My X3350 has arrived and i will be putting it in my rig right now and i will report back with results...:D
Printable View
My X3350 has arrived and i will be putting it in my rig right now and i will report back with results...:D
Very nice, looking forward to seeing results :)
Mine is coming tomorrow! :)
Can't wait to see your results ;)
Nice to hear more people are jumping on the train. :) Here's hoping someone can finally boot into Windows at 4GHz. I can only POST at 4GHz.
Mine should be coming soon as well.
Runawayprisoner, you should start a new thread, or post to this one, with your results and experiences with the x3350. Seeing as there isn't much info about the chip right yet, any personal experiences would be appreciated :P
mine arrives tomorrow.
where r u guys buying them from ? moogr ? lagoom ?
is moogr legit ?
Eagerly awaiting your results as I have one on the way was well. A lot of later ES were hitting 480MHz FSB, so hopefully they have the retail cores tweaked even more.
Well i got it running, I think i may have a bad stick of memory... but otherwise everything runs great. CPU-Z detected it wonderfully and now its a matter of figuring out the C1 error i get with one stick and not the other.
This chip works great though!
We will find out soon... im not at my PC but i know it works. The key is to figure out my OC tomorrow.
I picked mine up from lagoom. They seemed to have the best price, but I think they've raised it.
Edit: They have, but it also seems they're out of stock. Here's another place, but more expensive.
http://www.excaliberpc.com/Intel_BX8...id-585292.html
http://forums.ngemu.com/hardware-dis...s-warning.html
^= All you need to know, guys. :) I can run mine at 475 x 8, and has been running at that clock for four days now. VCore is only 1.320V for a rock solid Prime stable system at 475 x 8. I can POST at 500 x 8, but it would hang as soon as Windows starts loading... so I'm guessing my G35 motherboard cannot handle 500MHz for a quad 45nm. Now I'm curious if anyone with a X38 or X48 mainboard can get into Windows at 500MHz. :)
Here's how I got it to POST at 500MHz x 8, in case you can't read much from that forum post I linked above.
http://youtube.com/watch?v=QzF4Vg6_R9Y
And... here are some results from Crysis Benchmark Tool:
DX9:
http://i235.photobucket.com/albums/e...isTimeDemo.jpg
DX10:
http://i235.photobucket.com/albums/e...sBenchDX10.jpg
Not too shabby... but I guess these quads are just what Crysis has been dreaming about all this time. ;)
Edit: Or heck, here are all I have at the moment:
Idle temp before I lapped my ACF7:
http://i235.photobucket.com/albums/e...C/quadidle.jpg
What VCore is needed to get into Vista @ 3.60GHz:
http://i235.photobucket.com/albums/e.../x3350cpuz.jpg
VCore needed to get into Vista (and watch a movie) @ 3.80GHz:
http://i235.photobucket.com/albums/e...am/PC/CPUZ.jpg
And yep, I don't think this thing has any problem running two instances of the PlayStation 2 emulator:
http://i235.photobucket.com/albums/e.../dualPCSX2.jpg
Impressive temps at that, too... considering I haven't lapped my ACF7 yet.
Now here's 3DMark06 with a 9600GT:
http://i235.photobucket.com/albums/e...DMarkX3350.jpg
Voltage for Prime completely stable at 3.60GHz:
http://i235.photobucket.com/albums/e...rime95SFFT.jpg
And... that's it for now. I think I'll leave her on tonight for either Prime at 3.80GHz or 100 times Crysis benchmark.
Mmm tasty, 3.8Ghz at ~1.3v. That's a hell of a lot better than my Q6600 B3's 3.5Ghz at ~1.5v :p:
It'll be good to see some more reports on them, but providing most of the X3350/Q9450's can hit 475FSB fairly easily, it's definitely a winner over a Q6600.
Eller
Well, thing is... I can't go any further than 475 and stay stable for long. At this point, I think it's the motherboard that is holding the chip back. Watch the video and you'll see that it's completely possible to POST at 500 but it won't go into Windows no matter how high the VCore is set.
