going to try my hand at a Xeon

[splat]

Well, since I overlooked the area in the manual which explained why my cpu/ram pair did not work in my new Gigabyte 965p-ds3, I sent it back to Newegg for a replacement. I have been planning all week to buy a new Xeon 3050 w/ 2gb Geil DDR2-800 ram to put in that replacement DS3, but Newegg said today that they don't have them in stock. So they refunded my money and now I need to pair up a new motherboard with my cpu and ram. I know there are a lot of people who will post "get the asus p5b", but I want to be a little different, and after reading Before's nice post with his 500+ fsb overclock on a Biostar (yes i see he had several mods on the board), I think I'm going to follow his footsteps and buy that board, even tho I'm only going to be using air. I gotta say, I like the price. So, I guess this post is mostly foreshadowing for what will be delivered in a few days.

I'll post all my results here.

[splat]

the Xeon came in last night. Got it burning in now. I'd post a screenshot but for some reason the nic on the TForce965PT isn't working...gotta figure that out...

[splat]

just a quick benchmark I noticed, using the benchmark built in to BOINC -

my powerbook g4 867mhz:
435 floating point MIPS (Whetstone) per CPU
1354 integer MIPS (Dhrystone) per CPU

my pentium d 805 at 3.7, windows xp:
Measured floating point speed 1777.1 million ops/sec
Measured integer speed 1232.36 million ops/sec

pent d 805 at 3.7, linux 32bit:
Measured floating point speed 1117.04 million ops/sec
Measured integer speed 2872.6 million ops/sec

Athlon 64 4000+ stock, linux 64bit:
Measured floating point speed 2164.37 million ops/sec
Measured integer speed 4675.98 million ops/sec

Xeon 3050 stock, windows xp:
Measured floating point speed 2003.05 million ops/sec
Measured integer speed 4166.5 million ops/sec

*updated the numbers

[||Console||]

just a quick benchmark I noticed, using the benchmark built in to BOINC -

my pentium d 805 at 3.8, windows xp:
Measured floating point speed 1777.1 million ops/sec
Measured integer speed 1232.36 million ops/sec

pent d 805 at 3.8, linux 32bit:
Measured floating point speed 1117.04 million ops/sec
Measured integer speed 2872.6 million ops/sec

Athlon 64 4000+ stock, linux 64bit:
Measured floating point speed 2164.37 million ops/sec
Measured integer speed 4675.98 million ops/sec

Xeon 3050 stock, windows xp (not exact numbers, i'll update them when I get home today):
Measured floating point speed 2000 million ops/sec
Measured integer speed 4000 million ops/sec

So the single core 4000+ is faster than the DC new arch Xeon ?

[splat]

it appears BOINC thinks so. Once I get home today I'm going to run thru some 3dmark tests, super pi, and others at stock to get a baseline, then see if I can find my max OC on stock volts, then see how far I can really go. I'm also going to boot in to linux and do some tests there, tho I won't have anything to compare it to really other than itself stock vs. OCed.

[MikeDVB]

What's the stock voltage/vid on your 3050?

Mine is 1.2250 and I wanted to see if this was what it was default or if I got lucky... On the box it says max 1.35v.

Mike

[shimmishim]

that's a very low vid Mike!

mine is 1.3125!!! :O :O

i have mine testing right now at 3.7 ghz

http://www.johnshim.com/forums/anandtech/8x463smalltesting.1hour.xeon3050.jpg

[(Clanger)(DOA)]

What's the stock voltage/vid on your 3050?

Mine is 1.2250 and I wanted to see if this was what it was default or if I got lucky... On the box it says max 1.35v.

Mike


that is indeed a nice chip.

[splat]

my stock voltage is 1.235v.

batch #: L626B153 Pack Date: 11/02/06

now for some runs at stock to get a baseline:
http://img244.imageshack.us/img244/7200/3050stockgs5.png

3dmark 06
http://img244.imageshack.us/img244/6267/3050stock3dmark06zx2.png

3dmark 05
http://img244.imageshack.us/img244/8522/3050stock3dmark05gg6.png

Super Pi 1m time:
http://img92.imageshack.us/img92/3157/3050stockspi1moa6.png

and my max clock at stock volts 32m super pi stable 293mhz, shows as 291. (found using the biostar supplied T-Utility):
http://img244.imageshack.us/img244/2884/3050stockmax32myr2.png

[splat]

well, I think my power supply is not pulling it's weight now. I'm usiing a Antec Neo480 i had lying around from my A64 build. I couldn't boot in to windows at 350mhz fsb. I got the voltage all the way up to 1.42v but decided that something had to be wrong. I bumped he mch voltage up to 1.35, then 1.45, stll wouldn't go in to windows.

