FRONTPAGE Intel i7 4790K Devil's Canyon CPU Review

Woomack · Jun 12, 2014

RubyBeats said:
FX is a much more powerful series overall with a higher multi-threaded performance. APU's will always have less CPU power.

Quick browsing:
http://cpuboss.com/cpus/AMD-FX-4300-vs-AMD-A8-6600K
About the same specs except that FX has L3 cache while APU doesn't.
I can't see much higher FX performance here. FX will be faster but L3 cache is not helping in everything especially that AMD uses large but really slow cache.

Here is also 4.2GHz FX vs 3.5GHz i3:
http://cpuboss.com/cpus/Intel-Core-i3-4330-vs-AMD-FX-4350
Results are similar in most tests ( also not included in link ). When I was comparing APU to i3 then I had to set ~4.3-4.4GHz to get about the same results as 3.5GHz i3 ( depends from test ).

Also sorry for off topic.

RubyBeats · Jun 12, 2014

Woomack said:
Quick browsing:
http://cpuboss.com/cpus/AMD-FX-4300-vs-AMD-A8-6600K
About the same specs except that FX has L3 cache while APU doesn't.
I can't see much higher FX performance here. FX will be faster but L3 cache is not helping in everything especially that AMD uses large but really slow cache.

Here is also 4.2GHz FX vs 3.5GHz i3:
http://cpuboss.com/cpus/Intel-Core-i3-4330-vs-AMD-FX-4350
Results are similar in most tests ( also not included in link ). When I was comparing APU to i3 then I had to set ~4.3-4.4GHz to get about the same results as 3.5GHz i3 ( depends from test ).

Also sorry for off topic.

Actually i was talking about the FX8xxx series but nevermind, i rest my case

Lvcoyote · Jun 12, 2014

Ok guys, I have a few clock for clock comparison numbers for you. i7 4770K at 4.4 GHz on M6F Z87 board. Same memory set to same speed as DC was reviewed with.

Cinebench 15 scored 905 for the 4790K, and 900 for the 4770K. Roughly a 1% difference here.

Intel XTU scored 1118 for 4790K, and 1117 for 4770K, so basically a draw here.

wPrime 32M finished in 5.333 sec for 4790K and 5.287 sec for 4770K, pretty much a draw here too. wPrime 1024M finished in 164.47 sec for 4790K and the exact same 164.47 for 4770K.

SuperPi 1M finished in 8.159 sec for 4790K and 8.175 sec for 4770K, pretty close there. SuperPi 32M finished in 429.28 sec for 4790K and 432.53 sec, this is the biggest difference among all the comparisons, but is still within 1%.

Thanks fellas!

Ajay57 · Jun 12, 2014

Thanks for the Review Dino and the Benchmarks, so one would assume no need to rush out and buy the newer CPU then!! :shrug:

If your already using a 4770K in your present set up!

AJ.

RubyBeats · Jun 12, 2014

1% difference :rofl:

Audioaficionado · Jun 12, 2014

Thanx for the great review. :thup:

This sure makes me want to upgrade my whole system, but the kilobux plus price tag keeps me at bay LOL.

One of these days... :comp:

EarthDog · Jun 12, 2014

RubyBeats said:
1% difference

Why is that funny? It was expected as there was no intended change from what you seem to be looking for. This CPU was really made for overclockers with zero intent to improve on IPC.

Broadwell will do what you want.

Ajay57 · Jun 13, 2014

+1 for Broadwell as ED just said!! This was a correction on the Bad TIM used in the past CPU's that's all folks!

AJ.

Lvcoyote · Jun 13, 2014

EarthDog said:
Yep. The only difference in the chipset, for the most part, is the addition of M2/SATAe(xpress), DC and Broadwell support.

M7G wont run the 4770K Dino? Just drop the CPU in and go.

I would have, but I have another vid card review to do, and we're still doing those on Z87.... so I had to change it out anyway....LOL

Woomack · Jun 13, 2014

Both CPUs are the same so I wouldn't expect any performance differences clock to clock. The only difference is TIM and improved power regulation inside CPU what matters almost only while extreme overclocking. If you have good clocking 4770K then I see no point to change to 4790K ( unless you know it's some good sample ).

It doesn't change fact that I sold my pretty average 4770K some time ago so I will buy 4690K or 4790K soon anyway

xrror · Jun 13, 2014

One small feature Intel enabled vs 4770K

A small thing that's easy to miss, Intel didn't disable TSX on the 4790K. They disabled it on the 4770K (and all earlier K skus I think).

