online for like 3 hours, no reboots, no lockups, etc.
When I run Prime95 small FTT, I get it freezing on me in like 1 min.
I've raised Vcore even to 1.48 for 4.2 Ghz and still the same thing.
I think I may know what it may be. Tell me what you think.
I am using Gskill 7-7-7-18 1.9V memory.
My memory settings are on auto, but I think it is trying to run 1600 MHz on like 1.6 V, which I think is causing the instability when all 4 core are being stressed.
What do you guys think?
I've ordered some OCZ gold 8-8-8-24 1.65 1600MHz memory which should be arriving next week.
I've run Prime 95 on stock and it works flawlessly, but then again the memory is not clocked to 1600.
So, one way to test this theory would be to lower the Mem ultiplier to 6 and see if I still have that problem?
@Lukee:
Most games doesn't utilize all cores (i don't know any game which can do 8, even 100% on 4 will be hard to find) so prime is doing much more load on your cpu than games.
Simple, cause nehalem uses the PCU, the thermal spec of the perviouse C2 was the Tcase max, but there is no such value for Nehalem.
Instead Nehalem uses the PCU to stay within the thermal envelope (Tjmax or max TDP) and gives the users the best performance.
Also just for info, there was no way to measure the Tcase max (Intel thermal spec) for C2 unless you would modify the IHS and house a thermo couble in the center of the IHS.
So the thermal spec for the C2s was kinda worthless. :yepp:
Here are my settings
20x200 bios f4e gigabyte- X58 extreme
turbo on
HT off.
eist, c1 also off.
qpi 36x
uncore 16x
ram auto.
loadline on
1.40vcore.
qpi/vtt 1.29
cpu pll 1.84
qpi pll 1.18
ioh core 1.120
ich 1/0 1.58
ich core 1.140
Any suggestions?
Try my chess test..it's a full working chess program and will use your 8 threads at 100%
Most people have to clock down their speed,because not stable anymore like you find out with prime!
Look to my topic and these are the highest stable clocks for the moment who has a core i7:
http://www.xtremesystems.org/forums/...d.php?t=202139
JP.
Today I found stability for the 4050Mhz:)
"214x19" - 4066Mhz - V.core 1.23v:
http://img176.imageshack.us/img176/2...5187di7.th.jpg
Tomorrow I try stability for the 4100Mhz.:)
I think the 4200Mhz are very difficult to achieve stable.:(.....
because the temperatures are high.:shakes:.......
but I'm not complaining.:up:
you and your golden cpu :P
I only show C0. Is C1 better?
http://chuckbam.com/i7_P6T-D/-CPUZ1.PNG
Use this version of Cpu-z:
http://www.cpuid.com/download/cpuz_149.zip
By:)
I have done some more tests, interesting ones I would say. :)
Well first I have to say that I tried to complete the Prime 95 test but it was not possible due to the high temperatures
That is the reason I keep it 4.2Ghz with 1.35V Hyper-Threading off.
First: Intel recommends it below 1.35, not to care nonetheless.:p:
Second: Load times are 10 to 20c degrees less.
The first screenshot is at 4.2Ghz with 1.37,5 which finished 2 tests as Super PI 1M and 32M.
http://dpweqa.blu.livefilestore.com/...ETh_Uww/20.JPG
The second screenshot is the reason why I can't go ahead with this. I do respect 10%. An example is if the chips runs with a TJMax of 100c then the maximum I would let is 90c. This is far higher than my 10%, check the Realtemp program well. You will see it reached 100c and it was on the 3rd Prime 95 test 1024k which is a bit further than the beginning of the test.
http://dpweqa.blu.livefilestore.com/...IhapRuA/21.JPG
The conclusion is simple and can see by anybody interested on the topic. If you need to use 8 threads with this speed at 4.2 using the best of the best aircolling solutions the Thermalright true copper 2 kilos then It would be better for you to go water cooling solutions as you will get a better voltage than me with 1.38V 8 threads hyper-threading on.
As for me I will stay with my 4.2Ghz at 1.35V 24/7 4 threads no problems. The reason is again very simple, it is cooler with 4 threads, less voltage, programs which I most use can not even use 4 threads let alone 8 threads.
