>tfw i3 6100 is better than i5 6400
YOU JUST CAN'T MAKE THIS UP
http://www.trustedreviews.com/intel-core-i5-6400-review-performance-and-verdict-page-3
>>54717761
DELETE THIS NOW!!!!
Is there even any difference from the two chips aside from the fact the i5u can turbo?
>>54717918
I5 is a gimped quad core.
I3 is a non gimped dual core with hyperthreading.
OY VEY SHUT IT DOWN
>>54717761
Holy Shit My i7 870 Is Faster Than That Shitty i5 HaHaHaHaHa
>2.7ghz
Kek
>Dual core
>for gaming
>2005+ number of years to current year = your a faggot
>>54717953
>I5 is a gimped quad core.
no shit
>>54718227
You would be surprised how capable that i3 is
>>54718247
its okay to be poor if you buy i3 :^)
>>54718307
>using the smiley with a carat nose
holy shit big surprise in games that don't need lots of threads the higher-clocked, higher/equal IPC part will do better!
You'd be an idiot to buy either of these intel chips at this time, frankly.
What about the 6500?
>>54720364
Also loses to the i3 in single core. Only the i5 6600 or 6600k beat it.
If you really need a powerful cpu on your laptop, why pick an i5? They are glorified i3 , which are usually more than enough for a laptop.
>>54720507
Why are you bringing up Laptops?
Will Kaby Lake even be any different? I heard they're literally Skylake cpus with hardware 10 bit HEVC decode/encode built or something like that. And pretty much any Skylake CPU is powerful enough to do that through software anyway.
>>54717761
>buying anything Intel other than Xeon or i7.
Poor people should be blown up.
>>54720422
>m-muh cache!
Can someone spoonfeed me if OP statement is true for laptops as well?
I'm looking to get a laptop with an i5 skylake 6200u processor. Basically this
http://www.bestbuy.com/site/acer-aspire-15-6-laptop-intel-core-i5-4gb-1tb-hard-drive-gray-black/4961000.p?id=bb4961000&skuId=4961000
>>54720573
Some of us have HTPCs that we'd like to transcode stuff with at decent speeds without using any power.
>>54720711
>Poor people should be blown up.
Cherry Trail does that for even less then...
>>54717761
Who needs 6400 when you can buy 6600k and overclock it to 4.8 ghz
>>54720775
this tbqlrh
rocking that stable 4.8ghz 24/7
>>54720569
You can use your graphics card to that anyway.
>>54720523
Because I'm stupid and thought it was about the i3-6100U, the laptop one.
>>54717761
For fucking real
That single core performance of the 6300 is up there with the big boys
>>54717761
>muh gaymin
Isn't the i3 6100 just a dual core tho? It can't play games that require quad core
>>54717761
>>54717822
He defines CPU performance as FPS in games, back to >>>/v/ you go you fucking retard
I3 3220 master race reporting in.
I run almost any game on it without a video card. Also, 4gb ram. Poor and proud for not dumping my money on Jewish apparatus.
>>54720569
>powerful enough to do that in software
HEVC encode in software is fucking dicks right now, go fuck yourself.
>>54720799
Fuck you too pal, GPGPU encoding algorithms
are really inefficient, they're fast as fuck but they produce larger files, which entirely misses the point of using HEVC in the first place.
>>54721487
i3 has 4 threads, most of gaymes treat them like a core. So, we might tell that you have retarded quad core cpu.
>>54720569
50%* performance improvement for lag-free gaming at 125%** of the cost, goy
*- marketing bullshit stats
**- not representative of the final cost, which will be higher
>>54722331
I've heard that an 880K OC'd will actually yield much better performance in DX12, but most games are still on DX11 or earlier where the i3-6100 is far better.
>There are cucks in this thread that buy locked CPUs to save a few shekels
>There are cucks in here that don't realize that these processors are literally failed versions of the unlocked processors
>They're literally buying expired cartons of milk from Intel Milk Company and crying about why it's sour
>>54722597
i3 dies and i5/i7 dies are NOT the same.
i3-43XX/63XX are the full dual core die
i3-41XX/6100 are the full dual core die with parts of their L3$ disabled
Pentiums are dual core dies with defective L3$ and HT disabled
Celerons are Pentiums that have worse voltage requirements to run at the same clock speed and even more of their L3$ disabled.
i7 are the full quad-core die
i5 are those same dies with parts of their L3$ disabled and HT turned off.
>>54722597
Overclocking is not worth it for me man I'd have to buy an aftermarket cooler, and then my electricity bill grows, and then my CPU life is lower
>>54720689
No, since mobile i5s are dual core.
Here's a rundown
atom<core m< mobile i3, i5, i7 U-series < mobile i3, i5 (even old ones) < Desktop i3 < mobile i7 quad cores (slight) < Any desktop quadcore
>MUH GAMES
eyl
>>54722646
>Disabled
>Disabled
That's just funspeak for failed the silicon performance test and got lasered off.