So I guess it's time for a new psu. :(

[MikeDVB]

my stock voltage is 1.235v.

batch #: L626B153 Pack Date: 11/02/06



Mine is the same exact Batch and Pack Date.

http://www.mikedvb.com/CPUzCoreTemp.jpg

Mike

[splat]

Mine is the same exact Batch and Pack Date.

http://www.mikedvb.com/CPUzCoreTemp.jpg

Mike

thats nice. should give you some more headroom as long as you have good enough cooling.

I ordered an Enhance E5150GH powersupply, so hopefully I'll be able to start overclocking again later this week. For now, I don't want to risk using that Antec again since I think one of the capacitors is bad, and other will probably go soon too.

[inkfx]

I was about to say, those Xeons have pretty low voltages which should make for some nice overclocking. I havn't read much on them so I'll definitely be checking in for your results.

[MikeDVB]

I can't seem to actually test the memory.

I'm new to this overclocking stuff, and I dropped the multi to 6 (as low as it would go) and brought the memory to 1:1, The memory I am 99% sure should be able to handle a 400mhz FSB at 1:1 because it runs at 408mhz normally on 2:3. The processor is only at 2400mhz which should be no sweat for this xeon but I cannot get it to post at 400mhz FSB. I get some "Award BiosBlock" when I try to boot it up 6x multi 2400mhz processor 400mhz FSB (ram at DDR800). And then it gives me a "Bios Checksum Error" and I have to clear cmos to get it to boot up.

In the event that my memory were the weakest link I'm looking into PC2 8000 DDR2 ram (1000mhz) which will give me PLENTY of headroom and for sure have some good chips. Was looking at the Crucial Ballistix:
http://www.newegg.com/Product/Product.asp?Item=N82E16820146563

If the memory's not the weakest link ... do I need to up voltage on something else or is my mobo just not going to let me do 400mhz FSB?

Anybody have any tips or advice I'd appreciate it.

Mike

[batboy]

I answered your PM. You probably need more chipset voltage.

[MikeDVB]

I answered your PM. You probably need more chipset voltage.

Yeah this is the thing, I don't know what voltages need bumped, what should be fine, what to test and how to test it, I've not got the experience. If you had Yahoo or MSN or AIM I could talk to you on my laptop in realtime as I fooled around with it !!!!

Oh well guess I learn the hard way =)

Mike

p.s./edit If anybody has a messenger and has a good bit of experience overclocking an intel chipset/mobo/processor shoot me a message on any of my 3 contacts to the left under my Join Date & Location =)

[splat]

Well, the processor itself is rated up to 1.35v. You should be pretty safe as long as you keep the vcore below that value. most likely, you will only need to go above that value when you raise the multi back up to 8 and start testing the processor. You can raise the Vcore slowly, i usually go one step then reboot and test.

[batboy]

To do 400 FSB, you will probably have to raise the MCH and ICHIO voltages up a little. Try one bump at a time, like 1.3v for the MCH and 1.55 for the ICHIO.

Default MCH for the AB9 = 1.25v, default ICHIO = 1.5v. Try 1.3v and 1.55v and see if that helps. Thought I was pretty clear.

[MikeDVB]

Edit: I just purchased a set of Crucial Ballistix DDR2 800mhz 2x1gig 4-4-4-12 at 2.2v sticks. They *should* handle overclocking (or at least running 400mhz at 1:1) much better I hope, we'll see.

Default MCH for the AB9 = 1.25v, default ICHIO = 1.5v. Try 1.3v and 1.55v and see if that helps. Thought I was pretty clear.

