Guild Wars 2

I am using an AMD Phenom II 1100t black edition processor. In lions arch at max settings I am getting from 15-20 fps. Now I also have a GTX 650 2 gig video card used to have a Radeon 5750 (using that one in my other computer now lol). I got the same fps on both cards. I read that Intel processors show higher fps than AMD because of the hyperthreading they use. Now AMD uses hyper transport instead could that give false readings on an AMD fps number lets say 15 fps on AMD is equal to 30 fps on an Intel? They are different architectures so why wouldn’t that change also. If that is true then AMD is as good or better than Intel is.

GW2 loves raw clocks in addition to a strong single thread.
If you haven’t overclocked that black edition then I suggest that you do, as long as you have a good aftermarket cooler. There are tons of resources on the net, especially for that processor, to aid you in overclocking.

Intel has a stronger single thread performance and clock for clock will always beat out AMD in GW2.

I went from a X6 1055t OC to my i7 and noticed a massive difference in the game.

You are making assumptions that AMD is as good or better, truth is, AMD cpus have always been the budget choice. Intel obliterates AMD procs in single thread useage, hyperthreading has nothing to do with that. On top of that if you wanted to upgrade, an FX 8350 would be a side grade, so your only logical upgrade would be to an i5 or i7.

In the days of Duron and AthlonXP AMD was good to go with as they went head to head with the Pentium4, but after dual cores started making the scene Intel stepped ahead and AMD essentially fell behind. The last good AMD imo were the Socket939 Opterons, after that they went downhill compared to Intel offerings.

Well this turned into AMD bashing pretty quick… lol.

Anyway when it comes to Guild Wars 2, intel’s single core performance is king. Which is unfortunate that this late into the cpu game so many modern games are still using so few cores. Cool thing is some of the recent shooter games and some others are starting to take advantage of many cores. That definitely helps AMD’s cause, but doesn’t help us in GW2!! lol.

I had the AMD 8150 when it first came out, and switched to an intel 3570k solely because of GW2. There was a noticible FPS increase. And MMO’s do tend to be CPU hogs. Otherwise, Stormcrow gave the best advice for the hardware you do have. Overclock it if you can. Good luck!

His post made an assumption that AMD was better at the end, sadly it is not the case. I personally wish AMD were more vested in the cpu space as it would put more pressure on Intel for pricing and performance.

In the case of OP a new FX AMD cpu would be a sidegrade instead of an upgrade. AMD is budget, nothing wrong with them at all, but its one of those you pay more for better type of things.

If a game is GPU bound, like Metro/Crysis, then CPU brand won’t make a difference. However once it’s not, then Intel pulls ahead a lot more often than not. If a problem can’t easy be split into a bunch of equal sized parts then throwing cores isn’t going to cause it to scale linearly to cores available.

There may be many minor things that the main code could split off and check back when it’s done, that doesn’t automatically means significant performance improvements if the main code simply ends up waiting for the other thread to get done before proceeding or those other threads don’t take long to do their work. And if threads are fighting for the same resource like the hard drive, you aren’t going to get much scaling at all.

Sadly most SMP benchmarks are with apps designed to scale well. I think this causes some misconceptions about how powerful the CPU is with normal apps. That’s why you don’t see a huge disparity between PCMark scores compared with a SMP task like CineBench.

Pulling from TomsHardware charts.

Athlon II X2 250 – PCMark: 3201 – CineBench: 1.72 – Crysis II: 52.9
Athlon II X4 640 – PCMark: 3451 – CineBench: 3.36 – Crysis II: 86.6
I7-3770K – PCMark: 5360 – CineBench: 7.91 – Crysis II: 122.4

So thats 1.67x with PCMark and 4.6x with CineBench between low end Athlon dual core and high end i7 (1.08x and 1.95x between the Athlons) . So guess which end of the spectrum games fall? Crysis II here is tested at low settings and resolution to emphasis CPU impact and we see 2.31x between the Athlon dual core and the i7 and 1.64x between the Athlon dual and quad. While the ramp from 2 to 4 cores is evident in games it’s nowhere near 4x we see in SMP apps and much closer to 2x if that.