I hope this helps,
Thanks for the info on HT Metroid, sounds like good logic to me :)
I originally had the same chip as Gligy, but when I went to pick up the rest of the system today they had a fresh order in, so I grabbed one of those instead, as well as upgraded to 6Gb of ram.
The batch number I ended up with was 3839A459, I haven't seen anyone with one yet so I felt brave. Well I'm off to throw the system together. I should have some info later this weekend.
I am getting old. I go more for power management and 24/7 operation. I am happy with 30%. And, that was my target. I idle at 31C.
Does anyone have what would be the default volts for the setting (other than CPU and DRAM). I still like to set all my volts manually. If you have that you could save me some time.
Great OCs people!!! I wish I could see some 4.0 GHz benchmark results.
Sandra 1560 Summary 12/4/08
----------------------------------------
Processor
----------------------------------------
Benchmark Processor Arithmetic
Dhrystone iSSE4.2 : 84.28GIPS
Whetstone iSSE3 : 71.73GFLOPS
igher index values are better.
----------------------------------------
Benchmark Processor Multi-Media
Multi-Media Int x16 iSSE4.1 : 174.53MPixel/s
Multi-Media Float x8 iSSE2 : 135.16MPixel/s
Multi-Media Double x4 iSSE2 : 74.27MPixel/s
Higher index values are better.
----------------------------------------
Benchmark Multi-Core Efficiency
Inter-Core Bandwidth : 41.66GB/s
Higher index values are better.
Sandra 1560 Summary 12/4/08
----------------------------------------
Benchmark Cryptography
AES256 CPU Cryptographic Bandwidth : 620MB/s
SHA256 CPU Hashing Bandwidth : 579MB/s
Higher index values are better.
========================================
Memory
----------------------------------------
Benchmark Memory Bandwidth
Int Buff'd iSSE2 Memory Bandwidth : 24.56GB/s
Float Buff'd iSSE2 Memory Bandwidth : 24.57GB/s
Results Interpretation : Higher index values are better.
Higher index values are better.
-----------
Total Memory : 11.99GB
Memory Used by Test : 6.00GB
----------------------------------------
Benchmark Memory Latency
Memory (Random Access) Latency : 65ns
Speed Factor : 55.70
Lower index values are better.
----------------------------------------
Benchmark Cache and Memory
Cache/Memory Bandwidth : 76.47GB/s
Results Interpretation : Higher index values are better.
Speed Factor : 21.60
ower index values are better.
-----------
Cache Information
Internal Data Cache : 272.39GB/s
L2 On-board Cache : 241.40GB/s
Higher index values are better.
========================================
Storage Device
----------------------------------------
Benchmark File System
Drive Index : 141.54MB/s
Results Interpretation : Higher index values are better.
Random Access Time : 11ms
Lower index values are better.
----------------------------------------
Benchmark Physical Disk
----------------------------------------
Drive Index : 132.12MB/s
Higher index values are better.
-----------
Random Access Time : 13ms
Lower index values are better
========================================
Video... sucks!
----------------------------------------
Benchmark Video Rendering
Dx10.1 (SM4.1) Float Shaders : 72.29MPixel/s
Dx10.1 (SM4.1) Double Shaders : 10.88MPixel/s
Higher index values are better.
-----------
Dx10.1 (SM4.1) Float Shaders : 108.06kPixels/s/MHz
Dx10.1 (SM4.1) Double Shaders : 16.26kPixels/s/MHz
Higher index values are better.
----------------------------------------
Benchmark Graphic Processing
Native Float Shaders : 107.02MPixel/s
Emulated Double Shaders : 10.06MPixel/s
Results Interpretation : Higher index values are better.
----------------------------------------
Benchmark Graphics Bandwidth
Internal Memory Bandwidth : 45.08GB/s
Data Transfer Bandwidth : 4.73GB/s
Higher index values are better.
----------------------------------------
.
I am loving these new chips. Still testing. This is with HT off an mem at 1520. Stock cooler in my very cold basement.
Multi x21 running prime ????
