Why is the game a laggy shitheap [SOLVED]

[jorb]

Glad you're doing better, jcoles.

[MightySheep]

awesome now all the rest of us have to do is go buy 16 GBs of RAM, thanks scooby you really solved that mystery!

[pedorlee]

awesome now all the rest of us have to do is go buy 16 GBs of RAM, thanks scooby you really solved that mystery!

+1

jajajajajajaa

[sabinati]

Welp you can add me anyways if you'd like
Steam ID is jordancolesie

Sent a request.

You should try out some of your other games as well, they should run better too.
Edit: Glad this worked out, cause for some people digging into their computer can be scary. The down side would have been if you damaged something.
Edit2: Its also funny looking at the title of the thread after it turns out it was you the whole time. Its performance is still crappy considering I can load up diablo3 at 1600p and get a steady 60fps.
ZZ

Copy of terraria sent your way, thanks for all of the help!

87 replies to this thread AND it was solved
Fucking crazy

ZeggyZon is a hero ITT

[loftar]

That's actually quite fascinating. Did I interpret the thread correctly if I conclude that configuring the memory into dual-channel mode almost doubled the client's performance?

[ZeggyZon]

That's actually quite fascinating. Did I interpret the thread correctly if I conclude that configuring the memory into dual-channel mode almost doubled the client's performance?

Yeah that totally correct and it was running at 666 instead of 800.

So he was basically running with less than half his total possible memory performance/bandwidth.

It wasn't Havens problem at all. (in his case )

ZZ

[loftar]

It's still very fascinating, since the natural conclusion to draw from that seems to be that the client is completely memory bound for performance, which honestly just sounds quite strange to me. Usually when I've seen benchmarks comparing single-channel vs. dual-channel performance, the difference has been only a few percentage points apart except for workloads that are intrinsically memory-bound, like memory-streaming benchmarks. I find it difficult to reconcile the notion that the client would be so entirely memory-bound.

If anything, though, I could imagine that video-capturing software could very well have that characteristic. Was it only with the capture program enabled that performance jumped so dramatically?

Also, as for AMD vs. nVidia, I would not be entirely surprised if the client does run better on nVidia hardware. Both I and Jorb only use nVidia, so I haven't had any good opportunity to try and tune the client for AMD cards. I've mostly been hoping that they have somewhat similar performance characteristics, but really who knows how those AMD VLIW GPUs work.

[MagicManICT]

Boosting the speed was likely the change that worked. After all, you're talking about a 17% boost in throughput.

Coles should be able to go in and explicitly set single channel mode to test. (It's an ASRock board, and these are designed for custom builders.)

[loftar]

Boosting the speed was likely the change that worked. After all, you're talking about a 17% boost in throughput.

In memory throughput, indeed, but as I said, I find it very hard to imagine the client to be completely memory-bound.

If it is indeed, then that completely changes how I'd go about optimizing it further.

[ZeggyZon]

Boosting the speed was likely the change that worked. After all, you're talking about a 17% boost in throughput.

In memory throughput, indeed, but as I said, I find it very hard to imagine the client to be completely memory-bound.

Up to a point memory bandwidth does matter. When you hit DDR3 1600 speeds and dual channel, increasing the memory speed has little impact on performance. He was so bottle necked it effected him drastically.

If he went out and bought 2400 DDR3 he would not see any real improvement. (in most games)

He would be better off overclocking his CPU or GPU to get more FPS then faster main ram for example.

ZZ