Edit: Oh, and I lapped my chip today, too. Idle temp dropped by 10C and load temp dropped by 5C after lapping both the chip and the ACF7. When I lapped just the ACF7, idle temp dropped by 5C and load temp did not drop at all. And beware, at 1.35V, this chip can really heat up to 80C (with Prime95 small FFTs) when you don't monitor her.
Thanks for all the info. Looks like you have a great chip there. I think most would be happy with what you've gotten. There seems to be quite a few people with these on order, so it wont be too long for the mb questions to be answered. Mine will be here on Wed and I have an X38 mb.
Awesome chip!! However after being spoiled with a QX9650 @ 4.0 I am going to wait to see if more of them 3350/9450 hit 4.0 on high end water before I consider one. :)
Cheapest available X3350 is still $390 in Canada... garrr
i wonder how well one of these will do on a P35 mobo on air.. i doubt 4 ghz
I'm currently on air, and it's 3.80GHz stable. Temps can reach 70C sometimes, though... and definitely don't try Prime or stress testing too much at 3.80GHz without TRUE or some high-end air cooling, lest you kill your chip.
PCSX2 0.9.5 svn 329. Look it up in Google. :)
The processor runs faster with only one instance, so I'm guessing one or two of the same cores are still being used by the emulator.
And by the way, Crysis really DOES love quad-core processors...
http://www.betterit.com.au/scripts/p...product=106225
x3220
http://www.betterit.com.au/scripts/p...product=106211
x3350 - showed "19 available" over the weekend...now "NO ETA"
Wow... these are going so fast. I wish I didn't lap my chip now... but hey, it's nice enough, so I guess I'll keep it. :)
I just got to 3200ghz easily and now im gonna test my 3dmark score a bit...
I got 11,600 with my videocard at stock speeds. A 4000 point improvement
I kept my old card so i could compare the difference...
Old- http://service.futuremark.com/compare?3dm06=2185229
New- http://service.futuremark.com/compare?3dm06=2185229
Whats LOLTASTIC? 3dmark thinks my CPU is a P3
MOre pics to come...
Mmm, I think you're giving the wrong link for the new 3DMark score? And nice score, by the way. ;) Try going to 450MHz FSB now? I know that's reachable, and it's much much cooler than 475MHz, too.
Actually, it's only completely stable (Prime95) at 1.328V to 1.344V (1.335V and 1.350V in BIOS respectively). At any other settings, it might BSOD or crash every hour or so if I'm doing something intensive. But I'm happy nontheless, I've seen initial retail Q9450 results that need 1.350V (1.375V in BIOS) or even more to be stable for 3DMark06 only. And I think none of them could even POST at 500 x 8 so far.
I could POST at 520 x 8... but you can guess how long that will stay before it freezes. :) Will take a picture next time if I have a better camera.
bookmarked! hoping to see more results
Russion,
Thats great but what board are you using and what bios version?
Mine is going to arrive hopefully in the next 7 to 10 hours from the time of this post :D .... and then it is going directly into a server build... BUT before that happens I need to test it out and fire up some benchmarks :yepp: :yepp:
So what board should I use??
evga 680i ???
evga 780i ???
evga 790i ???
or some other board like an X48 or something... ??
Definitely X48, and please, please, please... PLEASE get it to 4GHz beyond the POST screen. ;) I know such a thing is possible.
I swear... if I only I could get to that Validate tab in CPU-Z...
And yeah, I was lucky! I almost got into Windows XP at 500 x 8! That's after cranking all kinds of voltage options I can find in the BIOS to their max settings and putting a whole bunch of fans right beside the CPU HSF. The system froze right at the Welcome screen this time. Bummer...