No I understood you... what I mean is I don't know what it is that you know, and that you are helping =)

Ok I did a lot of testing to see what I could get to boot and not and I got several higher FSBs to boot up to 400mhz (where I stopped for testing ram)

Here are my results... I think my ram IS going to hold me back but I'd like your guys' inputs before I make any decisions. I have about 15 more days to Return this ram and I have up to $350ish to spend on HIGH QUALITY ram if need be (1000mhz ram for example)

400mhz FSB
2.2v Ram
1.25v MCH
1.5v ICHIO
WOULD NOT POST! (2.2v Ram will not post at 400, 2.3v Ram will)

400mhz FSB
2.3v Ram
1.25v MCH
1.5v ICHIO
I could get this to boot but when I got into memtest I got a lot of errors before even leaving Test 1.

400mhz FSB
2.3v Ram
1.30v MCH
1.55v ICHIO
I got to test 3 before errors but then they came flooding in.

395mhz FSB
2.3v Ram
1.25v MCH
1.5 ICHIO
I could get this to boot and memtest got 16 errors in Test 4 (made 1~3 no problems)

I don't really want to go any higher than 2.3v on the Ram and I tried bumping MCH, ICHIO and even the processor up and booting and none of it helped me pass memtest. I did try 2.4v on all of the above mentioned setups and it didn't help me pass memtest. I *think* my memory is holding me back - I can get the computer to post at higher than 400mhz but I cannot get it to pass memtest. If you have any advice on what I can try to get Memtest to pass at 400mhz then shoot it over.

Otherwise if you think my Memory is crap let me know and let me know what you suggest I get to get 450~500fsb (450x8 = 3.6ghz 900mhz ram) etc. I was thinking about going ahead and getting some Crucial Ballistix 1000mhz Ram (which I am pretty sure would eliminate my memory overclocking problems)

Mike

[greenmaji]

Mike.. start a new thread ;)
that and post what bios functions you have disabled in that post :)

Let us know when that enhance PSU is suppose to show up would ya splat :D

[batboy]

He already has threads going in several sections. Looks like increasing MCH and ICHIO voltages helped a little. Try 1.35v and 1.6v now. Bumping up vcore another notch might help too. What are your default timings and what are they currently set at? That makes a big difference too.

[splat]

Let us know when that enhance PSU is suppose to show up would ya splat :D

oh, you'll know. Believe me. I would assume the earliest it'll be here will be Wednesday, but I would bet Friday...which is a lot of down time...

[MikeDVB]

Mike.. start a new thread ;)
that and post what bios functions you have disabled in that post :)


Yeah I responded here and then I decided to make another thread but I couldn't delete the post here so I just updated it.

Mike

[leojharris]

i've got an old 500mhz dual zeon system with a SCSI LVD raid0 array ... i *still* love that computer!

bought it around 1999, paid around $8000 dollars for it. the sustained disk throughput of 40gig a second was unbelievable at the time.

so funny. it's currently on the shelf, but i might fire it up and do some benches on it just for kicks and comparisons ...

good luck with your new chip ... looks good ...

[splat]

Finally got my new psu. But, it's not going so smooth. I can't really boot at 350mhz fsb. I have my ram set to 2.1v and 1:1, raised the mch all the way to 1.55 (max) and raised this fsb limit voltage to 1.5 which is max in the bios too. I raised the vcore all the way to 1.5v but I still could not stay in windows for more than a few seconds.

So, i'm sitting in windows at 335mhz right now, 1.35v, and I just got an error running SuperPi 1m.

Temps at idle right now are 36, 35 in coretemp (53c, 53c load), and when i raised the proc up to 350mhz, 1.5v, I was booting in to windows at around 45-50c.

So, I don't know whats up. Any thoughts? oh, and I'm on the latest stock bios from biostar.

[greenmaji]

2.6Ghz? if so, lower the Vcore to something like 1.25.

Test the ram frequecney, timings and Vdimm with test 5 in memtest.

[splat]

i was at 1.25 and i booted in and started using clockgen to raise the fsb. I got up to 345 and was able to get thru superpi 1m, but at 350 it rebooted, so i started raising the voltages and could not get windows to stay stable for more than a few seconds. I'll run thru a memtest, but at 700mhz (rated 800mhz), I don't think ram is the problem. I've got the ram set to spd, so it was loading at 5-4-4-10 @ 2.1v, 700mhz. (spd for 800mhz is 5-5-5-15)

[greenmaji]

Go strait for 401FSB..
Stock settings on your ram.