Core i7 920 = multi x20
multi x21 only with Turbo ON , and I think its only in single thread App , no??
btw: Update
Core i7 920 @ 3.9Ghz
BCLK : 195
Multi : x20
ddr3-1560Mhz 9-9-9-24
Turbo OFF
HT ON
1.28v Full
(prime95@Blend)
http://img152.imageshack.us/img152/1038/390015be9.jpg
:up:
Got my i7 920 rig up and running last night, I have some questions.
i7 920
giga x58 ud5 bios F4i
3x2gb patriot 9-9-9, 1.65v
21x196
QPI 3600
mem ratio 8, uncore 16
with cpuz showing 1.45v, multi 20 in bios,
four cores enabled, HT off, I'm priming at 4.1 Ghz, but seems I don't know how turbo mode works. I have Intel's turbo function set to 'on' and giga's enhancement set to 'standard'. At all times, as soon as I'm on desktop, CPUZ reports 21 multi stays at 21 through priming. I thought I was suppose to get the extra +1 multi only when I'm loaded, like when priming. Is there a way that I can get a +1 boost over idle, like when I'm playing a game?
Also, I'm seeing a large difference in my 4 cores temp, 1, 2 are around 63C during prime, 3,4 are 76C. This 13C difference exists all the time, regardless of loading. Is this missreporting, are two cores broken?
So far, seems the only thing that affects my OC is Vcore. Is there any other voltage, like Vtt, that gives me more OC? I have Vtt at 1.36, but I really have no idea what it should be.
Finally, at 4.1 G I'm getting only 10.06 super pi for 1M calc. This seems low a little low in that I think I've seen sub-10 at around 4.0G from others. I believe that perhaps tighter memory timings would help performance for a given CPU clock, I now have my mem set to SPD/auto. What about some of the other ratios? How can I get more performance other than increasing CPU clock?
As promised here are the 4104Mhz stable with "Prime 95".
"Prime95" a 4104Mhz " 216x19" - V.core 1.26v:
http://img65.imageshack.us/img65/346...8950vo5.th.jpg
Settings used:
V.core: 1.2950v on bios - 1.26v on Windows;
Cpu pll: 1.84v;
Qpi/Vtt: 1.26v;
Qpi Pll: 1.14v;
Ioh Core: 1.120v;
Ich I/C: 1.50V;
Ich Core: 1.14v;
Multi: x19;
BCLK: 216;
QPI Link: x36;
Vdim: 1.62v;
Loadline: on;
Uncore: x16;
Memory: 8x;
Turbo: OFF;
Pcie: 101;
Performance: Turbo.
Very good cpu!:up:
Update
3800Mhz @ 1.20v
200BCLK
1.235v QPi/VTT
http://img87.imageshack.us/img87/9391/3800120vev4.jpg
:up:
I will follow you on this one, my goal was 4.0Ghz 8 cores (4 threads for the WCG 24/7) and 4 Threads for the system. The temperatures reach 90c 8 threads priming but I doubt 8 threads will ever be working at full. So at the end 4 threads 24/7 will be at maximum 77c.
Maybe it is just me, but I notice that I am either rock stable or BSOD. There is no in between where Prime95 is failing or getting glitches. If I can run Prime, then it seems to run forever and everything is good. Just a small tweak, and bang BSOD.
dj883u2, what sort of cooling are you using? Are you on air or water there???
Download the latest Realtemp program and you are going to be very surprised.
Right on guys I finished my 24/7 crunch system:
Air Cooling True Copper lapped:
First lets see 4.0Ghz 8 threads ran for almost 20 minutes maximum 88c.
http://ssyybg.blu.livefilestore.com/...YQULyeDA/7.jpg
4Ghz 20 Minutes 8 threads =
Maximum: 88c
Average Maximum: 85c
http://ssyybg.blu.livefilestore.com/...201.168%20.jpg
3.7Ghz 8 threads 7 Hours Prime 95 =
Maximum: 73c
Average Maximum: 72
Load minimum and maximum average: 2 hours = 540 / 8 = 67.5
3.8Ghz should be a little higher.