From my own experience, this chip seems to put a lot of strain on the motherboard even at 475 (Windows Vista refuses to even start until NB volt was at 1.45V, to get Vista to cooperate nicely with no sudden freeze, NB volt needs to be at least at 1.51V. I set this to 1.55V just to be on the safe side, and just because I already put an aftermarket cooler on my NB). At 500, my board couldn't afford to stay stable for long... but I would bet ANYTHING that a stronger board with higher voltage settings (at least higher than 1.80 PLL, 1.50 FSB Term, and 1.71 NB) can at least get into Windows at 500 x 8.
Edit: In case you have any doubts, I can already POST at 500 x 8. And it's also possible to POST at 520 x 8 with those settings. Or actually, I'm going to see what the max stable FSB my board can handle... right now.
you got a dual boot there ... XP needed for PS2 emu ?
XP is needed for stability testing and such... since it's easier to boot into XP than Vista.
K will grab the Emu then , tx for the info... I always enjoyed the Mame Emulators... hope these are a blast too
Okay. :) And for the non-believers:
480MHz limit? :p
http://i235.photobucket.com/albums/e...C/X3350486.jpg
Hmm... 490MHz is also bootable! :)
http://i235.photobucket.com/albums/e...C/X3350490.jpg
And ultimately... the closest I can ever get:
http://valid.x86-secret.com/show_oc.php?id=334848
So... I'm missing 5MHz on a G35 board. X38 and X48 boards are just going to let that 5MHz go? ;)
Edit: And... I think it's possible to stay at 3.96GHz 24/7 on water cooling, since it was idling at 45C on air (low-end). I'm getting a Thermalright SI-128 SE and a really good 120mm fan this weekend, and I'll see if that makes a difference. Oh, and VCore for 3.96GHz is 1.42V.
ES but 500FSB
We need a method to unlock these woeful multipliers :wasntme:
3.80GHz is a keeper IMO. The difference in performance between 3.80GHz and 3.96GHz (it's stable at 1.46V) is tolerable. 3.96GHz requires much higher voltage for both the CPU and the board as well.
And from the 500FSB result link, I think... you can't stay stable even at 3.60GHz for Q9300? And by the way, if I drop the multiplier down to 7.5x, I can go beyond 500FSB and stay stable with ease. What I want to see is 500FSB at 8x. ;)
I caved and ordered two from moogr
http://www.moogr.com/INTEL_CPU_XEON_...id-368325.html
Use the coupon marchmadness
Found the deal here.
Never mind... you guys ready for this? Here you go, 500MHz at LAST! :up:
http://valid.x86-secret.com/show_oc.php?id=334952
I think that's all I'll do at this point... it's time for good ol' X3350 to take her break. I have been pushing her quite hard the last few days. :p
And by the way, batch #: L803B196
Now I'm even more positive that a X38 or X48 motherboard would be able to go beyond 500MHz, or at least be able to do more at 4.00GHz than my board... probably... hmm... they can even keep this FSB 24/7 stable! Oh yeah, and voltages needed for 4.00GHz:
VCore: 1.500V BIOS, 1.492V CPU-Z
PLL: 1.80V
VDimm: 2.30V (just to be on the safe side)
FSB Term: 1.50V
NB: 1.71V
SB: 1.20V
SuperPI 1M at 4.00GHz is 12.500s.
Got it now lets see what she can do.
Congrats! And let's see some results with a true overclocker's board! :)
I'd watch that vCore if I was you. The 45nm CPUs (at least the Wolfdales, and I doubt Yorkfields are much different) have a habit of dying if given much more than about 1.45v actual for 24/7 use. I personally wouldn't want to be using more than about 1.4v actual with a 24/7 rig.
Eller
Guys, anymore places selling these CPUs?
3.6GHz: 1.22V
3.8GHz: 1.32V
Thanks for the reminder. Someone said the same thing to me when I bumped my E8400 to 4.65GHz the other day. I'm just posting a validate for 4.00GHz to show that there's really... no such thing as a 480FSB wall for this CPU. Of course it's not a matter of VCore, since I can also boot into Windows at 4.00GHz with 1.35V (just found out). The problem is that the strain on the motherboard at 500FSB is too great for it to maintain stability over long-term use, so 4.00GHz is really unrealistic, and not just for the vCore it requires.Quote:
Originally Posted by karl_eller
I'm currently running 3.80GHz still, and it's stable enough at 1.32V. There's very little difference between 3.80GHz and 3.60GHz, though. Probably 1.1 seconds in SuperPI 1M and that's it.