>>54722665
Don't overclock, buy unlocked processors and keep them for over 2 decades because you're a cuck.
>>54722597
>overclocks
>Stability test, everything is fine
>Random crashes at the worse moment
Not overclocking again, not worth it at all.
>>54722799
That's totally fine, but buying defective/locked processors are terrible
>>54722817
There's not defective with locked processors.
>>54722976
i5 6400s are clearly defective 6600ks considering they're clocked about 1 ghz lower
>>54723743
And all the perfect silicon just gets made into i7s anyway
>>54721850
>he thinks a better cpu doesn't offer better in game performance
>>54726622
no here's merely calling you a faggot for using videogames as evidence to your argument.
>>54726677
And here's me calling you a faggot for not considering video games as valid benchmarks
>>54722751
Since skylake not all mobile i5 are dual core, i.e. i5-6300HQ
>>54726724
They aren't. Performance differences in a child's toy is not a good way to measure the effectiveness of a processor.
>>54727834
There's nothing else you'd use a powerful desktop processor for. Excel and photoshop can only use so many cores. And any real enterprise grade software or database software is virtualized so it could run on a toaster with the same performance. Of course, if you had a real job you'd know that
>>54727834
You're an idiot.
>>54718227
>you + are - a + ' = you're a faggot
>>54727911
hi q
>>54727867
>There's nothing else you'd use a powerful desktop processor for.
Absolutely absurd. Why am I even taking this bait?
>Excel and photoshop can only use so many cores.
CPU isn't even the most important factor for Excel or Photoshop.
>any real enterprise grade software or database software
Well you sure covered all of your bases. I don't think there's anything else you can use a computer for in a business. Nope, surely not.
>>54727834
>thinking games aren't demanding on the cpu
What are you? 12?
>>54727867
> And any real enterprise grade software or database software is virtualized so it could run on a toaster with the same performance.
>buzzwords make my point automatically logical and valid
>>54728467
>>thinking games aren't demanding on the cpu
>What are you? 12?
I don't think that was the point he was trying to make at all.
He was talking about how most video games (especially older ones) don't properly utilize all the cores of a quad-core.
>>54722768
No, they are literally gimping some of their chips on purpose so they can sell their full featured CPUs at a higher price. It's one of those things where capitalism is completely absurd when you think about it.
>>54728696
But most newer ones do.
>gta v
>bf4
>crysis 3
>fallout 4
>thief
>the witcher 3
>anything dx12, vulkan, or mantle.
>>54728744
Well no shit, sherlock. Those are the most CPU intensive games on the market too.
>>54728744
I think the problem is more that you shouldn't benchmark CPUs with basically random software. You simply can't tell whether a CPU performs worse because the game is just programmed like shit (which gets more and more common lately) or because the CPU is really worse. That's why the i3-6100 performs better than a i5-6400: The game has just pathetic threading.
The only reason to compare CPUs by game performance is if you intend to use them specifically for gaming.
>>54728777
Therefore, these games are one good method of testing cpu performance, making his argument invalid.
Am I wrong in this? His point was that games are not a valid benchmark for CPUs.
>>54722665
>>>54722597
>Overclocking is not worth it for me man I'd have to buy an aftermarket cooler, and then my electricity bill grows, and then my CPU life is lower
Literally this. Obese neck beard buy low-end processors and overclock them to match high end processors, then crow about how much money they saved. But they don't take into account the fancy aftermarket coolers they need to buy, the higher electricity costs, and the lower lifespan of all components in their oven of a case. Would be cheaper and simpler just to buy the better processor in the first place
>>54728802
The i5 is only like what, 5 or 10% faster in multithreaded workloads? Of course even a multithread application doesn't mean the load is spread across all cores equally, that's why the i3 typically wins.
I'm an ausfag, so the price difference between the i3 and i5 is roughly $100 here. Considering they trade blows, I fail to see why the i5 even exists.
>>54728847
>The i5 is only like what, 5 or 10% faster in multithreaded workloads?
Let's say 15–20%.
>Of course even a multithread application doesn't mean the load is spread across all cores equally, that's why the i3 typically wins.
That is not really a problem of multithreading itself, but of its implementation in current games, which is piss poor generally.
>I'm an ausfag, so the price difference between the i3 and i5 is roughly $100 here.
Here in Europe it's just 50 bucks.
But I agree, for by far most people i3-6100 will be a much better deal than a i5-6400. The i3-6100 is probably the best value Intel CPU in that performance range atm anyway.
>>54728805
Those particular games are, but they are exceptions to the rule.
>>54728947
How? They're the only games still getting benched.
>>54727867
>And any real enterprise grade software or database software is virtualized so it could run on a toaster with the same performance
>he actually believes this
Must be why all the workstations at my government job are running Celerons from ten years ago instead of $10k+ machines with more hardware than your average gamer has LEDs.
