Because it feels terrible to basically buy old tech.
The 10600k more or less existed in 2017. It has only minor improvements, mainly thermal, and no real performance gains over its 3 year old predecessor.
1 Like
this is nothing new for intel, see 6600k/7600k, 6700k/7700k, 8600k/9600k and more. That said the 8700k/10600k still matches or beats anything from Ryzen at stock and when you OC it you gain a small edge over Ryzen. For $300 its an excellent gaming CPU and you need to spend $100 more to get the 10700k.
1 Like
The absolute performance isn’t the problem. I’ve already outlined in this thread how they perform, and how I feel about the performance. (well, as Salhezra anyway, LOL)
The 8700k was compelling in 2017 because it was a 50% jump in resources over the 7700k and was the first time a 6/12 chip was available as a mainstream processor from Intel.
The 10th gen overall isn’t very impressive; it’s just more of the same with better thermals.
If I were buying now, I would absolutely get an 8/16 chip because I feel that’s the best choice moving forward. Also, since i already have a 6/12 chip, any upgrade I feel would be stupid to not go wider.
But given the 3600 is so damned cheap, it’s not hard to recommend as stop-gap for people with no machines.
The 10600k performs fine, but in 2020 it’s and old configuration, much like the 6600k/7700k 4/8 configurations of old that people complained about.
some useful information for anyone else researching CPUs for WoW (although obviously any modern CPU can handle it). I’m sure some people have already seen this but it’s from a Belgium site hardware info using an i9-7900x and ryzen 2700x w/ a Nvidia 1080ti to review WoW -BoA
| Framerates | 2C/4T | 4C/4T | 4C/8T | 6C/6T | 6C/12T | 10C/20T | Ryzen 7 2700X |
|---|---|---|---|---|---|---|---|
| 1920x1080 - medium | 190,3 fps | 197 fps | 198,1 fps | 197,7 fps | 198,8 fps | 195,4 fps | 179,9 fps |
| 1920x1080 - ultra | 45,2 fps | 44,8 fps | 45,1 fps | 44,6 fps | 46,7 fps | 46,1 fps | 45 fps |
| 2560x1440 - medium | 192,8 fps | 196,7 fps | 198,1 fps | 197,4 fps | 198,6 fps | 193,5 fps | 174,1 fps |
| 2560x1440 - ultra | 40 fps | 42,3 fps | 44,5 fps | 45,9 fps | 48,1 fps | 46,2 fps | 42,9 fps |
| 3840x2160 - medium | 188,7 fps | 193,6 fps | 183,4 fps | 187,3 fps | 192,5 fps | 185,8 fps | 166,6 fps |
| 3840x2160 - ultra | 39,6 fps | 45 fps | 46 fps | 46,7 fps | 48,2 fps | 52,1 fps | 45,1 fps |
*Also in the frame times we see that four cores lead to better results, and that is less the case at higher settings. (sic translation)
| Frametimes | 2C/4T | 4C/4T | 4C/8T | 6C/6T | 6C/12T | 10C/20T | Ryzen 7 2700X |
|---|---|---|---|---|---|---|---|
| 1920x1080 - medium | 7,4 ms | 6,3 ms | 6,2 ms | 6,3 ms | 6 ms | 6,2 ms | 6,5 ms |
| 1920x1080 - ultra | 31 ms | 29,4 ms | 27,3 ms | 27,4 ms | 26,6 ms | 26,6 ms | 28,6 ms |
| 2560x1440 - medium | 7 ms | 6,2 ms | 6,2 ms | 6,2 ms | 6,1 ms | 6,5 ms | 7,5 ms |
| 2560x1440 - ultra | 32,4 ms | 29,3 ms | 29,7 ms | 27,4 ms | 26,4 ms | 26,3 ms | 28,5 ms |
| 3840x2160 - medium | 7,4 ms | 6,6 ms | 7 ms | 6,8 ms | 6,5 ms | 6,8 ms | 8,1 ms |
| 3840x2160 - ultra | 37,4 ms | 28,6 ms | 26,8 ms | 26,8 ms | 25,6 ms | 26,1 ms | 28,2 ms |
1 Like
That’s probably an effect of ringbus getting stretched to its limit at the 10 core sample.
In 1080p medium frametime (least GPU limitation), you see a continual improvement going from 2c w/SMT (7.4) to 4c w/SMT (6.2) to 6c w/SMT (6) then a dropoff at 10 back to 6.2.
If anything, this just goes to show more cores DOES help…as long as the architecture itself isn’t gimping it.
I notice there’s no 8/16 sample here. Even if we call it even and say 6.1ms, that’s still a huge improvement over the dual core SMT sample, and not enough of a detriment to rule it out for its other benefits.
