![sandybridge[1].jpg](https://i.imgur.com/cEPrH9E.jpg "sandybridge[1].jpg")
5 years later and still literally no reason to upgrade unless you're editing 8K movies for Paramount.
This is the most future-proof beast of a CPU ever created. With overclocking you're set for another 5 years. Prove me wrong.
>>54144100
Nehalem here. I'm toying with the idea of getting a Broadwell-E when they come out, but even for me its not a pressing need.
>Z77 motherboard busted few weeks ago
>Z77 boards are discontinued
>either have to resort to buying a motherboard off ebay or throw down $400 on a completely new haswell build
Fuck you Intel
>>54144165
Tech retard here, is there any reason Intel and AMD can't just make newer CPUs backwards compatible with older motherboards?
>>54144189
the pins dont match up
>>54144189
AMD does by having sockets last forever
for example, AM3 was launched in 2009 and will be in service until Zen
>>54144210
Why can't they just make the pins in the same configuration as the older ones?
>>54144165
Damn. Sorry about your mobo, bud.
I hope my shitty mobo won't crumble away like that; the prices for Z77 shit is terrible.
>>54144228
ALL CPUS are future proof. CPU power hit a ceiling many years back.
The LOL MAH INTEL and MAH AMD dickwaving contests are pointless. They're both the fucking same.
>>54144248
But how does it benefit CPU companies to make new pin configurations unless they're selling the motherboards too? Wouldn't they make more money by leaving them backwards compatible since then most people wouldn't have to fuck around with building a whole new computer or taking everything out to put it all back on a new motherboard?
>>54144228
Because it allows them to be more creative in designing new chips. The fraction of people that would upgrade their CPU and NOT the motherboard along with it is tiny, regardless of what the inevitable AMD fanboys will say. The CPU already lasts the longest of any part in the PC (besides the case or PSU which don't count) so whatever
>>54144276
they make money licensing the base board design and socket.
>>54144286
>Because it allows them to be more creative in designing new chips
Does the pin configuration really matter that much? I mean, does it have an effect on how the CPU works?
>>54144303
They also make the chipsets those motherboards use.
>>54144221
>implying
I am still using a Q9-550
>>54144100
>implying quantum cpus aren't going to be available soon
lol dumbass
>>54144258
well there was just no demand for higher power. In serverfarms though, cpus are officially the bottleneck. There was a really good article about it let me see if I can find it
>>54144100
>bought an i5-750 in 2009
>still runs single threaded applications better than anything AMD
>i literally have no reason to upgrade
>>54144276
Regular consumers don't upgrades very single cycle. Its usually every 4 or 5 years, and you want all new hardware at that point.
Also, motherboards are very different from what we had in 2006, where we had the old CPU->FSB->NB->SB; most of the 'features' on a modern mobo are purely physical links to the CPU, all the actual features are ON the cpu package itself, and the mobos simply facilitate physically connecting hardware to the CPU, and power distribution.
>>54144100
All true, but I didn't realize how important heat output is... until I had a 7870XT in my 10x10 room in the middle of summer.
I left FC3 open, came home from the bar after a few hours and my room was literally 130 degrees. Now I try to balance between power and low heat, but I caved in and got a MSI 390 Gaming so I'll probably be getting a Portable Air Conditioner for my room this summer.
It's about a 275w card according to most reviews (like Nvidia shills, to an extent) and I'm hoping I'll be able to undervolt it down to around 190-200w. From the quick googling I've done it seems possible. All AMD chips I've had have been very conservative on stock voltage and have had tons of room to undervolt and lower heat.
Also my CPU is a A10-7850k and I'm trying to decide how far to OC it. I'm leaning towards 4.2GHZ for low TDP but 4.6 is tempting. I was really happy with my FX-8350 at 4.2 but the 7850k obviously only has 4 'cores'.
>>54144100
>Literally no reason to upgrade
I'd love to know how you would do passthrough on a non x79 Sandy bridge to a VM.
>p-p-p-performance is the only thing that matters
>>54144100
Dat soldered heatspreader.
>>54144100
is the i5 sandy bridge 2500k still good?
>>54144618
Jokes on you m8, I have a 3930k. I have the most future proof of the sandy empire :^)
>>54144100
I work in data mining and it really shows and upgraded a month a go to an i7 skylake. Solid 50% improvement, at least for me.
I bought the i5-2500k on release and I expect it to last me until 2020 at least. CPU technology has stalled and games graphics are stuck in the past thanks to consoles, though I rarely play games these days anyway.
Athlon XP would still be mostly fine for my purposes if not for SSE2 and Windows 7 not having support for hardware accelerated audio.