Which... yea I'll avoid my rant on how dumb it is that they disabled it for ANY haswell sku but... anyways - Transactional Synchronization eXtensions, doesn't really mean anything NOW, but MIGHT be rather important in 5 years - but that's of course if long term is a consideration for you.

Also I just noticed they "enabled" VT-d on the 4790K vs 4770K - which amazes me (xeon competition).

Just wanted to bring these small changes up, because it's more things to consider when looking at 4770K vs 4790K for a new build.

EarthDog · Jun 13, 2014

Vt-d is big for those that want to run VM's... Good point. Almost males me think these are rebadged Xeon's, lol!

Shell · Jun 13, 2014

EarthDog said:
Vt-d is big for those that want to run VM's... Good point. Almost males me think these are rebadged Xeon's, lol!

If there's any xeons with the same caps on the bottom out now or coming out, then they are.

Intel/AMD only has a few actual CPUs and a million SKUs of them.

bluezero5 · Jun 13, 2014

How well can i7 4790k OC, that's what I really want to know.

what can we expect as max freq on Air / commerical water / custom water?

as I got a crappy 4770k recently, I don't mind trying a new one.

will the mobo need upgrade though?

wingman99 · Jun 14, 2014

Lvcoyote said:
Ok guys, I have a few clock for clock comparison numbers for you. i7 4770K at 4.4 GHz on M6F Z87 board. Same memory set to same speed as DC was reviewed with.

Cinebench 15 scored 905 for the 4790K, and 900 for the 4770K. Roughly a 1% difference here.

Intel XTU scored 1118 for 4790K, and 1117 for 4770K, so basically a draw here.

wPrime 32M finished in 5.333 sec for 4790K and 5.287 sec for 4770K, pretty much a draw here too. wPrime 1024M finished in 164.47 sec for 4790K and the exact same 164.47 for 4770K.

SuperPi 1M finished in 8.159 sec for 4790K and 8.175 sec for 4770K, pretty close there. SuperPi 32M finished in 429.28 sec for 4790K and 432.53 sec, this is the biggest difference among all the comparisons, but is still within 1%.

Thanks fellas!

Thanks for doing those bench marks, some people were saying IPC was improved glade to see the proof that is not the case.

bluezero5 · Jun 14, 2014

hmmm..

so given overall overclock headroom also appears to see no improvement.

4790 = 4770 with higher base clock only? basically the same chip?

Ajay57 · Jun 14, 2014

Blue, it boils down to a small increase over the 4770K and a reduction in thermal heat from the CPU! But it depends a lot from where you are up grading from, far better to go for the 4970K.

If you are coming from so many Generations back like 2500K or a 2 core duo etc!! :shrug:

AJ.

trents · Jun 14, 2014

Lvcoyote said:
Ok guys, I have a few clock for clock comparison numbers for you. i7 4770K at 4.4 GHz on M6F Z87 board. Same memory set to same speed as DC was reviewed with.

Cinebench 15 scored 905 for the 4790K, and 900 for the 4770K. Roughly a 1% difference here.

I'm getting 807 in my Cinebench scores with my 2600k overclocked to 4.588. I think I'll stay where I'm at for now.

RubyBeats · Jun 16, 2014

bluezero5 said:
hmmm..

so given overall overclock headroom also appears to see no improvement.

4790 = 4770 with higher base clock only? basically the same chip?

This ↑↑↑↑↑↑↑↑

EarthDog · Jun 16, 2014

bluezero5 said:
so given overall overclock headroom also appears to see no improvement.

4790 = 4770 with higher base clock only? basically the same chip?

It is the same chip, just binned a bit higher, with better TIM, some added caps, and some software bumps (Vt-d is the only one that may be useful). Whoever said IPC was increased, was smoking the good stuff. There wasn't a piece of literature that mentioned IPC increases.

FRONTPAGE Intel i7 4790K Devil's Canyon CPU Review

Benching Team Leader

Registered

Overclocked Moderator, Overclockers.com Lead Edito

Member

Registered

Sparkomatic Moderator

Gulper Nozzle Co-Owner

Member

Overclocked Moderator, Overclockers.com Lead Edito

Benching Team Leader

Member

Gulper Nozzle Co-Owner

Member

Winner, Rig-o'-the-Quarter, Fourth Quarter 2012

Member

Winner, Rig-o'-the-Quarter, Fourth Quarter 2012

Member

Senior Member

Registered

Gulper Nozzle Co-Owner

Similar threads