Is this CPU useless??
Who will buy this if there is 5820K
it has higher single core performance by about 15%, so if you need the highest single core performance you can get, it's a better choice than the 5820k.
That being said I got my 5820k for $290 at microcenter and I have it OC'd to 4.5GHz (can can get it stable up to 4.7GHz) so i'm pretty happy with my 5820k. It's not like the 5820k is lacking in single core performance, its just not going to be as good as the 6700k.
In anything properly multithreaded of course the 6700k is a good 50% slower.
>>55374094
it's a good upgrade if you do more than just play gaems
>>55374125
Motherboard?
>>55374158
Right now I have an MSI X99A MPower, before I had an ASRock board but it kept giving me issues with my RAM and microcenter let me pay the difference to upgrade to the MSI board.
It's been pretty solid so no complaints.
>>55374179
I'm thinking in buy 5820K + X99 Sabertooth and overclock it to 4.5Ghz
>>55374220
Shouldnt be too hard, 4.5GHz is pretty easy with moderately decent cooling.
>>55374094
Give me the 6700k over any placebo-cores of X99.
There is no good GPU rendering program that uses anything past 4 cores.
>>55374285
What about VMs
>>55374094
It's the best gay men CPU on the market right now. Even the 5960X can't match it in most gaymes.
https://www.youtube.com/watch?v=ocwwaVGUFtk
That may change in the future and the hex and octa core parts may hold up better five years down the line, but right now there's no reason to go X99 if all you care about is gay men.
>>55374298
A stupid endeavor.
>>55374285
>everyone only uses computers for rendering
encoding is still primarily done via software encoders run on the CPU because GPU hardware encoding is dicks quality, sure it's fast, but you have fixed quality settings that make it shit.
>>55374311
Who's speaking about gayming¿
>>55374285
id rather buy the x99 platform for the ease of upgrading
currently on X58 and it still kicks ass with those 6-core westmere xeons that cost 100 dollars a pop now
>>55374395
Yeah, X99 supports up to 10 core for consumer CPUs or up to a 20 core Xeon. In 5 years you can probably get a cheap drop in upgrade.
>>55374338
Me, you fucking retard. The OP asked what use that CPU has. I provided the major one. I don't give a shit about what the rest of you are circlejerking about.
>>55374094
Has CPU performance gains just stopped? I'm using a computer I built in 2010, and I feel the only real bottleneck is my GPU.
I was going to buy a new PC, but based on the tests I've been looking through, there really isn't that much of a jump from my ancient i7 950, to an i7 6700k. Obviously there is some difference, but the 6700k only performs about 65% better, which is nothing considering these CPUs are 6 years apart. Back in the 90s and 2000s, the performance gap would have been huge over a 6 year span.
So, what's the deal here? Should I just get another GPU and stick with this CPU until Intel pulls its head out of its ass and start making some gains in CPU development? I really don't give a shit if they keep making smaller resistors if that only amounts to an 11% yearly gain in performance.
>>55374311
unlikely in the next decade at least.
single core performance will always be a good investment for general use. the game only changes if you are doing heavy workloads
>>55374984
Lack of competition
Intel focusing on iGPU performance instead of CPU performance
CPU Performance will actually go backwards in the next couple years as intel focuses on power efficiency
>>55374984
>only 65% better
It's not only Intel, fucko, its the whole industry.
What you recommend me anons?
6700K or 5820K ¿?¿
I'll use it for a little programming + VMs (1-2 virtual machines at same time i think) and a little gayming when I buy 1070 soon
>>55376929
If you are too stupid to pick your own processor, you have no business on here.
>>55376094
I mentioned Intel, because they're currently the industry leader in desktop CPUs, dumbfuck.
>>55376994
You apparently are so dumb you don't realize the limits of silicon, and why 10ghz processors are unfeasible.
>>55374984
lot of it is internal, company cutbacks brought on by market forces. PC sales have been in the dumps for the past fifteen years. As well, in the past few years tablet sales have gone to crap. Simultaneously, the advancements in hardware have outpaced the overall programing needs of "99%" of humanity. "99%" still only used computers to run a browser; for Netflix, Facebook. check email; run a word processor; and play simple games like Minecraft, Farmville, Candy Crush, etc. You do not need hexa-core 4.X ghz CPU and "500-bit" 6gig RAM Video Card for any of that; the masses are satisfied with hardware as it has been for the past five years. Not to mention the Great Recession has left its mark and even less people have nor willing to part with the several hundred, or few thousand dollars, to buy the most "latest and greatest" hardware. This is the dilemma facing most, if not all, companies right now.
