should i get an i5 2500k or amd fx 8120 8 core??

[PG]

Looks like here is another review of the 95W version of the 8120.
http://www.xtremesystems.org/forums...w-diamond-from-FlanK3r-AMD-FX-8120-first-part
The guy said the english was from a translator so it can be a bit hard to follow.

It is interesting, but no, not the same or better than a 2500k. Don't get me wrong.

 

[skipsneeky2]

Captain Kirks bloody frozen tampon popsickle treats!

 

[BallaTheFeared]

Well I guess it does exist.

However his testing methodology is worthless to say the least.

He took power draw samples using linpack.

8120

1100T

1100T is almost three times faster.

He is from China though, so what's available to him isn't always available to us.

 

[digger72]

Core i5-2500K has been the best selling processor - and there are reasons for it. The only reason I could think in favor of FX 8120 is when you need a separate graphics. The FX8120 still lags - kind of.

 

[Puppies04]

You are now basing your numbers off interpolated data assuming the 8150 and 8120 power usage is a simple direct relationship

Nope. I linked you specific data regarding the 8120.

and even assuming it's accurate enough your worst case scenario has the 8120 using about 86W more than the i5. At 2 hours a day of usage, it would take 5 years before the power cost even equalized the $40 price difference

Now you show me some proof Mr bigshot. I pretty much 100% guarantee you that you pulled that out of your ass.

For less than $35, you can get a nice budget 500W power supply, which is more than enough for either system.

Now I need to educate you about the fact that almost all the 2500Ks and 8150s/8120s (heh lets group them together because no matter what you say they are effectivly the same CPU) are not going to be run in a system without a GPU which starts to take us into the mid-premium PSU market. That, added to the fact that you think that you can buy a 500W "budget" PSU for $35 that isn't going to run a significant risk of bricking your entire system shows me that you have no idea whatsoever of the subject you are taliing about (heh like your comments in this thread made anyone believe otherwise)

 

[Cstefan]

I do love popcorn and fanboys going back and forth playing soggy biscuit.

 

[infoiltrator]

newegg has a 1pm shellshocker on OCZ ZS750 power supply for $55 after $25 rebate. Yes you can get good 500 watt power supplies for $35 after rebate. I would say $40-50 is easier. Corsair, Antec, OCZ, Rosewill.

Unless you find the AMD motherboard features an advantage, Intel is the best choice.
OTOH the 8120 is not a bad choice, just not as good. It is slower in low resolution gaming (which is likely on a budget build), uses more power, has a screwy (at this time anyway) architecture, and costs more for most people.
The power cost, power supply cost, cooler, and case/case fans cost, do add up but are not deal breakers unless you plan 24/7 overclocked 100% usage.

8 pages of vehement arguement on a choice unlikely to make a lot of real world difference does seem excessive. We build these things for enjoyment as much as science. Over the 3-5 years we keep these things playing with one or the other is a what the hell why not choice.
If you change your mind selling off should be a small loss proposition.

 

[Chiropteran]

Nope. I linked you specific data regarding the 8120.

No, you actually didn't. The fact you don't even know what you yourself are writing just shows how delusional you have become. I was responding to your last post. Lets see, here is your last post:

Here is anands test setup including a 5870. Call his results lies if you dare.

http://www.anandtech.com/show/4955/the-bulldozer-review-amd-fx8150-tested/9

I5 2500K
Idle 76
Load 133

FX8150
Idle 85
Load 229

As your made up link shows a 1W difference between the 8150 and 8120 at load I will be kind and let you subtract 10W (10 times as much) from the 8150 to get a good 8120 estimate....

Still roughly 86W more on the AMD rig.

Edit.

I think you need a bigger shovel, that hole is getting deep.

Nope, not a single mention of the 8120. You even wrote "8150" in bold, yet you still try to claim you were talking about the 8120. This is incredible. More lies debunked. Keep lying, maybe you can fool some people, but not me.

Now you show me some proof Mr bigshot. I pretty much 100% guarantee you that you pulled that out of your ass.

I guess when you post lies in every single post you begin to assume everyone else is lying also. Luckily math is a very simple process and the proof is easy to see.

86W - your figure from the above quoted post
.1216 per kw/h- my cost for electricity.
365- days in a year (seems obvious enough but since you couldn't do the math yourself I guess I better spell it out so a 5 year old can understand)
365*2*5= 3650 (total hours for 2 hours per day for 5 years)

3650* .1216 * .086= $38.17 (Total hours times cost times killowatts, 86W is .086 killowatts)

$38.17 is less than $40.

Now I need to educate you about the fact that almost all the 2500Ks and 8150s/8120s (heh lets group them together because no matter what you say they are effectivly the same CPU) are not going to be run in a system without a GPU which starts to take us into the mid-premium PSU market. That, added to the fact that you think that you can buy a 500W "budget" PSU for $35 that isn't going to run a significant risk of bricking your entire system shows me that you have no idea whatsoever of the subject you are taliing about (heh like your comments in this thread made anyone believe otherwise)

I don't know why you put "quotes" around "budget", but I guess you are trying to "imply" that the "power supply" isn't really as "good" as it is actually rated for. Nice try, but thermaltake is not exactly a bottom tier producer and even anandtech had lots of great to things to say about their power supplies.