[splat]

thanks, that worked. Odd how my ram/mobo didn't like running in the mid 3's. I'm passed 425 now, so it looks like things are going well.

[greenmaji]

Typical 965P behavior.. it has some rough spots in the mid to mid-high 300's FSB.

[leojharris]

Typical 965P behavior.. it has some rough spots in the mid to mid-high 300's FSB.

good job, mang!

[splat]

so i'm trying to get Prime stable at 446mhz fsb right now. This is the highest my ram (Geil Ultra) will run at 2.2v which is the max Vdimm I can set in the bios. I think the processor still has a good amount of kick left in it, and I'd bet with the right memory I could hit 500mhz fsb with it. I tried using a couple Rebels Haven bioses that Before linked me to, but I couldn't get any to load in the flashing utility that is built in to the current bios. Tonight I might try them out with the award flashing utility that is on the Rebels Haven page.

I can't remember specifically what Vcore i have set right now, it's somewhere around 1.33v. I was able to get thru Super Pi 1m and 32m at 1.30v @ 446mhz (8x446 = 3568mhz), and I'm hitting 53c load on both cores, so I think I've still got some room left to bump the voltage some more (using Arctic Cooling Freezer 7 Pro).

Also, I've got FSB Termination Voltage set to 1.2v (stock), and MCH Voltage set to 1.35v (1 bump up from stock 1.25v). Hopefully I'll have some prime stable pics for you guys tonight or tomorrow.

[deathman20]

Damn thats nice man.... 3.5Ghz on 1.3ish Volts wow.

I'll take that chip if you have a limited ram speed :) Trade ya

[splat]

if anything, i might sell this ram and buy some better sticks. I'd really like to get some more Vdimm in the bios, I'm hoping I can figure out a way to flash to the same bios Before was using. I think I remember seeing an article somewhere (probably anandtech) that they took these Geil sticks up to around 500mhz but on 2.4v...i gotta track down that article and figure out what those exact numbers were.

*edit: ok, found the article. It was Anandtech, but they were using Geil Ultra sticks with Micron chips (makes sense). They got 515mhz @ 2.4v, 4-4-4-12. Their sticks are GX22GB6400PDC, while mine are GX22GB6400UDC. Still, if I can get the options in the bios, I think I can get above 450mhz with 2.3v or 2.4v (the memory is rated at 1.8-2.1, and those Ultra Plus sticks are rated up to 2.3v)

*edit again: just noticed on the geil website that they state a working range of 1.8v~2.3v, so that gives me hope.

[greenmaji]

yep.. link below :p

[Gautam]

I think gm might be confused?

http://www.xtremesystems.org/forums/showpost.php?p=1631098&postcount=29

Have you tried 5-5-5 btw?

[splat]

Yeah, i've actually been reading that XS thread all afternoon, as well as the Rebels Haven forum. Both have some good info in them. I'm going to have to find a 10k ohm VR and give that a go.

And yes, I've been running at 5-5-5-15-auto set in the bios, i get errors in Memtest86+ at anything over 446mhz. But I also saw that the mobo can some times set some odd timings with that last auto setting, so I'll have to explore that in windows tonight.

[greenmaji]

What mod do you need a 10K VR? The Vdimm mod says 50K VR.

[splat]

on the rebels haven page, the guy calculated it out to be a 10k-20k VR...and if you look farther in to the pages of that XS thread, one guy changed from a 50k to a 10k and said the adjustments were better.

anyway, on to the chip:
http://valid.x86-secret.com/show_oc?id=160170

http://img99.imageshack.us/img99/2259/xeoncpuz446mhz4ln.png
Thats the max my ram seems to want to do.

http://img408.imageshack.us/img408/1984/xeon1m446mhz4bh.png

but I can't get orthos stable at that fsb. I got the Vcore up to 1.42v, but my temps in CoreTemp started touching 65c stressed, so my AC Freezer 7 Pro isn't able to handle the task.