>>54144228
Because the memory controller is part of the CPU now.
>>54144618
>pci-e passthrough on sandy bridge-e
Not happening brah...
VT-d is broken is SB-E (even though it's supposed to be there and work)
x79 is fine though, just move to IB-E or Haswell-EX (protip Haswell-EX is still 2011-1)
The Oculus rift and Vive states recommended minimum is i5-4590
>>54146121
>listening to recommended system req's
kek m8
>>54144221
I remember back when I brought my AM3 board... years ago... Then they made AM3+. And all the new processors used AM3+. AMD made it so most AM3 mobos could upgrade to support AM3+ with just one bios/firmware update.
AMD's most powerful processors still uses AM3+ even tho they came out with FM2 and FM2+. But the FM ones are basically designed for APUs. Funny thing is.. most benchmark websites these days don't even bother measuring AM3+ processors for vidya, they're focused only on FM2 shit. Then they compare that to Intel's i7.
True, there's really no reason whatsoever for most people to upgrade their CPUs at this point.
5 years of improvement and you'll get about 30% of more processing power. That's ridiculous.
Desktop market is going to be fucking dead in few short years if this continues.
The future is in the mobile market for the consumer grade stuff, that's where people actually buy new shit on a yearly basis.
I'm pretty sure that soon you'll have to look for the enterprise level stuff if you want any real change in your processing power.
We've already seen this shit happen on the SSD market.
The professional world is already hitting 16TB on a 2.5" form factor and the consumer market still keeps on pumping out new products, that start from 128GB and often cap out at 1TB.
>>54144100
I run my 2500k at 5GHz daily. Even if I'm just browsing the internet.
Considering how good the i3 in 6th gen was I'm considering upgrading to a new i3 every 2 - 3 years to keep maximum performance for what I do - this means 100$ every 3 years or even less.
>>54146310
Nice electricity sink.
>>54144165
>tfw P67 mobo still going strong
thanks based MSI
>>54146327
My room is consistently 8-10 degrees warmer than the rest of the house.
Almost 1.5V.
>>54144100
Even better: I'm still using two Nehalem based processors in one of the last sane Mac Pro that Apple build.
Still powerful as fuck. And fast. Even considering that this machine doesn't have an SSD.
I have an i5-3570k don't see any reason to upgrade for a long time.
Waiting for new cards to upgrade GPU though.
>>54146164
>most benchmark websites these days don't even bother measuring AM3+ processors for vidya
Because every casual will know Vishera cpus are fine for everything except 4k gayming.
emulation.
But that's mainly because most emulators are retarded abusive resource whores who are just dating for your cpu and gpu cycles.
>>54144100
>5 years later
>Not 7 years later
I only upgraded from my i7 920 because the stupid ROG motherboard shat itself
Ended up with a i5 4690 16gb
asus z97mplus
Considering its one of the fastest gaming cpu's on the market its still blazing fast 3 years later.
I just wish i got a better GPU
>>54144384
SOON
O
O
N
>>54144133
Haha exact same here.
Nehalem clocked up to "only" 3.5GHz, on an awesome X58 motherboard that has SATA III and can read PCIe SSDs. Upgraded last year to 24GB of DDR3 and a new GPU, and I do work that involves a half-dozen VMs running while I occasionally break to watch YouTube or fuck around on 4chen.
I'm interested in Broadwell-E but in my heart of hearts I know I don't need it. It would just be unjustified, mindless consumerism. I mean, I guess with a slightly-higher base clock my test suites would run... a few seconds faster? Maybe? Is that worth dropping a grand on an X99 system? I could take an entire vacation with that money and that sounds nicer when all is said and done.
>>54144100
8 years later and the only reason to upgrade would be playing Witcher 3.
>>54144133
Lynnfield here, still going strong. Sandy bridge a shit, i7 870 still kicks major ass
>>54146368
Actually vishera is better than Intel for 4k gaming according to tweaktown
>>54147160
HaHaHaHa
>>54144100
Nice meme.
Upgraded from an i7 2600k to a 6700k last year and its well worth it. The difference in applications is huge, and even in gaming, the minimum framerates improved shit tons.
Sandy was great but its not the juggernaut some o you make it out to be anymore
>>54147163
HaHaHaHa
8 year old core2 e8500
too fast so I use my 11 year old pentium M
>>54144100
It's not that that CPU is time-proof, you cuck. It's only that CPUs are going towards a wall in their evolution and the ones we're using right now are most likely gonna last forever. The actual clock speed of 4 GHz or a bit more is close to speed of light, and the 14 nm is using so few silicon molecules for transistor that we're most likely unable to go further.
