Separate names with a comma.
Discussion in '[H]ot Deals Discussions' started by primetime, Sep 29, 2013.
Cool. Ordered 1 to upgrade the 920.
If you have a 3770k run the test post your result. Then we will run the same with the L5639. This would be the easiest way to help you and everyone else out here.
i posted the cpu tests 2 pages back or so, you can take a look, it's a toss up really, but L5639 is almost five times cheaper than 3770k and probably only better if you can overclock it to 3.6ghz or higher and are going to utilize all cores and threads
this info has been posted already...but ok here we are again...keep in mind the biggest factor in power draw is the video card used in test pc...anywhere from 100-200 watts difference just at idle is possible...and much more while gaming something like bf4
right now mine is drawing 170 watts from the outlet...but push the gpu to 99% usage and cpu to heavy load like say running bd_rebuilder at the same time and you can easily pull 500 watts...id say around 450+ watts just playing bf4 alone.
I m planning to built new pc just for gaming..is it worth investing in this old tech or go with new haswell 4770k.. i upgrade my pc every 5 to 6 years so planning to keep it for long time..one more thing i already have 16gb ram(8x8) can i use it with X58 mobo.. can i just use 2 stick of ram or do i need 3..if i get 4770k processor worth around 280 and if i get Xeon both mobo and processor is around 269..any advice would be helpful..
In your case I would just go with a Haswell 4770K and here is my logic. The Haswell setup is not as robust, but equally as fast. Also the power difference isn't as significant when you throw in the fact that most likely you will be running a gaming class Video Card. Also since you have 2 sticks of DDR3 8GB ram it is perfect for Haswell, but you would need a 3rd stick to do triple channel on the x58 platform.
This platform is great for folks that already had an x58 setup and wanted to extend its life. If you have an application that can take advantage of all the threads available this will beat the Haswell parts at stock, and maybe even on mild overclocks (4.2Ghz), but as Haswell climbs closer to 5Ghz I think it would pull away. The newer boards will have more USB3 ports, and many times less legacy connectivity for faster boot times.
Now you could always go with the latest generation Xeon if you just need to have a Xeon in your life . Something like the E3-1275 v3.
Are there any Asus p6t deluxe people here who have no 16x throttle when pushing more than 2 threads...I read about an actual asus 06 bios that was never actually put on asus downloads page that....but people were apparently using it to stop the throttle issue...problem is Im scared to flash to a much older bios that might not even support the xeon hexa core...hopefully another one of the old school experts will check back into this thread and answer if this is doable, but no biggie if it isn't....this board is so stable and i hate to screw it up now
I previously posted about having an issue with 5 minute or more POST times after installing the L5639 processor.
I have apparently remedied the problem by not turning on my monitors on before turning the computer on. If I only turn on one or two monitors not all three it will post fine. It is still a little slower than the 920, but we are talking about 10 extra seconds compared to minutes. So, I don't think it is related to processor.
Just wanted to post this for any googlers.
Just switched from a quad core Intel i7 920 OC'd to 3.8ghz to this Xeon OC'd to ~3.7ghz (205 BCLK). The performance gains in FPS are quite significant.
FRAPS results taken on 64p Large Conquest server -- Floodzone:
Setup: X58 Sabertooth, GTX 780 @ 1100, 6gb ram, 1920x1200, Ultra 4xMSAA
Min fps Avg fps Max fps:
52 77 110
Six core :
67 94 134
Core usage on the Xeon averages about 70% across all six cores, with a few spikes to 75% in heated moments. I forgot to check the i7 core usage before I replaced it, unfortunately. The max and min fps do vary quite a bit across my bench logs since no two game sessions are identical (high fps max with no action; low minimum with lots of explosions, etc) but the average is pretty consistent.
Conclusion: about a 20% gain in avg FPS by switching to this CPU.
Edit: Running the Xeon @ 1.32v, 1.36v QPI, with a Kraken X60. Temps never exceed 52c running LinX. I'll try for 210 BCLK tomorrow and see what happens.
Can you post a screenshot of the task manager while playing that map...was just curious if any difference using nvidia 780 card vs. mine
I ran Unigine Valley Benchmark on my 920 @ 3.6 and my Xeon @ 3.6. I gained 1.5 FPS.