>>55377364
Dennard scaling is dead, along with Moore's observation in 2011. It is simply too expensive to advance much further. Every node shrink requires billions and produces less benefit than before.
>>55377098
Nice strawman, asshole. I never mentioned clock speeds. I mentioned performance. But keep grasping for something to whine about.
>>55374984
If you make single core performance better shit gets hotter at extreme rates, but lots of cores aren't usually that useful for people
>>55377500
So what particular performance metric are you referring to? Or do you not have one, just a "general performance metric that is decided by me and only me?"
>>55374125
You forgot the botnet in the skylake arch
>>55374179
>before I had an ASRock board but it kept giving me issues with my RAM
Which board? OC Formula; have worked perfectly with RAM's (Crucial Ballistix Sport) that didnt work with that exact same MSI board, Asus X99 WS and Deluxe.
Any chip with non-soldered heatspreader is crap. Also, Intel CPU's more recent than Sandy Bridge or Sandy-E; irrelevant. Get used for decent price and you wont support Intel with new purchases either.
I have two 5820k's (with Asrock OC Formula) and will be selling them, because they provide no significant performance advantage over my Sandy-E and Ivy-E 6 core chips on X79 southbridge (both chips i got for less than 100$ and boards, Asus X79 WS for similar prices).
>>55377582
What botnet that is not present in all cpus made post 2011?
>>55377601
The only thing you will get buying post Ivy/Sandy Bridge is 25-30% less performance. Not a bad trade off if you live in a 3rd world, or are poor.
>>55377602
For Intel, Nehalem and newer have IME.
For AMD, Piledriver chips on AM3+, doesnt, but some APU's do.
>>55377630
That IPC improvement doesnt translate to my encodes in same clockspeeds, some 5-10% difference between Sandy and Haswell.
>>55377602
The one on skylake won't let you install anything below Win10
>>55377601
>which board
Asrock x99 extreme4
Also I upgraded from a Q6600, I have no complaints. Especially not at the price I paid.
>>55374094
in my country
x99 boards are expensive while z170 are cheap as fuck
btw i rather have a xeon i ever buy depreciated processor
>>55376964
>>55377736
>x99 boards are expensive while z170 are cheap as fuck
The money doesn't matter here
>>55377719
>Also I upgraded from a Q6600, I have no complaints.
And if you had upgraded to 2500k, 2600k, 3820 or 3830k, you would probably have no complaints either.
>Asrock x99 extreme4
That board doesnt even have decent enough VRM to operate 5820k@stock (well it does, but the mosfets sweat like hell). That is the one thing with X99 boards; overpriced as fuck and all doesnt even have proper components to run the cpu's intended for them.
>>55377679
really? You that dumb you fell for that lie?
>>55377666
In x264 encoding tests, you see solid scaling from Sandy Bridge up through Skylake. There is 28% additional performance to be gained moving from the oldest to newest design.
>>55377834
>And if you had upgraded to 2500k, 2600k, 3820 or 3830k, you would probably have no complaints either.
Sure but why should I have? I managed to skip the DDR3 generation entirely. Might as well get the newest platform since I upgrade my GPU every 18-24 months. A 5820k will last a decently long time and since I do a decent but of encoding it's not like I won't get some decent use out of it occasionally.
And again for $290 and the motherboard for $170 (which I returned and spent another $75 to get the MSI board) so I paid ~$525 for a $380 cpu and $275 motherboard. Saving about $125 off retail.
And if I ever need more cores for whatever reason a high core count xeon will be cheap in another 4-5 years and my motherboard supports up to the 22 core 44 thread e5 2699v4.
So sure, I'd have been fine on something older, but I see no reason to buy used parts or an older generation when I can get such a good deal on the x99 platform and a 5820k.
>>55377864
Like i said, from [email protected] to [email protected], performance difference (x264 encoding with ffmpeg included) is 5-10%.
Skylake or Broadwell might be a different horse, but definitely not 'solid scaling' when from Sandy to Haswell.