4Ghz to 4.2Ghz means:
200Mhz x 8 cores = 1600Mhz which is converted into heat into heat.
So it looks like is a general problem for aircooling users, this chip is too hot. The only way to get 4.2 Ghz to work with 8 threads is with a good watercooling.
By the way 1 fan 1300RPM 54CFM for me is better than 2 fans as the latter was explained.
This was an interesting week,
Thank you all who participated and contributed to this thread. This is a very good thread explaining many things about this amazing processor.
Well I will keep at 3.717Mhz 1.168V 24/7 8 threads mainly because my goal was to keep 4 to 6 logical processors for the World Community Grid xtremesystems team.
I will only use 4.2Ghz if I need to bench for any reason or if a game does not run at 60 frames per second and the same needs a CPU boost or to play the best Playstation 2 emulator PCSX2 at amazing frame rates which is a CPU eater.
I spent all my money building this system which I do not regret at all however if I only knew the best aircooling solution would not stand for 4.2Ghz then I would have built a water cooling solution. Well there are things we can not go back in the past. It was done and was for good.
I will change to water cooling in the next few months and keep at 4.2 Ghz 8 threads. I still think it is possible to keep such speed with a good water cooling solution. There are good radiators out there in the market.
The CPU-Z Validation: 3.7Ghz 8 threads 1.168
1st results withh my new Dual Loop Core I7 Heat Tamer
http://i293.photobucket.com/albums/m...1/corei728.jpg
here's my 24/7 setup in progress
http://users.telenet.be/nils-marleen/i920prime4.jpg
Cooling is a TRUE with MX2 batch is
http://users.telenet.be/nils-marleen/PICT1252.JPG
It's not a superchip but not bad also i think....
The perfect gaming system: 4.2Ghz 4 Threads 90c maximum.
http://ssyybg.blu.livefilestore.com/...ghz%201.35.JPG
Real Temp 2.70 is the latest version i see on google. do you have a link to version 2.90?
Edit i found it. Not much differance in temp readings more even if anything.
http://img360.imageshack.us/img360/1740/48526068kp8.jpg
http://img360.imageshack.us/img360/4...jpg/1/w652.png
With "Turbo"on! :)
Prime 95 - 4011Mhz - 1.18v! No problem!:)
http://img201.imageshack.us/img201/7...5241dt4.th.jpg
:up:
lol why search google, when the official real temp thread is here? :ROTF:
Super Pi 1Mb - 4218Mhz - FSB a "222x19" - V.core 1.23v....:)
http://img178.imageshack.us/img178/7...2fsbae5.th.jpg
Validation:
CPU-Z Validator 3.0
:up:
Everybody uses Asus, Intel and Gigabyte Mobos but no DFI LANparty UT X58 why? Maybe too much settings? :D
I'm sure, some users like Tony or Eva2000 are testing it. (non public) :D
For some of you guys using higher clocks, what are your other voltages at? I am just playing with CPU vCore cause i am not sure what everything else is for. And I crash after about an hour.
raise your qpi/vtt a couple bumps as well as pll
Thanks.
Metroid and DJ88 what are your batchnumbers plz?
nilos go back like 4 pages they already posted them
Nice 24-7 for me looks like 3.7- 3.8 is perfect for air 24-7 without stock heat sink. I need better cooling load temps are to high for 24-7 @4g , my load temps were way to high so i did a nice prime at moderate 3.3 27 hrs I dont like seeing it that high of a load temp. Love these new chips:yepp: I especially love the 4 cores 8 threads. Im waiting for the next round of heat sinks :toast:
http://i137.photobucket.com/albums/q...rboenabled.jpg
http://i137.photobucket.com/albums/q...er1st/full.jpg
http://i137.photobucket.com/albums/q...t/Snap1-16.jpg
http://i137.photobucket.com/albums/q...pusinjulia.jpg
http://i137.photobucket.com/albums/q...photoworks.jpg
http://i137.photobucket.com/albums/q...1st/sisoft.jpg
http://i137.photobucket.com/albums/q...t/prime3-1.jpg
On the stock cooler of the 920, I was able to boot up windows and get a cpu-z validated shot of 4GHz. All I did was set the bclock to 211 * 19, the vcore to 1.325v, vpll to 1.9v, and the memory, qpi and uncore multipliers to lowest. Again this is all on stock cooling.