The only hope for future evolution of CPUs is quantic computers, or just putting more and more cores together.
>>54147309
why would a quantum computer be any faster than a normal one?
you are getting meme'd
Eh, I'm writing a path tracer and I would have really loved to have a new skylake i7. The limitations of my 4.3GHz 2500k are really showing.
>>54147309
Isn't there a single atom transistor already in existence?
>>54147373
Study some physics man.
I'm not gona ive you a lesson so read here: http://computer.howstuffworks.com/quantum-computer.htm
>>54147173
Whatever makes you sleep at night.
>>54147309
What about just increasing core die sizes/building huge caches?
>>54147309
>actual clock speed of 4 GHz or a bit more is close to speed of light
>>54147383
Yep, but they're still quite slow. The best one right now spends like 20 minutes to do a simple factorization. When they'll evolve they can be the real deal in evolution of CPUs.
>>54147408
To be accurate, it's a bit more of 8 Ghz: http://hwbot.org/benchmark/cpu_frequency/
Which are most likely never gonna be reached for everyday use because of power consumption and heating issues. But it's still a very pathetic increase, if we think that in the past clock speed was doubled each 18 months.
>>54147406
That's the simple solution FOR USERS, however lots of people would lose their jobs at AMD/Intel this way, since there would be almost no more R&D work, only putting more old stuff together.
>>54147406
I think it'd reduce yields. And silicon is expensive. Maybe.
>>54145908
Owner here.
4.5Ghz and my Gpu (280x) is still the main bottleneck 70% of the time.
Games like planetside 2 and Mount and blade (modded) stress the CPU but it's still swapping back and forth.
And I can still overclock it to 5.4 comfortably.
It has a few years left in it.
>>54146164
And I'm sitting here with my FM1 a8-3850 Llano APU
>meme'd
>>54147132
And Lost Planet.
I went from an E6600 to a Q6600 because I knew games would start using more cores and even if they didn't I'd have two cores doing the background tasks.
http://www.xbitlabs.com/articles/cpu/display/core2quad-q6600_8.html
>>54147479
Is it worth getting one second hand? I'd love a cheap cpu upgrade because I'm now running this:
>>54147492
>>54144165
the only thing i dread
i have p65 motherboard with 30k hours in it.
Probably going to upgrade to zen or new intel at the end of the year just because of fear to be left without pc at the most inconvenient moment.
>>54147566
oh and i have magic i72600 non K that boosts to 3.9Ghz out of box without ability to OC
>>54147309
Photonic computers are the next step on the evolution of consumer grade computers, not quantic.
http://news.berkeley.edu/2015/12/23/electronic-photonic-microprocessor-chip/
And easily the most importan line from that article regardin this tech:
>They fabricated the microprocessor in a foundry that mass-produces high-performance computer chips
>Proving that their design can be easily and quickly scaled up for commercial production.
>>54147465
Shit is basically under yo feet anon
>>54147632
when i read paper on it 4 years ago i though it will take them 20 years to implement commercially
things sure move fast these days, now i hope we will get commercially available 300billion years living crystal discs for data just as fast
>>54144100
I agree. Great little CPU that 2500k/2600k is.
Unfortunately, my dumbass bought an i7 930 only a couple months before SB came out. I didn't know any better at the time, but looking back, wow, that really fucked me.
>>54147309
>The actual clock speed of 4 GHz or a bit more is close to speed of light
Ahahaha, you have no idea what you're talking about. Opinion Descarted.
>>54147659
Tech moves insanely fast when it needs to. It's pretty much just a matter of resources in the end.
After all, now we're hitting a point where the current CPU tech is becoming to the end of it's road, and an alternative NEEDS to be found.
Otherwise the huge companies are going to be fucked.
Without a doubt they're going to be throwing more and more money and brains at these projects in the very near future and this tech will advance way faster than anyone had even anticipated.
>>54147406
that's what they did back in the day, it was eating megawatts of power
>>54147687
RIP your purchase.
Why do people forget the 2700K?
>>54147696
Alternative techs exist, but conservative and stubborn companies have always gone for the “invest $X to increase our existing tech by a bit” option of investing in new technologies that will radically change computing.
Now our existing tech has hits its limits, and instead of researching new tech that they left by the roadside 10-20 years ago they have to scramble to save face and figure out how to get out of the hole they've dug themselves into.
>>54147696
>After all... an alternative NEEDS to be found.
>"It has to happen RIGHT NOW because REASONS."