>>55377939
Sure, you got a decent deal, but that is not the standard. You could have gotten better deal with previous generation, with insignificant performance 'loss'.
>DDR3 generation
There is hardly argument there. Four channel in practice doesnt give any performance boost over dual, DDR4 has sligthly more latency than DDR3, but runs at higher frequencies. Plus, there is the combatibility problems you have encountered yourself (meaning the platform or DDR4 is not mature yet).
>>55377959
I'm just going to assume you meant the 3930k, which was $999 at retail, compared to the 5820k, which is $380.
>>55378025
The memory controller on -E was always shit.
>>55378025
The argument was that I couldn't do any sort of partial upgrade. My Q6600 was on DDR2. So if I were upgrading it was going to be the whole package, since DDR4 is out and will be the standard from now on, it makes little sense to go with DDR3, especially since I obviously don't upgrade very often if I was coming from a Q6600.
I don't plan on doing a platform upgrade for a long time at this point.
>>55374145
Why would you do anything more? makes no sense imo
>>55378074
Correct, but that is obviously not the price i paid for it used. So, your point being?
The retail price of 3930k (in Europe) took a drop of over 200$ over the first year. Why, dont know, hardly ever happens with Intel cpu's.
The 5820k retail price hovers around 380€, and have done so from the launch.
>>55378127
They not only improved performance, but cut the price 60%. It could be argued that you should compare the performance from a $1k to another $1k chip. But it would make little scene to go with a dated platform if you had a choice today.
>>55378094
>since DDR4 is out and will be the standard from now on
>I don't plan on doing a platform upgrade for a long time at this point
And so, it might not be the standard anymore when you're going upgrade again.
>>55378186
That's the hope...
>>55378127
That old 6 core sandy is roughly the same compared to a 4 core devils canyon
>>55378174
>but cut the price 60%
Which means very little in Intel pricing scheme.
>But it would make little scene to go with a dated platform if you had a choice today.
Saving chunk of money and staying/getting a matured platform (where the performance difference compared to the latest is not significant) makes sense to me.
>>55374285
http://www.cpubenchmark.net/compare.php?cmp%5B%5D=2565&cmp%5B%5D=2340
The 6700k is nothing more than a glorified 4790k.
>>55378247
>where the performance difference compared to the latest is not significant
All right, this is before Broadwell (of which i have no experience), with Haswell-E, but then again, the pricing is higher again with Broadwell chips compared to Hash-E.
>>55378247
Have fun with your dated motherboard. Here's what 1k can buy you today.
>>55378274
...so an improvement?
>>55378299
LUDDITES
B T F O
T
F
O
>>55378299
I sure will.
I even have fun with AMD Piledriver (FX-8350) and Thuban (1100t) chips. In fact i have more fun with them, because i always feel like a sodomized whore when using Intel cpu's.
>>55378334
>he doesn't want Intel to penetrate all his holes
>>55378334
Yeah I get that sodomized feel also. Just not to me.
>>55378370
Enjoy TIM'd IHS and bending waffer.
>>55378370
My 5820k hits 4.5GHz comfortably and 4.7GHz is stable (but hits ~77c in stress tests so I don't keep it there 24/7)
That should match or get close to your single threaded performance while having 2 more cores (and hyper threading on top)
>>55378392
Yeah I'll enjoy my cooler temps compared to the housefire Devil's canyon. And to the bending issue: its non-existent, same as the OS controversy.
>>55378412
And muh 4 core is $240. Not exactly a raping by Intel. Post us some cpumark benchs for us to masturbate to.
>>55378322
An extremely minor one:
https://www.pugetsystems.com/labs/articles/Haswell-vs-Skylake-S-i7-4790K-vs-i7-6700K-641/
>>55378471
I'm not at home right now (holiday weekend with family).
But I paid $290 for the 5820k and microcenter regularly has it for $320. So for $290 it's a nice bang for buck CPU.
Stock clocks are laughably low however. 3.3GHz and mine hit 3.8GHz before needing any more voltage. Sitting at 1.28v @4.5GHz with Max temps around 65-70c.
>>55374312
Get fucked retard.
>>55377679
You can, they're not supporting after next year though.
You're still wrong though. It's okay this is /g/ it's just Linux users on their mom's old Dell bashing anything they don't like because they can''t afford to try it and some redditors who got their mom to buy them a $3K gaming rig for getting straight A's in summer school.
