Multi-core Benchmarks

There's not enough information to say yet, but there are few programs for home users that can actually tax 8 cores at once. It's common for work to not be perfectly parallel, and whenever that's the case you almost never benefit from adding more cores after the fourth as compared to getting faster cores.

I'd expect it to run perfectly well on either processor, but I also wouldn't bet against Intel without more information.

Yes, it is an interesting question, but there are too many variables even after we get the final code (without all of the debug stuff). Variables like the GPU and any assist it can provide, and how effective that assist is to either AMD or Intel come to mind. I can well imagine that there are partnering tradeoffs to consider. Add to that the way GeForce keeps trying to improve their game performance by providing game settings that match their GPUs. (Do other GPU manufacturers do that too?)

I doubt we will have any usable answers (or even clues) until 2015, so I think it is useless to raise this question at this time.

I think there is a goal here, reading Frogboy's post and watching the video from Oxide, which mind you Galactic Civilizations III is not using but, instead using it's own internal Stardock engine.  Frogboy has said that it's designed for the multiple cores, so brings up a question did they use what they learned in building this engine to help the new engine being designed?  In that video with the sample game that they quickly designed as a test with 100k batch files to the gpu, they were seeing increased preformance with the FX 8350 in benchmark test to that same intel i7 4770k and they were saying the AMD FX may be even beating it in some cases.

So with that information, if Galactic Civilizations III uses even a little bit from that model, or uses mantel, even though I don't think it is.  This could be very exciting for those that stuck with AMD went in most cases less expensive compaired to the intel equivelint and may be seeing simular or better preformance.

This for one is why I chose AMD FX 9370 when building my computer as, with games going to multiple cores I figured they may tie more closesly then with video editing and creation and therefore keep AMD in the playing field.  Going back to your posts you're right it's still most likely to early to tell but, I would imagine the Stardock with the Pre-Alpha working models have some idea if this will hold true in Galactic Civilizations III.

Otherwise, even though when us founders get Alpha access, it would be neat to see, even though not the complete game and a lot of aspects will still be missing if someone who has access to machines with the i7 4770k, and an AMD 8 core CPU, would run a benchmark using Galactic Civilizations III just to see, even though by the time of release in 2015 things may still be dramatically different.  After all by then Intel may have an even better processor at a lower cost and AMD may have something better at or around the cost of the FX 8000-9000 series.

Note that 'performance' in a game like GalCiv3 is a very very different metric to measure than in something like BF4. You'd likely need to be looking at several factors

1) 64-bit. As noted in the Oxide talks, strategic games actually are much more memory intensive than most 'brown/grey knee high concrete wall and dust simulator shooters'. Especially given the # of objects that need to be kept track of and that many such objects/terrain/etc are generally procedurally generated in strategic games. 64-bit architectures rae going to allow you to reduce things like 'pop in'  of texture and such. It also allows for certain sub-systems that usually are relegated to lower priorities in terms of memroy allocation (music/AI/sounds/etc) to have access to lots more memory.

2) multi-core/threaded programming. This is pretty tricky even if you decide to take advantage of it. You can actually make things WORSE if you don't architect things properly as you end up 'waiting' for stuff to finish, or you have race conditions, etc.

3) In terms of FX vs Intel. It appaers that the Bulldozer technology simply hasn't materialized the benefits it was supposed to from an architectural standpoint. The Windows OS itself also hampers the FX platform due to the way ti schedules tasks. I dont' really see much differnce in the future where AMD is going to out perform an Intel. Note that I"m not saying the AMD FX is bad. It's just that it really wasn't the 'game changer' it was supposed to be, and nothing indicates that's really going to change.

4) Mantle is an entirely different thing. It's an ironic throwback to 'lets have an API layer for every single video card out there'. Mantle as an API layer is supposed to improve FPS by allowing for greater optimization of GPU rendering pipeline. While GalCiv3 might take advantage of the API by the time the game is released, to me this isn't all that critical in that FPS isn't 'really' what I give a hoot about in a game like GalCiv3.

Realistically the only 'benchmarks' people playing GalCiv3 would really care about

1) Turn time processing

2) AI processing (and this is a veyr very subjective thing)

The original post for this thread seems to be more about how competitive the two manufacturers are going to be, now that I think about it. And it expresses a dilemma all of us users face. The only answer I can think of giving is that if one is better than the other now, the next version of the weaker manufacturer is going to be aimed at reversing which is better. Especially if they wish to remain competitive. So, in the future,  expect a continuing reversal of which is better until one of them goes out of business.

Also, always expect people to have their favorites, regardless.

For general purpose home computing, Intel CPUs have been beating AMD ones in performance for a while now. It's generally better on a home computer to have faster cores rather than more, once you get to 4. More than that is wasted too often, as perfectly parallel processes are VERY rare in home user workloads (video encoding would be an exception). Intel's typically been beating AMD in raw performance per core with the i7 line.

In terms of price performance? That can be a different story. I don't think you're going to go wrong with either one, but given the way that games tend to work I would expect Intel to wind up being faster simply because a lot of stuff can't be done in parallel.

All speculation based on current and past games, of course.