I'm sure a game like BF4 would be a better comparison.
like soon as you realize the storm in hitting full swing, and the waves are maxing out alt out and grab a screen shot lol one would think those waves be hard on cpus....impressive for sure
Edit: Windows 8.1 isn't reporting the correct frequencies (should be 3.7ghz). Off topic, but this is actually causing some problems with nVidia Shadowplay because it's unavailable due to the CPU frequency not meeting the minimum 3.1ghz requirement. Any solutions to this? Would appreciate it!
First thing i notice......you have hyper threading turned off in bios (always scored lower in beanchmarks for me...not to meantion games)
Secondly..if you switch the graph to logical process instead of overall it would be a more accurate comparison
third...its showing pretty accurate cpu speed for mine...what blck are you running at? you can double check with HWINFO64 for blck and current multipliers for each core....just bring up the summery screen...
here's mine with turbo engaged...win8.1 reports its nearly perfect
Uploaded with ImageShack.us[/IMG]
Hyper-threading seems to have no affect on game performance, but it is stable with this overclock thankfully. CPUZ and HWINFO are all reporting the correct frequencies. Could you perhaps post your BIOS settings so that I can compare them to mine? I can't figure out why Win8 isn't recognizing this correctly.
Oh wow...you using exact blck as me...yea i can do some bios pictures latter tonight...but theirs is a couple people in this thread with your same motherboard....you lucky as your board doesn't throttle to 16x like mine...im thinking our voltages will be pretty close for stable overclock, but the only time mine will match yours in using 2 cores...otherwise it throttles down
are you using win8 or win8.1?.....im using the 8.1 preview so that might account for the difference in reported clock speed...if not well have to keep looking for answers
Now looking at your task manager while playing bf4 it looks very close to mine now...was wondering if the 780 made a difference in cpu usage...looks minor though
Take it for what its worth..passes every test i can throw at it, 16 hrs of bf4 (so far), just can't make it crash. The only bug i have is sleep mode doesn't wake up above 200 blck.....so i just don't use sleep mode...The turbo only works on mine IF the power savings is turned on, so i keep that turned on. speedstep lowered my benchmark scores, otherwise i would have it turned on. Turning HT off will keep my multiplier from throttling down to 16x but every test gives a lower score, not to mention im sure making dvd 5 size blurray rips would take considerably longer, so it strays on.
Here's my first attempt at overclocking this amazing chip... The hard part was having to mod my Classified E760, but I think the results were well worth it.
yeah those are very nice cpus for the money, enjoy it, i am
I have an MSI X58M and it is working great with the latest bios.
Just got the chip in and running yesterday on that board and just running stock right now. So I can't confirm any overclocking as of yet.
I've been reading this topic for while and I thought I'd make a account and post. So here is my first post. I'm on other sites as well. I don't come across a lot of X58 topics nowadays so this is great to me. I received my chip about a week ago, I have been thinking about upgrading my X58 for time now, but it appears I won't be doing that for awhile now.
I was originally running my i7-960 Quad+HT @ 3.9Ghz and 4.1Ghz. I have successfully overclocked my new L5639 to 4.4Ghz [stable]. I usually run it at 4Ghz for the lower Vcore settings. I have also included some comparisons. I wanted to do more before and after benchmarks, but I couldn't resist and HAD to install my new Xeon.
Here is my basic setup. The only thing that changed is my CPU.
Asus Sabertooth X58
Xeon L5639 OC'd @ 4.4Ghz and 3.9Ghz
Antec Kuhler 620
12GB [6x2GB] 1600Mhz RAM Triple Channel
x1 SSD ---- x2 HDD 7,200rpm High Performance
GTX 670 2-way SLI
Now I'll post my benchmarks. Note: I ran all test @ 4Ghz on the L5639
i7-960 @ 3.87Ghz = 6.64
7-960 @ 4.1Ghz = 7.03
Xeon L5639 @ 3.971Ghz [4Ghz] = 10.12
Performance Test 7.0
[I know 8.0 is out. This will have to do for now]
i7-960 @ 4Ghz = CPU Mark - 8237.4
Xeon L5639 @ 4Ghz = CPU Mark - 12345.9
i7-960 @ 4Ghz = 8,519
Xeon L5639 @ 4Ghz = 10,991
Got 109.4 Frames in my Tomb Raider Benchmark as well [1080p]. I used to get around 100. Minor bump. I'm thinking about upgrading my GTX 670s. I don't know. I might upgrade to the 7990 Quad or something. I'm probably going to just wait until next year.
shows the l5639 and it's turbo ability as upto a 4x turbo so 4ghz should not be to hard to achieve.