Also, maybe you should educate yourself if you think 500W isn't sufficient.

http://www.anandtech.com/show/2624

I really question why you are even posting on a tech forum. I mean, you have been posting CPU wattage figures for the last three days, so I have to assume you know how much power the CPUs use. It's not exactly super complex math to add the power consumption of a video card to that figure, and an additional 100W to cover MD, drives, and act as a buffer, and add about 15% for power supply inefficiencies. You want to assume the system is going to include some big power hungry $300 video card, but I'd like to know where you came up with that assumption. I suspect you just made it up on the spot because you ran out of arguments, but it changes nothing.

Lets add a 6970, which might draw up to just over 300W.

So, based on my calculations I'd need just under 700W with the 8120, and i'd go with something like this for $60-
http://www.microcenter.com/single_product_results.phtml?product_id=0339039

Since you said you could save the $40 difference (or more) since you won't need such a strong power supply with the i5, I'll just let you go ahead and find a quality power supply for $20 or less that will drive the i5, MB, accessories, and a 6970. Go ahead, I'm waiting.

 

[Markfw]

You do realize that 700 watt PSU puts out 560watt usabeble, but that 80% is not at 100%load. So you would be luck to get 500 watt usable, and then you are running at full power. PSU would self-destruct in months.

My Corsair AX850 is over 90% efficient, and then I would only get 765, but that should just barely do the job, and thats $170.

 

[BallaTheFeared]

You do realize that 700 watt PSU puts out 560watt usabeble, but that 80% is not at 100%load. So you would be luck to get 500 watt usable, and then you are running at full power. PSU would self-destruct in months.

My Corsair AX850 is over 90% efficient, and then I would only get 765, but that should just barely do the job, and thats $170.

Wut? That's wrong I believe.

The 90% efficient means it needs to draw more than 850 watts from the wall to deliver 850watts to your components.

 

[Puppies04]

Wut? That's wrong I believe.

The 90% efficient means it needs to draw more than 850 watts from the wall to deliver 850watts to your components.

Correct most reputable power supplies actually are capable of producing the rated power output, the efficiency just relates to the amount of juice pulled from the wall as ballathefeared pointed out.

That said the 90% efficiency is not tested at 100% load, the PSU has to hit certain efficiency points at certain output levels to attain the certification.

 

[Puppies04]

I guess when you post lies in every single post you begin to assume everyone else is lying also. Luckily math is a very simple process and the proof is easy to see. 86W - your figure from the above quoted post .1216 per kw/h- my cost for electricity. 365- days in a year (seems obvious enough but since you couldn't do the math yourself I guess I better spell it out so a 5 year old can understand) 365*2*5= 3650 (total hours for 2 hours per day for 5 years) 3650* .1216 * .086= $38.17 (Total hours times cost times killowatts, 86W is .086 killowatts) $38.17 is less than $40.

Ahhhh now I see why you used 2 hours per day, is that some magic number or did you just choose it because if we worked out 3 hours a day what you have been saying would be proved to be rubbish. Well guess what I can prove that even at 2 hours per day you are wrong, see below.

Also are you aware that PSU's become less efficient once you get past about 50% load (as well as the one you linked only beign 80% efficient anyway). Assuming most sensible people buy PSUs that will power there rigs at about 50-75% of the rated PSU load (which I can assure you they do) they you forgot to factor in the extra juice being pulled from the wall to compensate for the extra 86W which is almost 20% of a 500W PSUs capacity.

So we are looking at 86W +20% standard inefficiency 86 + 17.2w = 103.2w

Lets plug that into your maths....

3650* .1216 * .103= $45.71 (Oh noes, proved wrong by your own maths )

Edit.

But lets ignore that for a minute, you say I need to find a power supply for $40 less than your $60 choice actually I dont I have to find one for $1.83 less (assuming we use your calculations based on a measly 2hours usage per day and not mine) so I choose the 600W version of the one you linked which is $10 cheaper on the same website which will happliy run the same setup with the 2500K.

Whichever way you look at it I was right and you were wrong.

 

[BallaTheFeared]

Correct most reputable power supplies actually are capable of producing the rated power output, the efficiency just relates to the amount of juice pulled from the wall as ballathefeared pointed out.

That said the 90% efficiency is not tested at 100% load, the PSU has to hit certain efficiency points at certain output levels to attain the certification.

Johnny Guru does some great PSU reviews, check out his website: http://www.jonnyguru.com/index.php

Here is my HX850: He tests cold, and in a hotbox. At different power draws.

70A is the max rating for the 12v rail, which is 840 watts.

 

[Markfw]

Thanks, I stand corrected !

 

[Chiropteran]

Ahhhh now I see why you used 2 hours per day,

Yes, I used that number because I wanted to post something accurate and truthful. I didn't just pick a number out of thin air like you seem to do.