>>54729365
That's one reason. Another reason is that upgrading is expensive because they'll have to replace all machines
>>54728935
>15-20%
Intel's HT isn't that good, you give the Jews too much credit. It's more like 30% and above. A logical thread post-Haswell is only about 40% as efficient as a a physical core.
>>54729384
I was being sarcastic, dumbass. The cheapest workstation we have runs about $8K, and that's because it's a back-up machine for when one of the complete towers fail.
>>54729432
Oh.
>>54729412
The i5 is 4 cores at 2.7ghz.
The i3 is 2 cores at 3.7ghz, with HT.
Considering HT usually gives 30% better performance, I can even see the i3 outperforming the i5 in many ways, and only losing when the load is scaled evenly across all 4 cores.
>>54729365
Gov sector is an entirely different ballgame. Virtualized that kind of stuff is just an unnecessary additional risk. virtualized is far cheaper than individual workstations however in terms of uptime, reliability and maintenance. but That's why I specified enterprise grade, no need to be both an ass and an idiot just because you can't read.
witness me
>>54729681
Additionally, hhs in most states typically use amd gecko boards. As do many other government facilities. Deployment cost is really your primary cost. You also must consider heat, reliability, uptime, physical access to the data and power usage as factors. Often times virtualization is the only solution. Which is why a lot of agencies rely on low power implementations. The infrastructure simply isn't there in many rural areas to run multiple high power workstations and I think a lot of people forget that.
>>54729681
All government workstations are the exact same as enterprise (except for the really exotic stuff, which is only found in departments with bigger budgets and fancier needs). You're just pulling shit out of your ass now.
Most departments can't rely on virtualized environments due to security concerns and availability. For instance, a lot of our workstations require that we physically access those machines in order to use them. Some of the programs used have no need to be virtualized since the workstations are powerful enough to handle them. A lot of places don't even have any network connections outside of buildings, sections, or even rooms, making virtualization pointless.
>>54729706
Now post multithread
>>54729801
You're literally repeating what I said. Did you not read my post? I cited a specific case of hhs offices using and gecko boards. It's very common because a constant network connection is required to do anything. So in terms of deployment cost virtualization proved to be the best route. Otherwise yes, in government applications individual workstations are used due to security concerns. I already said this. You're arguing with me when I agreed with you. Virtualization does not necessarily mean remote virtualization either. Here on the chem plant I work at, we use small celeron workstations/laptops to connect to the intranet and access a variety of virtualized software from a central location. This includes AutoCAD, SAP, primvera and office. Alongside many other applications. Thats why these applications are designed with virtualization in mind. It's cheap and low mantainence by comparison.
>>54729804
>>54729801
What may also be confusing you is me using the acronym hhs. That stands for health human services. As in food stamps. Which are typically in ghettos/rural areas. Can you imagine what would happen if they used workstations instead of virtualizing everything? Crackheads would literally have a field day.
>>54729887
>Did you not read my post
No because I'm drunk
Fuck you and the cunt you came from
>>54729916
>less than 20% faster than 3 year old laptop cpu
Kek
>>54728703
>gimping some of their chips on purpose
Only if demand exceeds their failure rate for higher binned chips.
>mfw i5 6400 @ 4.4ghz
>>54730042
didn't intel release some microde to kill all overclocks?
>>54729972
4 vs 8 hurr durr
>>54730076
ya with the newer bios revisions. im on an older bios with skyoc feature intact. its an asrock z170a, bios version: 1.4
>>54730159
The laptop is a quad core.
>>54730076
After they saw how fast the i3's performed i didn't think they would let BCLK OC live for too long
>>54730191
3.6ghz/8 threads
>>54720422
>single core
its like Im still in 2004
>>54730250
8 threads, not 8 cores.
Still interesting that it's not so far behind the latest desktop i5s from intel as I would have expected, especially considering its age.
>>54730296
It's an i7, buddy. They age well.
>>54730326
I just checked, it has a tdp of 45w and turbos to 3.6ghz, with hyperthreading.
Ina a laptop.
What the fuck? What insane lunatic would put a 45w CPU in their laptop, and pair it with a GPU?
http://ark.intel.com/products/64891/Intel-Core-i7-3720QM-Processor-6M-Cache-up-to-3_60-GHz
>>54729706
How much did you overclock?
>>54730163
@ 4.4 not 2.7 obv
>>54730502
That's good for a desktop laptop. Some people actually use laptops like this.
>>54717761
I love this guy
>>54722786
I don't buy unlocked processors
>>54730779
You shouldn't have to to be able to overclock
>>54719744
>carat
>being this acoustic
>>54718227
Implying u can use 4 cores for gaming
>>54733120
>implying your not the one whose guitarded
>>54733140
You can't?