What motherboard and settings are you running to be able to get that high? What cooling as well?
Ohhh the devil in the details.. Got my L5639 in this morning, but not booting. Cleared CMOS, tried taking memory out, simply nothing.
Double checked compatibility list, and it isn't specifically listed, but found a thread on evga forums that suggests a revision 1.2 of my board supports the Westmere EP's. Trying to identify my board revision, as i couldnt see it on the board itself.
Any program that can assist here? 132-BL-E758-A1
If you read the first part of his post....it says Asus sabortooth with Antec Kuhler 620...I think one of the very few boards that can run that high of blck.....or even boot for that matter......unclear weather its 100% stable yet either
Maybe try SiSoftware Sandra? On our asus boards its easy to tell revision by which bios rom its using.....but that idea didn't work for yours....try a bright flashlight and look near the model number on the board itself?
You could also contact EVGA customer support with your serial number and they can tell from that. Also look on your serial number as it might have the revision number...Not sure on EVGA boards.
OK so i did find the revision number tucked away in the corner under some routed cables.
Revision 1.1 . According to evga forums, revision 1.2 will support the L5639. After more research, i found that there is a simple hardware modification that can be done to this board that effectively disables the second QPI link, and enables this cpu to work (i'm already running bios 83).
Now, here's the really cool part: i spoke with evga customer service: if i ship the motherboard to them, they will mod it for me. however, as it is out of warranty, there's no problem with me doing it myself, if i'm comfortable with it, they said. I am. So they will be emailing me the directions on what to modify on the board! How cool is that?
Do us a favor and run intel burn test 2.54 at very high stress level 2 different times and ways. first is checking to see if your turbo multiplier is really stable. set it to use 2 threads only and once it starts set it in task manager to use only 2 cores. this will run your cpu at 20x turbo (4.4GHZ) multiplier the hole time. take a screen shot of it when done.
Second test set it a very high stress level like before but set it to use 12 threads and let it do its thing at default number of 10 tests like the previous test. your rampage will use the 18x (3.96GHZ) multiplier the hole time and take a screen shot when done.
If you can do these relatively quick test's in my book your overclock is totally stable..feel free to include a summary view of HWINFO64 next to your test just to show actual clock speeds of each core when done.
Good luck and you got me impressed if you get it to pass them both and im sure more than a few asus sabertooth board users will want pictures of your bios settings so they can get the same results.
Guys..i just registered. i really need help, if someone can help me, i be sooo grateful.... some company is selling me a workstation with this cpu... i do hardcore video encoding,rendering, and hardcore gaming, i want to know friends, at stock speeds, can this cpu handle modern gaming? Currently i have an Athlon x4 645 3.1ghz, and im looking for a performance equal to Amd Fx's and Intel i5s at stock speed, Can this CPU do that? im getting it really cheap with 12 gigs of ram, and will add a Very nice Graphics card. what intel processor does it compare to at stock speeds?
Thank you guys so much.. i only got like 2 days for this offer. thank you all.
Can i ask why stock speed? think there's only one person here had to run at stock speed so he replaced it back with his older guad core so HE could overclock again...Im guess its some kind of board that wont overclock? whats the total price? might be better going different route like with different board if thats the case....none of us bought these to run stock lol
I can check tomorrow some time if you really need it
edit for whats its worth here: look here and find it...scores better than these:
Intel Xeon X5647 @ 2.93GHz
Intel Xeon E3-1225 @ 3.10GHz
Intel Core i5-3330 @ 3.00GHz
AMD FX-6120 Six-Core
Intel Core i7 965 @ 3.20GHz
Intel Core i5-2400 @ 3.10GHz
AMD Opteron 6212
Intel Core i5-3335S @ 2.70GHz
Intel Xeon E5-1607 @ 3.00GHz
AMD Phenom II X6 1090T
as just a FYI Mine overclocked score better than these 2 easily (and all for 79 bucks) and programs that make full use of 12 threads it pulls ahead much further (like bd_rebuilder type uses) ( bf4 it doesn't even break a sweat)
Intel Core i7-3770 @ 3.40GHz
AMD FX-8350 Eight-Core @4.2GHz Turbo
from here http://www.cpubenchmark.net/high_end_cpus.html
Thank you very much for your reply, i dont like overclocking, never tried and seeing as im on a huge budget, i cant. im getting the entire system for $350, and thats with 12gb ram. and its a workstation and stock. and if you could find out for me how it performs compared to other chips at stock speed that be great!
