i7-2600K vs. i7-8700K - Is Upgrading Worthwhile?

[SKYMTL]

1440P Gaming: GTA V / Overwatch

Grand Theft Auto V

In GTA V we take a simple approach to benchmarking: the in-game benchmark tool is used. However, due to the randomness within the game itself, only the last sequence is actually used since it best represents gameplay mechanics.

Overwatch

Overwatch happens to be one of the most popular games around right now and while it isn’t particularly stressful upon a system’s resources, its Epic setting can provide a decent workout for all but the highest end GPUs. In order to eliminate as much variability as possible, for this benchmark we use a simple “offline” Bot Match so performance isn’t affected by outside factors like ping times and network latency.

As you recall, the i7-8700K ended up being quite a bit faster than the i7-2600K within these two games at 1080P. While it still leads, the delta between these two processors has narrowed substantially, to the point where its almost impossible to detect any visible difference in-game.

 

[SKYMTL]

1440P Gaming: Ghost Recon / Witcher 3

Ghost Recon Wildlands

While the latest Ghost Recon game isn’t compatible with DX12, it happens to be one of the best looking games released in quite some time. It also has an extensive set of in-game graphics options. This 90-second benchmark is based in the tropical jungle of Espiritu Santo as well as a vehicle drive into a slightly more arid zone. As with some other games, the in-game benchmark on this one is out to lunch and doesn’t give a good representation of what you can expect within gameplay.

Witcher 3

Other than being one of 2015’s most highly regarded games, The Witcher 3 also happens to be one of the most visually stunning as well. This benchmark sequence has us riding through a town and running through the woods; two elements that will likely take up the vast majority of in-game time.

As we close out the gaming results, nothing has really changed with each CPU’s positioning relative to one another. Basically, if you have a Sandy Bridge system and want to get a bit more distance out of it for gaming, look towards your video card rather than a full system refresh.

 

[SKYMTL]

Moving Further On With Overclocking

Moving Further On With Overclocking

Up to this point the task of finding distinctive performance differences between the i7-8700K and i7-2600K ranged from easy to nearly impossible. However, both offer some pretty substantial overclocking chops and as a matter of fact our i7-2600K sample is one of the best CPU’s we’ve had in relation to how far it can overclock relative to Base Clock. Let’s quickly take a look at what we were able to achieve and remember: these clock speeds were tested with an intensive 6 hour stability test so I’m confident they are 24/7 stable.

While the 7-8700K hit 5.145GHz on all cores with a memory speed of about 3600MHz, I think the real star of this particular show is that little gem of a 2600K. It settled at a final speed of almost 4.9GHz without batting an eyelash. As a matter of fact, I was able to push things to almost 5.1GHz with some judicious BCLK adjustments but it ended up consistently failing our hour long 3DMark Time Spy loop so things needed to be scaled back a bit. Meanwhile, its memory ended up running at just over 2300MHz which is also a pretty good result considering Sandy Bridge is rated for 1333MHz operation.

But will these clock speed really lead to the i7-8700K running further afield in general tasks and could an OC’d Coffee Lake CPU also stretch its legs in gaming? Let’s find out.

In both Adobe Premier and Handbrake the jump in clock speeds results in much better performance for the i7-2600K but it still doesn’t come anywhere close to beating a stock i7-8700K. The 8700K doesn’t really extend its lead by all that much in Handbrake but that extra clock speed does wonders in Premier.

Switching to gaming and the results are pretty much the same as before since the graphics card ended up being a bottleneck in all of the cases. The GTX 1070 is without a doubt a very good option for a relatively inexpensive upgrade for Sandy Bridge users but I have a funny feeling that performance differentiation may have been possible with a slightly faster graphics card. I’ll get into that on the next pages.

Despite overclocking not doing all that much for average gameplay performance, it did highlight one fact: the i7-2600K’s in-game fluidity can be improved with some clock speed adjustments. What you can see here is that the overclock improved the 2600K’s 99th percentile frametimes in Overwatch, bringing its overall performance closer to the 8700K.

 

[SKYMTL]

GTX 1080 Ti Gaming Results (1080P)

GTX 1080 Ti Gaming Results (1080P)

By this point in the article I am sure you are wondering about what would happen if these systems were paired up with a faster graphics card. Indeed, in the last few pages it was extremely challenging to distinguish one system from the other.

But let’s assume for a moment that someone who paid big bucks for an i7-2600K and GTX 680 wants an in-place upgrade that will bring them back to the forefront of modern gaming systems. In that case, a GTX 1070 wouldn’t be the cream of the crop. That distinction falls to NVIDIA’s GTX 1080 Ti and after seeing the results on the past pages, I just had to see if adding one would allow Coffee Lake to finally stretch its legs. That task fell to the ASUS GTX 1080 Ti STRIX OC Edition.