I don't know where you got your MBA from you doofus, but Tech R&D isn't just some transmogrifier that you throw dollars into and out comes magic pixie dust. If you want a snowballs chance in hell of understanding this quantum compute business (rather than shoving your foot further into your mouth) here's a place to start:
http://www.scottaaronson.com/blog/
For my part, I'll bet we end up with Gallium-derived processors well before we get the whole Quantum Computing thing worked out and refined to a level that it's useful. >>54147659 is probably not far off when he quoted the 20 years figure.
And EVEN IF both of them appeared instantaneously, right now, fully-formed as though they were Athena from Zeus' forehead, we would not know how to meaningfully program 99% of consumer applications (i.e. the OS and Browser you are staring at right now) for a quantum processor. We would, on the other hand, be able to directly port code over and keep chuggin' along on a microprocessor made of indium gallium arsenide transistors.
>>54144100
I have an i5-2500k that I've never overclocked and it still works fine after 5+ years. I got to a point where I realised I don't need to overclock (so yes, I wasted money on the K-model) but it is probably for the best since OC'ing reduces lifetime anyway.
i5-2500k and Radeon 7950 for about 6 years now. Literally no reason to upgrade.
I built this computer last year with parts from a friend and about $200. all the components are old as fuck.
I'm gonna give this to my little brother.
Can you guys recommend a build to me that will blow my mind?
I can probably invest in the $1000 dollar range, give or take.
>>54145908
clock it to 4.5ghz+ and you're good to go
>>54147788
Are you even aware of the current situation?
The desktop market has been steadily tanking for years, mainly because there hasn't been a real reason for people to upgrade in ages.
The performance increases are fucking pathetic.
Also it has to happen in the very near future, because where the hell are they going to go from here?
If they delay any further, they're going to be in a pretty damn tight spot.
Also technological process is heavily driven by the amount of resources you throw at it.
You give some research department 1000% bigger budget and you can bet your ass that you'll see faster results.
And where the hell did you get "quantum" when I was talking about photonic processors?
>>54147788
Quantum computing is a meme and a red herring, with virtually no applications aside from breaking crypto.
>>54144487
You could just make a lower clocked , under-volted profile for usual shit 390 is more than plenty for most of stuff even if you under-clock/under-volt it. Also power-consumption droops dramatically at around 850 mhz core clock and some -50 core voltage and - 25 power limit. 130w max in dark souls III at 2k60fps all high (except blur), textures max.
If I want to go 4k tho I just turn it into a housefire by going 1200/1420 + 60 volts on core and +%30 power limit
But bye bye silence if I want it literary not to crash. Also +250w at all times
>>54147894
>2k60fps
Why not just write 1080p?
Is it because of people like me who will quote your reply like
>1080p
irrelevant?
Intel is shutting down desktop VCPU production in the long run
https://www.wired.com/2016/04/intel-smart-declare-cloud-company-now/
But in any case as has been said CPU advancements outside of Xeons have been limited.
GPUs have gone leaps and bounds, CPUs have remained much the same
>>54147914
He's probably one of those retards that think 1440p and 1600p are 2k, when they are closer to 3K than 2K.
Still, people calling 2160P 4K needs to realize that you need 4000 or 4096 as it really should be to be called 4K.
UHD is a standard used for 3840*2160 for a reason.
>>54147926
>4K needs to realize that you need 4000 or 4096 as it really should be to be called 4K.
4096x2160 smugface reporting in
True 4K master race, plebeians need not apply
>>54147914
I meant 2560 x 1440 by 2k.
>>54147940
>plebeians need not apply
They shouldn't even try.
>>54147950
read >>54147926
>>54147950
That's 1440p, anon.
2K is 2048x1080 (or sometimes 1920x1080), just as 4K is 4096x2160 (or sometimes 3840x2160).
>>54147960
This.
Also proper 4K is either 4096*2304 or 4096*2560 if you wanna use 16:9 and 16:10
>>54147960
Fair enough I guess, I'll just go with horizontal pixels x vertical pixels form now on.
The point is Hawaii can be cool and silent if you under-clock it when full power isn't needed
>>54147990
The A series are known for being low wattage which is perfect for HTPC ad with a much faster iGPU they don't really need a graphic card for most non-gamers.
>>54144304
Yes and yes.
2600k here, 4 cores 8 threads with hyperthreading... What else do I need?
Great performance so far after 5 years of use... I spect another 2 or 3 more...so it is worthy to spend money in a good processor
>>54147980
>Also proper 4K is either 4096*2304 or 4096*2560 if you wanna use 16:9 and 16:10
No, proper 4K is exactly 4096*2160. You don't use 16:9 or 16:10 when dealing with 4K footage.
>>54148131
Newer instructions, better support for native USB3 and 3.1 and support for NVMe.
Faster IPC and much more.