[deathman20]

Hand the chip over ;) I'll clock it :)

Will orthos run at say 3.45Ghz or something with the 1.3 or 1.4V?

[splat]

i had to drop the vcore and fsb to get the temps down. I'm at 425 @ 1.375v right now, 1 hour in to orthos. temps touch 60c every once in a while, mostly run at 58 or 59.

To be honest, 425mhz was my initial goal before i got everything ordered, so if i can get it orthos stable at 425 then i'll be happy.

[greenmaji]

Yea.. looks like you found the limit of the AC freezer pro :-/ not a big fan myself (~$10-15 more gets you a top performer)

[splat]

funny thing is, it kept my 805 D running at 3.7ghz at under 40c under load. I thought these C2D's were supposed to run cooler. But yes, the AC Freezer 7 Pro is definitely not up to handle the heat.

When i left for work this morning, i was just passed 9 hours in orthos running at 425mhz, 1.375v. I've left taskmanager open so I can look for throttling, and I've noticed that every once in a while I will see a tiny blip in the cpu chart, but it will only be one core, not both. Is it normal for small, infrequent spikes? or is that a omen for problems?

*edit: btw, the SPD for my ram at 400mhz is 5-5-5-15. I wonder if I drop the timings to cas6 if i might be able to clock the sticks higher than 446mhz.

[greenmaji]

you don't know what a netburst chips core temperature is ;)

yes.. its fine to test other ram settings :)

[deathman20]

Its possible you just got a hotter C2D.

I've had 2 versions.. Ones that run cool, and ones that run about 10C hotter. Its possible.

[splat]

i gotta let the as5 cure too.

thats true, coretemp is reading a different temperature than other chips, so is 65c in coretemp the one you want to stay below, or is it safe to go higher?

[deathman20]

i gotta let the as5 cure too.

thats true, coretemp is reading a different temperature than other chips, so is 65c in coretemp the one you want to stay below, or is it safe to go higher?

Coretemp or TAT. Either way 65C is preferable to stay under on the high end.

[splat]

http://img161.imageshack.us/img161/2794/xeon465fsb2qa.th.png (http://img161.imageshack.us/my.php?image=xeon465fsb2qa.png)

http://valid.x86-secret.com/show_oc?id=160545

465mhz fsb at cas6. I have to set the timings with memtest, I can't boot if i set it in the bios.

on a more stable note, I got thru 20 hours of orthos with 425mhz @ 1.375v.

*edit: note, that pic and the cpuz valid post were set using different programs. the screenshot was using the biostar utility, and the cpuz validity was taken using clockgen to set the speed. I did a few tests in between the two that resulted in reboots and cmos clearing.

[leojharris]

When i left for work this morning, i was just passed 9 hours in orthos running at 425mhz, 1.375v. I've left taskmanager open so I can look for throttling, and I've noticed that every once in a while I will see a tiny blip in the cpu chart, but it will only be one core, not both. Is it normal for small, infrequent spikes? or is that a omen for problems?


is task manager an adequate tool for monitoring the occurences of throttling? i've used it to monitor cpu utlization but does it also do throttling detection?

i've always used rightmark cpu clock which will actually record all instances of throttling as well as give visual and audio alerts when internal throttling occurs.

give it a try if you haven't already done so; it's a free app: CPU Righmark (http://cpu.rightmark.org/products/rmclock.shtml)

seems to me, that orthos, when occasionally cycling beginning of new tests and what not, would naturally show a drop in cpu utilization as it ended one series of tests and began another; this would only be a momentary thing but seems reasonable (you see it also in your temp readings when temps very briefly drop 5 to 6 degrees and then a second later are right back up to full load.

[greenmaji]

Has anyone run trottlewatch with C2D and actually got the CPU to make it report trottleing?
http://www.panopsys.com/downloads/ThrottleWatch_2_0_1.zip

[leojharris]

i've never used throttlewatch but am going to check it out.

when using a pentium D 820, i got throttle alerts like crazy at temps past 73C and when using the CPU Rightmark program.

i've had my core 2 chip hit 72 at least and so far no throttle reports; maybe the throttle threshold is higher for the core 2 6400 than it is for the pentium D?