P.S.: And this chip is showing weird signs... of voltage requirement. It required 1.344V to prime stable for over 4 hours just yesterday. Now if it's degraded, it should require more, right? Today, it's requiring only 1.320V to prime stable for 7 hours straight. Slapping a Big Typhoon over it cooled it down drastically, too, and now I'm hovering at 20C-55C in normal office works and gaming sessions. Also to note that 475FSB no longer required me to boost the PLL and FSB termination voltages. I can leave them at stock now, and instead, just bump the northbridge voltage to 1.55V and it's rock solid. I have a Thermaltake Extreme Spirit II on my northbridge chip so I'm not afraid of temperature... but... that's just plain weird. Just yesterday, it required 1.64V PLL and 1.40V FSB Term to stay stable at 3.80GHz!
And... 480 FSB wall is busted. :p
http://i235.photobucket.com/albums/e...C/MaxBench.jpg
That's 3.92GHz stable at 1.44v... so if I'm not mistaken, it'd take 1.52v or 1.54v for 4.00GHz stable. vCore for frequencies are as follow:
Stock stable: 1.14v
3.00GHz stable: 1.16v
3.20GHz stable: 1.18v
3.40GHz stable: 1.20v
3.60GHz stable: 1.23v
3.80GHz stable: 1.33v
3.90GHz stable: 1.42v
4.00GHz stable: ???v (never tried)
That's how it is for my chip so far. :)
That's a killer chip you have there. 490 x 8 at 1.44..... when you say stable, you mean prime stable right?
Slight bump... and... max 3DMark06 with 9600GT:
http://i235.photobucket.com/albums/e...Mark0696gt.jpg
Seems to be around or above 8800GT scores now. :)
That looks great! :)
These ones are so much fun, I got mine today, and started easy with
8x400, all stock and its nicely stable so far, and I'm on the stock cooling (yeah will have to get something else later) But for now I'm quite happy since I'm
upgrading from a X2, and working with mainly 3D I'm more than happy :)
Motherboard used so far: ASUS P5E-VM HDMI. It's G35, by the way. :) This little guy can really overclock something when pushed to his maximum limits, eh?
Man, I took the day off work, Im just sitting here waiting for my X3350 to show up, along with my X48-T2R. COMEON FEDEX Man!
Ahahahah, I know how it feels like. And I got really frustrated last Friday when the delivery was attempted at... 8AM, and I wasn't even awake then! I had to run down to UPS at 6PM to get the darn package. It's all going well now. I'm currently waiting for more Arctic Silver 5 and a pretty cool aftermarket cooler to come. Hopefully, I can stay longer at 4.00GHz to... maybe do a quick 3DMark06 bench? :)
Edit: Here are the temperatures when the processor is idling at 3.80GHz... and when she's being stressed by Prime95 large FFTs, me playing some Youtube videos, loading these websites, chatting on Yahoo, and MSN, and probably... uploading something via Torrent all at once:
http://i235.photobucket.com/albums/e...am/PC/load.jpg
Idle: 26/26/17/20
Load: 64/61/59/55
All on AIR with a Big Typhoon.
This is just so you know what you can expect from your chip. :) Ambient temp here is current 23C.
Yes. G35 doing quad at 490MHz FSB... STABLE~! :up:
I even posted a validate link for 500MHz FSB, too... too bad it wasn't that stable. :p
would a X48 DFI be worth getting over a DK X38?
someone message me there opinion please
AS5, I bought some awhile ago as well..Idk just seems good lol
Nothing wrong with AS5, been around for years. The reviews usually show MX-2 to be a little better, but AS5 is still pretty good.
Anyway, looks like 3 803b196's. I hope mines is as good as yours.
