[ARTICLE] Every Core i5 generation benchmarked from Lynnfield to Skylake

Random Guy · 21 Apr 2016 at 08:02

http://www.hardwareunboxed.com/ever...-lynnfield-to-skylake-7-years-of-i5-goodness/

Super information for anyone who is pondering an upgrade for purely FPS-related issues, and extremely interesting when DX12 is brought into the mix with the conclusion being that more efficient use of the CPU will only prolong the lifetime of the older generations.

Sure, platform upgrades are worth considering for other reasons, but the results here are empirical proof (unless someone wishes to disagree) that may be interesting for people wanting to swap out their older CPUs.

I know that most who have gone on to Skylake are actually very happy with the move so this is not to take away from their decisions but more to have a baseline (something I am admittedly obsessed with) to make a choice.

David Bisset · 21 Apr 2016 at 12:21

Nice set of reviews.

To be fair, I may be still using a Lynnfiled CPU but that's slightly flattering for them as they'll not go far past 4GHz while Sandy would easily go a fair bit faster, but it still copes fine with almost all current games.

Freddie1980 · 21 Apr 2016 at 13:20

The guy have been bored stiff doing those benchmarks. 90% of those results you could only tell you were running an older CPU if some told you or you looked at the FPS counter.

Nexus18 · 21 Apr 2016 at 15:16

May my i5 750 live a long and prosperous life

The only way I will replace this beast is when it gives out!

Absolutely no point spending £350+ when the upgrade will be minor in the majority of games.

Would have been nice to see results for the division though.

Orcvader · 21 Apr 2016 at 16:05

I was just speed scrolling past the first half as those games were mostly GPU bound - and then I started to pay attention when I saw the ARMA 3 section

.

... I didn't realise the jump from Broadwell to Skylake was that small compared to other tocks.

Besty · 21 Apr 2016 at 16:49

it was a waste of time doing the all the benches with max settings for many of us as in competiive play max settings are rarely used.

what we really need to see is the benches run in same resolution but lowest detail.

and skylake using 2666 ram? why would you run it so slowly? ddr 3400 would have been the absolute minimum for me, 3800 being optimal.

abui1806 · 21 Apr 2016 at 17:15

Besty said:
it was a waste of time doing the all the benches with max settings for many of us as in competiive play max settings are rarely used.

what we really need to see is the benches run in same resolution but lowest detail.

and skylake using 2666 ram? why would you run it so slowly? ddr 3400 would have been the absolute minimum for me, 3800 being optimal.

I have never heard of competitive The Witcher 3 or GTA 4 before but it sounds amazing.

breadturbo · 21 Apr 2016 at 18:08

Should have done the test with two gpu's, one isn't going to tax an AMD(lul) let alone an i5.

abui1806 · 21 Apr 2016 at 18:34

breadturbo said:
Should have done the test with two gpu's, one isn't going to tax an AMD(lul) let alone an i5.

they did CF R9 390X.

d_brennen · 21 Apr 2016 at 18:38

abui1806 said:
I have never heard of competitive The Witcher 3 or GTA 4 before but it sounds amazing.

LOL quite

The point of ramping settings up is to expose more weakness. Who cares if you get 198 or 199 FPS minimum.

Besty · 21 Apr 2016 at 19:49

abui1806 said:
I have never heard of competitive The Witcher 3 or GTA 4 before but it sounds amazing.

Pls, don't bait.

The notion that Battlefield 4 and Black Ops players play in max detail online is ridiculous.

60fps Battlefield 4/Black-Ops is of no interest to anyone who wants to actually win on-line against guys who are running at 120fps or 144fps in low-detail.

Jarrastafari · 22 Apr 2016 at 01:42

Hz crutch. I'm no pro (who is really) but to play competitively in BF4 you need stratergies not visible frames. 144Hz saved no-one from a well executed flank or an RPG in their cockpit...

Respect to this guy for putting it together! Even included ARMA for the critics who thought GTA4 wasn't harshe enough, heh.

Purgatory · 22 Apr 2016 at 04:28

He also did this one too for all that missed it :-

5 Generations of Core i7 Processors: 2600K, 3770K, 4770K, 5775C & 6700K Gaming Comparison

Every Core i5 Generation Benchmarked: Lynnfield to Skylake, 7 years of i5 Goodness!

AaA91 · 22 Apr 2016 at 10:07

Lynnfield still shows strong numbers even 8 years later

. Lynnfield CPUs in 2016 are about the same age as the early Willamette Pentium 4s in 2008.

eddyr · 22 Apr 2016 at 10:17

Interesting results.

Overshadowed by the way he places the cpu into the socket at the end ~15.50.
*shivers*

Nexus18 · 22 Apr 2016 at 10:50

My i5 750 and 60HZ IPS screen certainly doesn't stop me from doing well in FPS shooters like BF 4

A good workman never blames his tools

los · 22 Apr 2016 at 13:18

Shame they didn't include 4790k, 6700k and 5820k.
Great to see an actual real world usage in games though!

Caracus2k · 22 Apr 2016 at 17:18

Besty said:
Skylake using 2666 ram? why would you run it so slowly? ddr 3400 would have been the absolute minimum for me, 3800 being optimal.

Despite repeated tests showing basically no significant improvement for pretty much any games and applications when using faster memory modules?

See here and here and here for example....

Buying uber quick (and expensive) DDR4 for Skylake is largely pointless unless you value a memory bandwidth score over everything else and if that's the case you should be using an X99 setup anyway which can give you even more (likely unused) memory bandwidth with quad channel memory!

'Overall, we've found that DDR4-2666 memory is definitely the optimal choice in terms of performance and value on both the X99 and Z170 platforms, and while faster memory will benefit the Z170 platform, it comes at a cost that likely isn't worth it for gamers when the money could go to other components like CPUs and video cards.'

Back to the OP...

Lynfield to Sandybridge was a big gain but its certainly been small increments since then!

Randomface74 · 22 Apr 2016 at 17:21

i7 920 for life

abui1806 · 22 Apr 2016 at 18:02

Caracus2k said:
Despite repeated tests showing basically no significant improvement for pretty much any games and applications when using faster memory modules?

See here and here and here for example....

Buying uber quick (and expensive) DDR4 for Skylake is largely pointless unless you value a memory bandwidth score over everything else and if that's the case you should be using an X99 setup anyway which can give you even more (likely unused) memory bandwidth with quad channel memory!

'Overall, we've found that DDR4-2666 memory is definitely the optimal choice in terms of performance and value on both the X99 and Z170 platforms, and while faster memory will benefit the Z170 platform, it comes at a cost that likely isn't worth it for gamers when the money could go to other components like CPUs and video cards.'

Back to the OP...

Lynfield to Sandybridge was a big gain but its certainly been small increments since then!

why would he care? I already got the 3600Mhz memory as he wanted.