BUT, the thing is do YOU need it?
>>54147980
>>54148161
Pic related also top kek at that last sentence
>>54147960
>>54147980
>>54148161
4k is a marketing term for 3840 × 2160 so just dial back your autism and get used to plebs calling it that.
>>54148174
Wiki is a good source for info but not always true.
4K 16:9 is 4096*2304 and it was part of the 4K wiki in the past as a standard resolution.
https://en.wikipedia.org/wiki/DVD%2BR
>A DVD+R is a DVD that can be written once and read arbitrarily many times.
Doesn't even bring up Multisession that allows DVD+R to be written to multiple times as long you have free space to write to the DVD.
- NTSC 4:3 standard: 640 x 480 resolution = 307,200 pixels at 4:3 aspect ratio.
- PAL/SECAM 4:3 standard: 768 x 576 resolution = 442,368 pixels at 4:3 aspect ratio.
- NTSC standard for DVD: 720 x 480 resolution = 345,600 pixels at 3:2 aspect ratio.
- PAL/SECAM standard for DVD: 720 x 576 resolution = 414,720 pixels at 5:4 aspect ratio.
- NTSC 16:9 standard: 854 x 480 resolution = 409,920 pixels; aka 848 x 480 resolution = 407,040 pixels at about 16:9 aspect ratio.
- PAL/SECAM 16:9 standard: 1024 x 576 resolution = 589,824 pixels at 16:9 aspect ratio.
- High Definition (HD) for Blu-ray: 1280 x 720 resolution = 921,600 pixels at 16:9 aspect ratio.
- Full High Definition (FHD) for Blu-ray: 1920 x 1080 resolution = 2,073,600 pixels at 16:9 aspect ratio.
- 4K Ultra High Definition (4K UHD): 3840 x 2160 resolution = 8,294,400 pixels at 16:9 aspect ratio.
- Full 4K Ultra High Definition (4K UHD): 4096 x 2304 resolution = 9,437,184 pixels at 16:9 aspect ratio.
- 8K Ultra High Definition (8K UHD): 7680 x 4320 resolution = 33,177,600 pixels at 16:9 aspect ratio.
- Full 8K Ultra High Definition (8K UHD): 8192 x 4608 resolution = 37,748,736 pixels at 16:9 aspect ratio.
>>54148020
I clearly presented a gaymen scenario here with under-clocked 390 driving dark souls III at 2560x1440 at 60fps.
Thanks for the contribution tho.
>>54148234
No problem.
I know OP is shitting through his mouth because I have 60FPS constantly with my GTX 3.5GB and with a stock 4690K.
>>54148162
Even if he did buy new comp, 2600K is fucking beastmode for a secondary computer.
Mine is two feet away right now, ready to go in case the main computer goes BÖRK or I want to do LAN with a buddy.
Though I may have to get a new case for it, since the old cheapo case can't fit real videocards.
>>54148334
Get a Fractal S of you want a cheap case that you're just gonna slap a single SSD in.
>>54147282
I had an E8400 clocked @ 3.6GHz until just recently. That processor ran like a champ.
On an 8350 here, im having 0 problems whatsoever, its amazing how little the cpu seems to matter now (outside of mobile and atom)
>>54144221
AMD products age really well indeed.
Look at the Phenom II x6 for example, released in 2010 but still on par with their current offerings!
>>54144133
If wonder if they overclock as badly as the normal desktop Broadwells. They can barely reach 4.2 Ghz
>>54144165
Get a Maximus V Formula or something. I'm not a fan of these gaymer designs but this is a decent board and it only cost me 80 €.
Not sure if my 3770K will live another 5 years, but at least until 2018. My next CPU will be 8 core with PCIe4.
>>54147882
>computing is a meme and a red herring, with virtually no applications aside from breaking crypto.
- guy ca. 1980
>>54144165
Get Skylake you disgusting fuck
>>54149688
No get used non-Skylake.
Suck my overclocked Ivy Bridge.
>>54144304
Guys, am i fine with a I7 4771?
>>54144100
>2600k
>"future-proof"
>"with overclocking you're set for another 5 years"
Kek. Even the Skylake's i3 is comparable to an overclocked 2600k.
>>54149894
In which dimension?
http://cpu.userbenchmark.com/Compare/Intel-Core-i7-2600K-vs-Intel-Core-i3-6100/621vs3511
That would mean the i3 had more than 100 % more IPC.
I'm pretty much set for a while.
>>54144100
amd x6 1055 t here w/ ati hd6870gpu.
All i do is music production with reaper, which takes full advantage of multi threading, and the occasional round of STALKER.
tell me why i should upgrade