Is that at all impressive or noteworthy compared to others? It was not prime stable, the system froze shortly after starting prime. However I can get 3.8GHz prime stable, haven't tried 3.9ghz. For the third and final time, this is all stock cooling and 2min in the BIOS. Temps have hit 100C with this configuration.
Once DD gets me my new top for my MPC-X38 and D-Tek gets the mounting kit for the Fuzion V2, I will be putting the CPU and chipset under water. About what can I expect with water cooling based on what I've done on stock?
EDIT:
Forgot to mention, turbo mode is disabled for all of this.
Just Got my i7 920 running, here are the results
GA-x58 ud5
920 C0 rev, F4i bios
hyperthreading off
Patriot 999 DDR3, 3x2GB
Tuniq tower 120
Attachment 90352
I sort of cheated in that I benched with HT off, but that is no good for normal everyday use. Reason is Vista sleep mode doesn't work if HT is disabled. Struggling with this reminded me of the bad old days of XP with it's 90 second boot times. With vista, 10 sec is the norm, so HT on is a must.
No problem.:)
Prime 95 - 3705Mhz - V.core 1.088v:;)
http://img514.imageshack.us/img514/3...8263qb2.th.jpg
By:)
No luck for me using R2E with a Core i7 940. I already tried two CPUs and the highest rock solid stable I can get is 3.8GHz (19x200MHz) with 1.28V in BIOS, QPI/DRAM voltage is at 1.43750 in BIOS. Ram is Corsair Dominator 1600 @ 1600MHz with 7-8-7-22 timings at 1.65V in BIOS. Air cooling with CoolerMaster V8.
I do not know how everyone is managing 4GHz OCs even with 920 when I cannot even get any of the 940s to 4GHz. Maybe it's my cooling? Temps are maxing out at 77C (ambient is about 22C).
This is starting to suck badly...why pay for a 940 when it works no better than any 920 out there...
Any others users with R2E and a 920/940 on air to give some tips on breaking that damned 4GHz mark? These chips are supposed to be higher binned than 920s but maybe w/out WC or something better they cannot fully stretch their legs?
Any tips appreciated!
RealTemp beta link:
http://www.fileden.com/files/2008/3/...alTempBeta.zip
Save that link. It hasn't changed for months.
It's winter time here so there has been lots of time for updates recently. :)
Why would loosening up your ram timings help? I always thought that tighter ram timings were a good thing....
I've been having the same problem. I've managed to get (197 x 21) a lot more stable than the other ones. It seems that for me that 200+ BCLK isn't working, so I'm ok with 4.137 GHz at 1.344(1.3875 in bios)
I'll experiment a bit more and see if I can get anything stable at 200+ BCLK
Thanks,
Well to tell you the truth. I found out if you feed it with 13 to 16x then my system will boot with low voltage and it is more stable. I know it is strange. The memory has to be 6x and Uncore 13 to 16x. If the memory is x8 then the Uncore should be 17 to 20x.
I hope this info will help you guys out on the quest of getting a very good lower voltage with higher clocks.
No pics, I'm using the gravity method right now. That is, I set my case down flat on it's side and the HS is just sitting on the top of the 920 with some light bracing. Tomorrow, I'm going to make a proper hold down by screwing down a piece of sheet metal to each end of stock 'H' hold down to effectively enlage it so that it extends beyond the mobo holes. With holes drilled in the sheetmetal aligned to the holes in the mobo, I'll run the stock screws down into a custom backplate. This will allow me to turn the case up onto it's normal upright position...at least in theory.
Yeah, well, isn't max temp 100C? 77C is not that far off and with extra voltage that heat goes up pretty darn quick. Also, just cranking up the volts on Core i7 is not enough...all of that other QPI/Uncore junk goes into play as well, no to mention the mass of multipliers business, etc.
I do always loosen up the timings when trying to OC to factor out the ram. Thanks for the tip though :)
Looking for more now!
Are you on air or water?