The Intel Core i5-7600K (91W) Review: The More Amenable Mainstream Performer
by Ian Cutress on January 3, 2017 12:01 PM ESTProfessional Performance: Windows
Agisoft Photoscan – 2D to 3D Image Manipulation: link
Agisoft Photoscan creates 3D models from 2D images, a process which is very computationally expensive. The algorithm is split into four distinct phases, and different phases of the model reconstruction require either fast memory, fast IPC, more cores, or even OpenCL compute devices to hand. Agisoft supplied us with a special version of the software to script the process, where we take 50 images of a stately home and convert it into a medium quality model. This benchmark typically takes around 15-20 minutes on a high end PC on the CPU alone, with GPUs reducing the time.
Here we report the overall time to complete the test – sub-test results can be found in Bench.
The Core i5-7600K does well in the sub-tests that are more IPC dependent, but there are more sub-tests that revolve around threads and frequency and pure single-core grunt. As a result the i5-7600K pretty much matches the i5-6600K for performance.
Cinebench R15
Cinebench is a benchmark based around Cinema 4D, and is fairly well known among enthusiasts for stressing the CPU for a provided workload. Results are given as a score, where higher is better.
All the Kaby Lake processors seem to do well in CB15 single threaded performance, given that all the K-processors can reach 4.2 GHz or higher one way or another. This puts the i5-7600K on par with Devil’s Canyon when we factor in the IPC change as well.
Common consensus is that CB15 scales with threads, however the Core i5-7600K sits above the Core i7-2600K thanks to frequency and IPC gains over the years.
HandBrake v0.9.9: link
For HandBrake, we take two videos (a 2h20 640x266 DVD rip and a 10min double UHD 3840x4320 animation short) and convert them to x264 format in an MP4 container. Results are given in terms of the frames per second processed, and HandBrake uses as many threads as possible.
Both the HandBrake tests essentially mirror what we saw in CB15 – the Core i5-7600K is there or thereabouts when frequency is the main factor, and when we stick a register-heavy threaded situation in the path, the effect from not having hyperthreading compared to the Core i7-7700K is relatively muted – in this case the i7 is only +20% performance over the i5, despite costing nearly 50% more.
Hybrid x265
Hybrid is a new benchmark, where we take a 4K 1500 frame video and convert it into an x265 format without audio. Results are given in frames per second.
Our Hybrid test is somewhat similar to the HandBrake HQ test, showing the fact that heavy threads reduce the efficacy of hyperthreads.
Facebook
Twitter
RSS
70 Comments
View All Comments
Magichands8 - Tuesday, January 3, 2017 - link
This Optane looks completely useless. But Optane DRAM sounds like it could be interesting. Depending upon how much slower it is.lopri - Wednesday, January 4, 2017 - link
I immediately thought of "AMD Memory" which AMD launched after the Bulldozer flop. But then again Intel have been going after this storage caching scheme for years now and I do not think it has taken them anywhere.smilingcrow - Tuesday, January 3, 2017 - link
Games frame rates to 2 decimal points adds nothing but making it harder to scan the numbers quickly. Enough already.1_rick - Tuesday, January 3, 2017 - link
One place where the extra threads of an i7 are useful is if you're using VMs (maybe to run a database server or something.) I've found that an i5 with 8GB can get bogged down pretty drastically just from running a VM.Meteor2 - Tuesday, January 3, 2017 - link
I kinda think that if you're running db VMs on an i7, you're doing it wrong.t.s - Wednesday, January 4, 2017 - link
or if you're Android developer using Android Studio.lopri - Wednesday, January 4, 2017 - link
That is true but as others have implied you can get 6 or 8 real core Xeons for cheaper than these new Kaby Lake chips if VM is what you need the performance of a CPU is for.ddhelmet - Tuesday, January 3, 2017 - link
Well I am more glad now that I got a Skylake. Even if I waited for this the performance increase is not worth it.Kaihekoa - Tuesday, January 3, 2017 - link
Y'all need to revise your game benchmarking analysis. At least use some current generation GPUs, post the minimum framerates, and test at 1440p. The rest of your review is exceptional, but the gaming part needs some modernization, please.lopri - Wednesday, January 4, 2017 - link
I thank the author for a clear yet thorough review. A lot of grounds are covered and the big picture of the chip's performance and features is well communicated. I agree with the author's recommendation at the end as well. I have not felt that I am missing out anything compared to i7's while running a 2500K for my gaming system, and unless you know for certain that you can take advantage of HyperThreading, spending the difference in dollars toward an SSD or a graphics card is a wiser expenditure that will provide you with better computing experience.Having said that, I am wtill not compelled to upgrade my 2500K which has been running at 4.8 GHz for years. (It does 5.0 GHz no problem but I run it at 4.8 to leave some "headroom") While I think the 7600K is barely a worthy upgrade (finally!) on its own light, but the added cost cannot be overlooked. A new motherboard, new memory, and potentially a new heat sink will quickly add to the budget, and I am not sure if it is going to be worth all the expenses that will follow.
Of course all that could be worthwhile if overclocking was fun, but Intel pretty much have killed overclocking and the overclocking community. Intentionally if I might add. Today overclocking does not give one a sense of discovery or accomplishment. Competition between friendly enthusiasts or hostile motherboard/memory vendors has disappeared. Naturally there is no accumulation or exchange of knowledge in the community, and conversations have become frustrating and vain due to lack of overclocking expertise. Only some brute force overclocking with dedicated cooling has some following, and the occasional "overclocking" topics in the forums are really a braggadocio in disguise, of which the competition underneath is really about who spent the most on their rigs with the latest blingy stuff. Needless to say those are not as exciting or illuminating as the real overclocking of the yore, and in my opinion there are better ways to spend money for such a self-gratification without the complication that often accompanies overclocking which in the end fails to impress.
Intel might have a second thought about its overclocking policies now, but just as many things Intel have done in recent years, it is too little too late. And their chips have no headroom anyway. My work system is due for an upgrade and I am probably going to pick up a couple of E2670s which will give me 16 real cores for less than $200. Why bother with the new stuff when the IPC gain is meager and there is no fun in overclocking? And contribute to Intel's revenue? Thank you but no thank you.
P.S. Sorry I meant to commend the author for the excellent (albeit redundant) review but ended up ranting about something else. Oh well, carry on..