Let’s start with the full batch of 1080P benchmarks since I don’t want this article to drag on overly long. While the exact same benchmarks showed the systems neck and neck with a GTX 1070, the higher end (and MUCH more expensive) GTX 1080 Ti somewhat eliminates the GPU bottleneck. Consequently, the i7-8700K pulls out ahead in almost every single test, sometimes by massive amount.

Another interesting thing that came up was performance beyond averages. Based on 99th percentile results it looks like the Sandy Bridge system just can’t deliver as consistently smooth gameplay, or at least not as smooth as Coffee Lake. This proves that you can’t always rely on average framerates to determine overall performance.

With all of that being said, at 1080P the 2600K does provide very good performance when paired with the GTX 1080 Ti. But I have to ask the question: how many folks will want to spend over $600 on a graphics card upgrade and limit their new system to such a low resolution? Well that’s what the next page will address.

 

[SKYMTL]

GTX 1080 Ti Gaming Results (1440P)

GTX 1080 Ti Gaming Results (1440P)

Moving on to 1440P puts additional stress upon the graphics card and I was thinking that the gap between Coffee Lake and Sandy Bridge would narrow somewhat. Did that actually happen….well let’s see.

It looks like my thought process was pretty much spot on with a few minor exceptions; the 2600K and 8700K once again run neck and neck when the onus is put upon the graphics card. Expect these differences to become even narrower as resolution / detail levels increase at 3440x1440 and 4K.

As for those exceptions to the rule, well they come from the usual culprits; those being 99th percentile framerates in DX11 applications. And then there’s Grand Theft Auto, at truly odd performer that seems to have been updated to predominantly favor Intel’s newest architectures.

 

[SKYMTL]

Conclusion - Sandy Bridge Still Holds It's Own

Conclusion - Sandy Bridge Still Holds It's Own

I had originally planned on spending a few days benchmarking and writing this article but the dreaded specter of mission creep ended up determining the timeline more than my intents did. It started with curiosity about how Sandy Bridge and Coffee Lake compared against one another but then as the gaming results trickled in, I knew testing at stock settings with a GTX 1070 would only be the tip of the iceberg. And so overclocking, more resolutions and the GTX 1080 Ti ended up factoring quite heavily into the equation.

At the beginning of this article I posed a pretty simple question: but is it actually worthwhile for you to upgrade a six year old system right now? The answer is a perplexing Yes, No and Maybe. Now before you start the hate mail, let me explain why there are three potential scenarios, all of which largely depend on an individual’s situation and their wants / needs.

Let’s start things off with why Sandy Bridge users will likely want to look very closely at Coffee Lake. Not only is this the first architecture that brings a 6-core, 12-thread CPU to the sub-$400 market but the various architectural improvements over the last half decade have enhanced performance in certain applications to impressive levels. There has been absolutely nothing revolutionary introduced by Intel since the onset of their Core series, content creators and other professionals will find an upgrade to Coffee Lake hugely beneficial. Everything from rendering times to photo manipulation speeds has been increased drastically over the i7-2600K and its ilk.

The “maybe upgrade” part I talked about is for those who might not need massive performance uplifts in professional programs but rather folks who want to replace their once high end 2600K / GTX 680 gaming system with its modern equivalent. That means spending upwards of $1500 for a new processor, GTX 1080 Ti graphics card, motherboard and memory. When coupled with a powerful enough GPU, the i7-8700K was able to distinguish itself quite well from the i7-2600K, particularly if in-game fluidity is factored into the equation. More importantly, the 8700K is much better positioned to provide adequate processing bandwidth to the more powerful upcoming graphics cards whereas the i7-2600K already struggles to feed the GTX 1080 Ti.

The problem with Coffee Lake is its price to performance ratio for gaming if you are on anything resembling a reasonable budget. The i7-2600K paired up with a GTX 1070 proved to be a very, very potent combination particularly at higher resolutions where it matched up evenly against the i7-8700K. I’ll even go out on a limb and recommend avoiding Coffee Lake unless you absolutely need the additional connectivity offered on the Z370 platform. The framerate uplift just isn’t there and the money you save could easily be put towards a better GPU….and an all inclusive vacation.

Coffee Lake may be a real barn-burner for general applications but in gaming scenarios it is certainly not the best upgrade choice for your money. Right now that distinction falls to unadulterated GPU power. For a number of reasons, games just don’t need a modern processor and as holistic DX12 / Vulkan optimized games continue to roll out, that situation may become even more evident. While there is certainly a tipping point where either limited core counts or clock speeds will negatively impact framerates in a big way, it wasn’t reached within the somewhat limited nature of this article.

This all brings me to what should be an obvious bit of advice: know what your potential upgrade will achieve or you could be spending money for nothing. Given the right set of circumstances, Coffee Lake or any current generation CPU can truly shine but they are costly and a complete system overhaul may not grant the results you were looking for.