Uhh... just received new cooler. Testing just the fan out, and... results are incredible! E8400 @ 4GHz (500 x 8) @ 1.350v stressing with Prime95 small FFTs... barely 45C max! Idling at 20C, and that was just the fan! I wonder what the heatsink can do...
Edit: small review of the Noctua NF-P12 fan on Big Typhoon:
http://i235.photobucket.com/albums/e...am/PC/temp.jpg
That's it, folks. That's the review. :up:
Got mine in and running P95 @ 3.6GHz with 1.3vcore set in bios (1.27 in CPUz), all other voltages at minimum. Once my DFI board gets here I am going to slap it in there are see how far I can push it.
Yes...running P95 for over an hour so far. We'll see how far it goes (after 8 hours I stop). So far this is a great CPU. It replaced a Xeon X3210 and this seems so much snappier in apps and even when windows loads. If I can get this stable on the DFI board at 3.8GHz I will be very happy.
Not the X48. Couldn't justify it so I ordered the LP LT P35. I already got one from motherboardpro but had to send it back. it was the oddest thing...I was OCing my E3110 with it and all of a sudden it started failing P95 in like 10seconds. Even after reverting back to all stock settings it would not even run SPi all the way through. I swapped out every piece of hardware (CPU, memory, HD, PSU, GPU, etc) and applied every trick I have learned about DFI boards over the years (memory in all slots, over night bios clear, etc.) and nothing worked. I even tried to reinstall the OS but the system would stay stable long enough to install Windows. So I ordered another one from the egg and sent the bad one back to DFI for RMA and a refund.
http://www.xtremesystems.org/forums/...d.php?t=181995
since alot of people here have this chip maybe u guys can help me out
but no 2.0 PCI-E
As far as I know there isn't any cards that max out a PCIe x16 v1.1 slot, so aside from the extra power circuitry in v2.0 (and since v1.1 is still widely used, cards have a PCie power plug anyway), there isn't a need for it at the moment.
And by the time cards come out that DO need PCIe v2.0, most of us will have upgraded to Nehalem boards anyway :p:
Eller
well I can understand that, but I don't want to upgrade anything but my gpu after this building point..thats why im looking at getting 2.0
anyone overclock the X3350 on a X48 DFI before?
I'm pretty sure the multis are laser locked.
will 1 x pci-e 2.0 be a bottleneck with g100/r700/larrabee cards later this/next year?:shrug:
actually meant to say will 1x pci-e 1.0 be bottleneck in the not too distant future?
x48 is only a tad more expensive than x38 dfi...but what do you actually get for that? 10 more fsb on quads?:hehe: or not much at all...
you would think it would be better on the x48 thats what im hoping for
If you're not sure, you can try getting one of the boards that have had great results with the chip instead. Some are going for quite cheap, too.
Well, so far it doesn't look like I have a super chip. Just played with it for a few minutes and got 3.6 at 1.34v. Ran Othos for about 10min and then shut it down. I hate running that stuff, lol. Of course it could be the ram as it has trouble around 450.
Youch... so you're telling me I'm holding the world record at the moment? Come on now...? :p: Hmm... and is it 1.34v read by CPU-Z, or 1.34v set in BIOS? I have to set mine to 1.3625v in BIOS now because... I forgot to pencil my motherboard. I think I'm going to pencil it again now.
BUMP... and for god's sake... why is it that the voltage requirement is going down so much? Now it requires only 1.312v to stay stable... huh?
http://i235.photobucket.com/albums/e.../PrimeTemp.jpg
And thanks to new cooler, small FFTs can not even touch 70C anymore. They reached 77C before. :up:
Edit: Oh, and real temperature is actually 3-4C cooler than what's being reported by Real Temp. I forgot to tick the "--" option in Idle Temperature Calibration.
You're running an old version of Prime95, use Orthos, it runs 2 instances per exe. Set Priority to 5, and run custom FFT of 10K IN PLACE. That's the hardest cpu test. People who pass 8K FFT stable for hours fail 10K in minutes.