[greenmaji]

The temperature reported with C2D with Coretemp and TAT is higher then the temperatures reproted for pentium D's via bios or speedfan, PentiumD's don't have a core thermomister.
I keep reading C2D throttles at 85C in coretemp or TAT, but havn't had anyone report it happening.

[leojharris]

The temperature reported with C2D with Coretemp and TAT is higher then the temperatures reproted for pentium D's via bios or speedfan, PentiumD's don't have a core thermomister.
I keep reading C2D throttles at 85C in coretemp or TAT, but havn't had anyone report it happening.
that's great to hear! meaning i've got quite a ceiling to work with if my load temps are 55 or so ...

i always back off on my stress testing around 72C due to my dual core training ... :)

thanks for the info.

[splat]

so you guys are saying that 65c in coretemp is not bad? I've never had to deal with throttling, but I thought I remembered seeing some where that it would drop the cpu cycles down for a second or 2 to reduce heat, so I figured task manager would show that. while i was viewing task manager i noticed those small spikes, and that makes perfect sense that it happens at the start of a new test, tho i haven't looked to see if that is the case.

update: I'm running 430mhz @ 1.375v thru orthos right now, ~2 hours in, temps stay below 60c

[greenmaji]

For benchmarking no.. but if you have to run high Vcore, high temps isn't a good idea long term 24/7.

[Vulcan]

Have any of you guys hit a wall on the 965pt around 515-520mhz?

I'm pretty much stuck at 500x7. Not bad, but I can tell this chip still has lots of head room and I'd hate to see it go to waste. ;)

EDIT: Also, I've been using the shader vmod to adjust my vdimm. Seems to be working just fine, but I'm thinking it might be a good idea to switch to the variable resistor somewhere down the line. I also need to find a DMM, my analog one doesn't quite cut it for stuff like this.

[leojharris]

so you guys are saying that 65c in coretemp is not bad?
if 65C represents the very rare and intentional overstressing of your CPU, sure ... it's fine. i think that's what greenmagi meant by benching.

your question brings up the issue of the difference in temps between the rare stress test and any given user's 24/7 use. for instance, my machine loads around 55 when running orthos for 12 hours, but my normal pc usage, including video rendering, etc. never results in temps higher than 44 or so ... i just never use the cores as heavy as an orthos test.

random joe user, however, might be folding 24/7 and his/her CPU might indeed get near the same temps as the orthos test ... in which case, 65C ... although close to being in spec for intel's core 2 specs, ... would still be rather chancy at constant 24/7 use.

again, this is basically what another user has posted ... but thought that clarification of certain usage scenarios might help.

BTW ... the xeon work is looking good!

[splat]

430 errored out, I don't think I'm going to raise the vcore above 1.375v until i get a better cooler. So, i'm back down to 425mhz fsb. 3.4ghz @ 1.375v aint bad.

now for some 3dmark scores:
http://service.futuremark.com/compare?3dm06=1081467
http://service.futuremark.com/compare?3dm05=2601792

[deathman20]

430 errored out, I don't think I'm going to raise the vcore above 1.375v until i get a better cooler. So, i'm back down to 425mhz fsb. 3.4ghz @ 1.375v aint bad.

now for some 3dmark scores:
http://service.futuremark.com/compare?3dm06=1081467
http://service.futuremark.com/compare?3dm05=2601792

Better then what I can do. I need like 1.425V to get to that speed.

[splat]

I think I'm done for now. Maybe in a few months i'll jump to a water system and better ram to really see what the proc can do.

[splat]

small update. decided i'd do a little cpuinfo

~ $ cat /proc/cpuinfo
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Xeon(R) CPU 3050 @ 2.13GHz
stepping : 6
cpu MHz : 3394.885
cache size : 2048 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe lm constant_tsc pni mo
nitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm
bogomips : 6792.88

processor : 1
vendor_id : GenuineIntel
cpu family : 6
model : 15
model name : Intel(R) Xeon(R) CPU 3050 @ 2.13GHz
stepping : 6
cpu MHz : 3394.885
cache size : 2048 KB
physical id : 0
siblings : 2
core id : 1
cpu cores : 2
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 10
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe lm constant_tsc pni mo
nitor ds_cpl vmx est tm2 cx16 xtpr lahf_lm
bogomips : 6789.57