-
So we are looking at 86W +20% standard inefficiency 86 + 17.2w = 103.2w

Cute, but wrong. The wattage figures from the AT article are A) for an 8150 (therefore useless), and B) measured for the entire system, AKA from the wall. The inefficiency of the power supply has ALREADY BEEN ADDED, you can't add it again.

Nice try though.

But lets ignore that for a minute, you say I need to find a power supply for $40 less than your $60 choice actually I dont I have to find one for $1.83 less (assuming we use your calculations based on a measly 2hours usage per day and not mine) so I choose the 600W version of the one you linked which is $10 cheaper on the same website which will happliy

What the hell are you talking about now?

This is what started the whole line of the thread-

Originally Posted by Puppies04
While we are talking about this supposed "extra 40 bucks" the 2500K rig is going to cost you can we take into consideration the fact that higher rated PSU's cost more money.

You didn't say "over the course of 5 years". I made that part up, pointing out just how long it would take before the i5's reduced electricity usage would actually "save" you money. A $10 cheaper power supply is $10 cheaper, not $40. I guess this is proof you failed from the start.

Also, you continually point out the 86W difference, so why are you going with a 100W weaker power supply? Either you are going with an insufficient power supply for the i5, or you are admitting I could do fine with a lesser power supply on the 8120. The inefficiency of the power supply was already included in the difference you assumed above, you can't double dip.

 

[Abwx]

Also are you aware that PSU's become less efficient once you get past about 50% load (as well as the one you linked only beign 80% efficient anyway).

This just show how incompetent you are...
Any EE will tell you that switching mode power supplies
efficency rise above 50% load...

Of course this render the whole rest of your post futile ,
so no need to quote it...

 

[BallaTheFeared]

You wanted to post something accurate and truthful, yet said it's $200 vs $220, then used a 95w part that is actually closer to $240 to twist the argument to better suit your needs while never revieling the large price difference and the common knowledge that lower leakage chips like the 95w Phenom II x6 have lower OC potential. It doesn't even include a heatsink and can only be obtained from a shady company.

Why don't you just admit you twist anything and everything you can to suit your desire to screw the OP over with an inferior product?

 

[Chiropteran]

You wanted to post something accurate and truthful, yet said it's $200 vs $220, then talked about a 95w part that is actually close to $240 and doesn't even include a heatsink and can only be obtained from a shady company.

Why don't you just admit you twist anything and everything you can to suit your desire to screw the OP over with an inferior product?

I was never talking about the 95w 8120, I'm not sure why you thought that. I only brought it up because you asked if there was a 95W version.

 

[BallaTheFeared]

lol you're so full of brown it's sad. I never asked about a 95w part ever, the cool thing about forums is posts are still there the next day when a lier is caught lying.

If you don't even know, why are you in this thread trying to pass along information? You don't really know anything except how to google, do you?

Here are some terms to try: "fx-8120 95w"

Learn something new every day. Now please quit trying to spout off like you know what you are talking about when you obviously don't.



He didn't even know about the 95W 8120, LOL.

If we were talking about the 125w part, which is what I stated there would be no reason to bring up a 95w part since that has no baring on anything we were talking about, not price, not power draw, not anything else.

Just admit it, you're twisting words as a means to screw the OP into an inferior product that every single review site out there will back up 100%. i5-2500k > 8120 & 8150.

Yeah it's not a 8120, it's faster lololololol.

OP wants something that will be good in the future, not something that still would have been awful even if it was released three years ago.

 

[Puppies04]

This just show how incompetent you are... Any EE will tell you that switching mode power supplies efficency rise above 50% load... Of course this render the whole rest of your post futile , so no need to quote it...

You see that red line going down around the 300W mark?

 

[Chiropteran]

I never asked about a 95w part ever,

Yeah the 8120 isn't the 8150, it's still a 125w part though isn't it derp?

/turd

More lies. Not going to bother replying to you anymore.

 

[Puppies04]

Cute, but wrong. The wattage figures from the AT article are A) for an 8150 (therefore useless), and B) measured for the entire system, AKA from the wall.

1. I have already shown you power draw for an 8120 is pretty damn close to an 8150.

2. Where does it say anands power consumption figures are from the wall, seem you just made that bit up.

 

[BallaTheFeared]

More lies. Not going to bother replying to you anymore.

My first post started at #121, #127, #141, #144 your next post after #144 is where you brought up the 95w part and acted like that's what you were talking about all along.

Lies will get you nowhere in this conversation, sorry. Acting like I brought up the 95w 8120 because I said it's still a 125w part after you relentlessly and stupidly argued the power draw difference between the 8150 & 8120 would be something drastic will get you nowhere.

 

[Rvenger]

Guys, we don't need another thread lock for calling names.

 

[MrTeal]

You see that red line going down around the 300W mark?

Results from the last 3 Jonnyguru PSU tests (hot results)

Results are pretty typical for almost all PSUs: terrible at low power levels, reach max efficiency at about 50-60%, then a gradual decline at increasing power levels. It's a conscientious decision by the designers to get as large and broad high efficiency band as possible.