currently i run an amd athlon x4 645 3.1ghz, i love this chip, and i hope that this Xeon out performs it. i want to get a r9 270x and game with this rig ): but am skeptical..
oh and iam a jobless highschooler if that helps. i was gonna build a fx6300 rig but its expensive.... and maybe this processor will beat the fx6300 at stock speeds. Looking forward for more info. Thanks a lot Sir!
3513 vs 5945 (and that was a slightly slower clocked xeon)so about 60%+ faster/more powerfull..really depends on the applications you use...so far were only looking at a score given by pasmark.....take it for what is worth..with overclock it almost triples the amd stock score...and thats with a 50% overclock only...huge upgrade anyway to look at it
I'm using a Sabertooth btw and thanks for the luck . I've tested the system @ 4.4Ghz over the weekends. No issues. Stable. I did test only two cores a week ago, but stopped. I don't really like doing that [1-2 cores above standard] to my CPU at all. Never have liked that idea. However, I see why you would want me to stress the cores. Since only 2 of them guarantees 4.4Ghz when only two cores are being used. All of the CPUs run at 4.4, but drop obviously to [x18]. Also when I use Intel Burn Test I use the "Maximum" setting. However, I used the "Standard" settings for two cores even though I was reluctant to do so at all as I've explained above. I guess I could run more test on two cores, but I already have and don't want to screw anything up. I'm satisfied with all of the test and comparisons I've ran over the past 2 weeks or so.
Here are my results.
Intel Burn Test v2.54
I by no means want you to do something you feel is damaging to your setup, as a matter of fact i would't be surprised if it did pass the test i suggested with 2 cores....I will admit im a bit baffled as to why you think it could cause damage. my system actually jumps from 16 to 20 when making the 2 core test as yours is only jumping from 18 to 20 and much less likely to find instability from a smaller jump.
Now for me i had to be sure, and im glad i did, cause mine kept failing that particular test ( running 10 times at very high stress) til i lowered my blck to 207....so i dropped it 2 more for the headroom...lol to me its actually a lot less stressful than testing 12 threads since it never really heats up much...its really just verifying its really stable at the turbo freq.....and ill admit its rare if ever an application or program would have it run at turbo sustained for that long of time.( if ever)
But i will say this even with just bf4 turbo is being used on and off all the time...just not a sustained load for anywhere near that long...lol the ocd in me had to be sure that mine was 100% stable
hey i still love my asus p6t delux...but i wish i had gotten the sabortooth back then
It will run circles around a stock Athlon 645 at stock settings!
That's good to hear. I'm glad you got it stable. You are right when you say not many, if any, programs will just use one or two cores. Unless specified. I started stressing two cores about a week ago. No issues. I stopped the stress testing on both of them after a several hours. I don't want anything to go wrong lol. I like where I am at the moment. Yeah I don't go from x16 to x20.....x18 is where my MB likes to sit. Otherwise it's x20. I'm just trying to overclock one my GTX 670 now. One of them is being a real b**** and the other is fine and passing 1306+Mhz. Ugh. I'm thinking about just going AMD and getting a 7990 eventually or one of those 290x or something.
Just my personal opinion...but heck man your good for now...wouldn't even worry about overclocking you cards especially with sli....what are we talking 5-10 frames a second when you already getting 80 plus fps......id say for your good till late next year when the real new cards come out...that's IF there even needed then...I don't consider it new unless these a die shrink....now if your running tripple monitors or needing 120 fps then that's different
let me know if ya want to trade motherboards...j/k im actually very happy with mine as well
What chip would this Compare to on stock speeds? i was going to build a Amd FX 6300 Gaming Build, but im getting a xeon l5639 system for dirt cheap, and budget is a huge impact for me. so just wondering how it stacks to others on stock speed, and i hope someone runs this on stock speeds so i can know