>>55377835
people on /g/ don't actually know anything about hardware or any of this crap, they read a headline and accept it as fact. real knuckledraggers on here.
>>55374094
P-please no bully!
I'm about to get it because it's on sale. Can anyone recommend a good gaming Mini-ITX board for it?
Might get either of these cases:
http://www.newegg.com/Product/Product.aspx?Item=N82E16811139033
http://www.newegg.com/Product/Product.aspx?Item=N82E16811133260
Will be using Photoshop/Blender/Unreal, doing lots of painting/animation/sculpting, and heavy gaming.
people who aren't actually gaming manchildren who still buy Intel, much less special snowflake Xeons with $300 motherboards and ECC RAM they don't need are hilarious. you could buy a three year old FX 83XX and get ~90% of the same multi-threaded performance without the 300% price markup. You guys are a riot.
>>55378860
Lol even at 5GHz the 83xx can't keep up with stock 5820k multi threaded performance and the 5820k easily OCs a fair amount as well.
>>55378214
Whats the clock they have set for the 8350fx?
>>55374285
Additional cores are still useful if you want to do something while you're encoding video.
>>55379111
Yea anyone doing slightly CPU intensive workloads in the background of their normal computer use will enjoy having more cores.
I like playing CSGO while encoding so I let 4 cores do the encoding and have 2 cores for CS.
ITT retards
6700k is better and superior in the long term because OCing your shit and running it 24/7 is stupid and this cpu is already good enough, find me one game or software that can give it 100% cpu usage on stock settings
if you buy 5820k, you're forced to OC it and then one day your shit wont start and you see a CPU led giving you the mordor eye.
>>55380124
>what is multitasking?
Go back to >>>/v/ if you use your computer only to play vidya.
>>55380279
>i encode four 4k videos at once every day
sure buddy kys
>>55380124
People are still running 2500k and 2600k at 4.5GHz. Something tells me my 5820k will live since I keep mine under 1.3v and under 70C.
>>55374298
Wouldnt touch skylakes anyhoo. Too quirky. Source: I've built a lot of virtualization workstations and servers.
>>55380382
Why under 1.3V , I see so much Cpu-Z overclocks with 1.3-1.4V
>>55381833
>>55381853
Can pretty much promise that overclock is good for nothing but logging into windows and popping open CPU-z to take a screen shot.
No way are you stable under 1.3v with 5GHz. That would be the top 0.001% of the silicon lottery.
>>55381833
Because past 1.35v is where degradation can occur internally. It might work for awhile, but you can damage your CPU long term if you run it with that voltage all the time.
I could probably go over 1.3v just fine, but as a rule I stick to under 1.3v and I try to stay under 75c.
For my 24/7 OC I sit at a comfortable 4.5GHz with 1.285v
>>55381914
>For my 24/7 OC I sit at a comfortable 4.5GHz with 1.285v
MOBO?
>>55381941
>>55374179
>Right now I have an MSI X99A MPower, before I had an ASRock board
What's the best bang for the buck, first-desktop-build dolphin emulator CPU?
I've been using a mediocre AMD CPU laptop for streaming, shitposting, and playing basic online games with, but now dolphin has great netplay and I'm dying to build a cheap desktop just for emulation.
Is the 6800k worth the price? Or does the 5820k hold up to it? I was planning on upgrade but now I have to wait till August.
>>55381833
Not really worth it. My 3770K does 4.4 @1.25V, but needs 1.35 for 4.5. Anything past 1.3 will lower your life expectancy. 1.32 might be OK in some cases, but 1.35-1.4 is a bit high.
>>55381853
Is this even real? The bigchips are harder to OC and 5 GHz even below 1.4V sounds impossible.
>>55382727
I saw 5820Ks for 250 €, I'm really thinking of selling my Ivy platform and get a 5820K. Fuck the single core IPC (I play 4K), these cores will convert much faster.
>>55382727
>http://www.anandtech.com/bench/product/1727?vs=1320
They are basically neck and neck in everything.
6800k has very slightly higher IPC and they OC to basically the same level, the 6800k will use less power at idle and be able to turbo boost higher if left at stock clocks (if you OC this will be negated though).
people who does calculations either academically or in business. You need those cores to calculate the extremely complex formula in the spreadsheet mate. That and running math simulations for a research paper
>>55380124
>OCing your shit and running it 24/7 is stupid
there's nothing stupid about free performance faggot, even the most extremely overclocks are only going to reduce the CPUs MTBF from 30 years to 15 years.