My reasoning for getting a higher core count chip is for wider applications and being a bit more versatile than just WoW, as well as looking forward to newer games.
That makes sense but from your perspective. But coming to WoW for a new build with a focus of just WoW, the 10700k is clearly overkill.
Wait, the 7900x is mesh.
Since the 10th gen mainstream boards are all ringbus, its probably different results.
Ringbus is faster core-core latency than mesh up to a point. I don’t remember at what point that is.
aaaand thats why I recommended the 3600 for them, lol. Cheap and since they’re PROBABLY playing at 60hz anyway, it won’t matter.
Not debating this point just stating what works with WoW and just WoW.
Most likely but WoW really isn’t a game that review sites test for new CPUs. TPU was the last site to routinely do it. We will have to wait for shadowlands to be released before we can see the latest CPU results.
I think any wider CPU upgrade will be good for him. With a 7600k, being a 4/4, all you need is discord, a bunch of RGB s*it like synapse or CAM, and whatever else you got going on to start hogging up resources and taking it’s toll on your gaming experience.
And looking at the results from that video I posted, any chip is going to give good minimum FPS, which is what we’re mostly concerned with when it counts, i.e. raiding etc.
Strangely enough, the one source I found that has in depth CPU testing for WoW specifically shows that even though Intel chips beat Ryzen at average FPS, the Ryzen samples trended toward better FPS lows.
I think amd ryzen is better for over all gaming and multi tasking
Ryzen is good for gaming when GPU limited. It’s generally better than Intel at everything else, though.
As it stands, in most games at 1080p with a 2080 ti, there’s as much as a 15-20% improvement in games going from a modestly overclocked 10700k over the 3900x. At 1440p+, or at 1080p with a weaker card than a 2080 ti, there’s little difference.
This is fine the time being, as rarely people use a 2080 ti and run at 1080p medium; however, it’s an indication on how these CPUs perform once a GPU bottleneck is removed. These tests mirror what a 1440p result may look like with a more powerful GPU.
Like a 3080 ti.
I predict there to be at 1440p with a new 3080 ti around a 15% performance gap between 8th/9th/10th generation Intel chips to Ryzen 3000.
Thankfully by then, Ryzen 4000 will be along and should handily beat Intel finally at gaming too.
Intel still holds the lead in overall gaming, MS office, and certain browser benchmarks. You see Ryzen do better on more work station type benchmarks. Those benchmarks don’t necessarily mean a real world difference the user will see as Salranas stated. I have both an Intel 9700k and Ryzen 2600x among other CPUs so I’m not carrying the flag for either company. From a pure gaming perspective it’s hard to go wrong with either company. It’s up to the user to decide who gives them the best performance for their dollar at the time of purchase.
Might get ryzen though i own intel always and its bad at multitasking while im playing
I think this is more to do with your specific CPU than intel in general.
Newer higher core count Intel CPUs don’t suffer this.
1 Like
Hello, looking at benchmarks and comparison tests, it seems Intel’s Core i9-10900k is the winner?.. on paper anyway, and that superiority is only apparent at lower resolutions, at 4k its a wash against Amd’s 3950X.
The 10900k’s higher frequency and lower latency are more important for gaming than the 3950X’s 6 higher core count. That’s what I read.
Ryzen 4th gen is really good. Intel 10th gen is also good. Ryzen 5 or Intel Core i5 is sufficient for WoW, although I went with the Intel Core i7.
Ryzen 9 or Core i9, however, is overkill.
It’s apparent whenever there’s no GPU limitation or when your are CPU limited.
It might not matter now at 1440p with a 2080 ti level GPU, but I expect that with the 3000 series GPUs boasting 30%+ more performance, 1440p with those cards will look like current 1080p results.
Got it. and losing it. and got it. So, right now, only at lower resolutions are video cards keeping up with, or pulling ahead of cpu’s, creating an environment in which an accurate assessment of a cpu is possible. And, while it seemed and was implied that these cpu’s became equals at higher resolutions. it isn’t at all true. I think you have the right of it.
2 Likes
It’s not a pure science, but it’s a good indicator.
People said that the i7-8700k wasn’t worth the extra money over the Ryzen 5 1600 because at 1080p with a GTX 1080, performance wasn’t that different. Many users dismissed the 720p results as “unrealistic” and ignored the 720p results.
But now that we have better GPUs there’s a 30%+ gap in performance even at 1080p.
Really, testing at lower resolutions just removes the GPU variable and allow the CPU to perform unconstrained. But it’s not always exact.
I’ll be eagerly awaiting the 1080p medium results of the new Ryzen 4000 chips when they launch.
1 Like
Get a few more cores then you need if your at 100% the system will likely feel sluggish. I was running a 2950x but dropped to a 3900x when I burned the power connector on my x399 board.