>tfw you fell for the 6700k meme
>>55386139
Why is it a meme
> inb4 the 5820k is better because will you will obviously oc and run it for errands that will use all of your cores everyday
>>55378274
>The 6700k is nothing more than a glorified glorification
I agree. It's only the 4790k (better than the 5820k in single/quad) with the added benefit of quad-channel.
Too awesome. can i even handle it?
>>55382776
>Is this even real?
it's not.
He probably bluescreens as soon as he starts prime95 for more than 10 minutes.
Since this thread is in the general gist of meme intel cpu's,
I5-6600k and I5-4690k. How much of a difference in performance have between eachother? Plan on upgrading soon along with a GTX 1060 later this month
>>55388058
>I5-6600k and I5-4690k
You generally never consider the i5 series at all.
Hyperthreading might be a meme, but it's still a more significant difference than the placebo cores of x99.
From i5 to i7...the MC goes up almost 50%.
>>55388058
>along with a GTX 1060 later this month
enjoy an extra month of OUT OF STOCK
What OS do you guys use with your 6700k? Windows 10? 8.1? 7?
Is Linux even compatible with skylake?
>>55374094
I don't have a goy card so the 5820k wouldn't have cut it. Also power draw and heat and even with more cores the 6700k is faster per corr so the 6700k is I reckon faster than the 5820 to up to 8 threaded processing, and that's more than enough for me. But if you're running 2-3 VMs and so on I'd get it.
>>55388439
windows 8.1, its better than 7 with the start menu but I'm not sure if its possible to install 8.1 on Skylake processors with a 2.0 stick
In gayming related applications. Realistically, will we see same situation with octa cores as with dualcores, when suddenly single core processor became irrelevant? Will new API ever reach that kind of heights?
My painful example, when AC2 came out on PC and I still had single core Athlon. 20fps was considered nice.
>>55374094
I OC'd mine to 4.85GHZ. What did you guys get?
>>55388929
did it help with anything or did you do it just to do it?
>>55388937
Just to do it. It's free performance so why not?
I used to run mine at 4.8GHz, but I think I have too many USB devices in, or some dodgy driver.
Either way it now bluescreens a bunch more than it used to (and not clockintteruptnotreceivedonsecondaryprocessor error)
One day I'll find which drive it is; but I can quite happily live at 4.2
>>55388950
I am stickler for temps, I hate it when it goes higher than it supposed to.
>>55388968
stupid question: isn't it better to have higher clock if you have lots of usb devices?
>>55388982
the more non-CPU things trying to interact and stay in sync with the CPU, the more likely something breaks.
A higher clock can benefit many devices. But I don't think this is one of them.
>>55380124
You do realise the motherboard will die long before a OC'd CPU dies right?
but yes a OC'd CPU will never live as long as a cpu at stock but stock with most cpu's is shit anyway.
don't get me worn i'm buying a skylake at anyrate but its not going to beat any boardwell /-e or haswell /-e chips in anything outside of IOP
skylake is near the end of its span and intel is going to release kabylake
I'm not interested in kayby lakes IGPU or its meme 200 union board.
but skylake is really nearly over.
I'm only geting skylake because
A :im tired of saving up
B: i'm not waiting more months
C: the premium estimate is silly for a 4 core
D : I can deal with 2 less fake cores for one real core.
E: IGPU is even worse than the oldest 7800 APU so i don't care to use it.
besides i doubt there is much improvement in kayby lakes IGPU
F: 6700k on my motherboard will have a max vram share of 512mb i mean c'mon, really?
that is only twice the vram of onboard video 10 years ago.
*humpf* why bother with any intel IGPU
>caring about cpu when it will never bottleneck gpus
>>55388439
Linux 4.6.3 (Though not using the IGPU)
>>55389857
I've even seen 3D games bottle-necked by the CPU. Just because it's unusual doesn't mean it's impossible
>>55380279
>>55380382
>>55384737
>>55388996
Can you all stop replying? He obviously is mentally ill. [email protected] 1.355v since a few months after launch here. She went through many months of torture tests in my old hell hole.