So while you can go look at Passmark and say “I have a powerful CPU see it roar”, in a lot of apps that potential simply isn’t being used and is difficult for an average single app to use. Running multiple active apps, great way to use all that power, single apps, not so much.

Behellagh.1468:

If a game is GPU bound, like Metro/Crysis, then CPU brand won’t make a difference. However once it’s not, then Intel pulls ahead a lot more often than not. If a problem can’t easy be split into a bunch of equal sized parts then throwing cores isn’t going to cause it to scale linearly to cores available.

There may be many minor things that the main code could split off and check back when it’s done, that doesn’t automatically means significant performance improvements if the main code simply ends up waiting for the other thread to get done before proceeding or those other threads don’t take long to do their work. And if threads are fighting for the same resource like the hard drive, you aren’t going to get much scaling at all.

Sadly most SMP benchmarks are with apps designed to scale well. I think this causes some misconceptions about how powerful the CPU is with normal apps. That’s why you don’t see a huge disparity between PCMark scores compared with a SMP task like CineBench.

Pulling from TomsHardware charts.

Athlon II X2 250 – PCMark: 3201 CineBench: 1.72
I7-3770K – PCMark: 5360 CineBench: 7.91

So thats 1.67x with PCMark and 4.6x with CineBench. So guess which end of the spectrum games fall? While the ramp from 2 to 4 cores is evident in games it’s nowhere near 4x we see in SMP apps and much closer to 2x if that. So while you can go look at passmark and say “I have a powerful CPU see it roar”, in a lot of apps that potential simply isn’t being used and is difficult for an average single app to use. Running multiple active apps, great way to use all that power, single apps, not so much.

This is very well worded ^

In short if all you do is play games and don’t run apps that make use of your cpu, an 8350 will perform just as well as a 4770k

sobe.4157:

This is very well worded ^

In short if all you do is play games and don’t run apps that make use of your cpu, an 8350 will perform just as well as a 4770k

not really, AMD offerings is really competitive to Intels. Unfornately, AMD new architecture favors multi threading performance over single threaded.

When the application is properly multi threaded. AMD chips became very competitive

BF4 is a good example of a properly threaded game
http://www.bf4blog.com/battlefield-4-alpha-gpu-and-cpu-benchmarks/

BF4 is a good example of an AMD optimized game ^

Avelos.6798:

BF4 is a good example of an AMD optimized game ^

I wouldnt say AMD optimized but heavy threaded.

I am not asking that I am asking because AMD uses hypertransport can the fps numbers be inaccurate? Since Guild wars 2 seems to prefer hyperthreading

Fine, I still have the charts up. Let me include the CPU we were “discussing” on the other thread as well as the i7-2600K.

Athlon II X2 250 – PCMark: 3201 – CineBench: 1.72 – Crysis II: 52.9
Athlon II X4 640 – PCMark: 3451 – CineBench: 3.36 – Crysis II: 86.6
FX-8120 – PCMark: 3987 – CineBench: 5.10 – Crysis II: 99.6
FX-8350 – PCMark: 4538 – CineBench: 6.94 – Crysis II: 116.2
i7-2600K – PCMark: 4817 – CineBench: 6.83 – Crysis II: 120.0
I7-3770K – PCMark: 5360 – CineBench: 7.91 – Crysis II: 122.4

If we compare the FX to the i7-3770K we get with the -8120 that the i7 is 1.34x faster in PCMark, 1.55x in CineBench and 1.29x in Crysis II. With the -8350 the numbers are PCMark 1.18x, CineBench 1.14x and Crysis II 1.05x while when compared to the i7-2600K we see 1.06x, 0.98x and 1.03×. Windows see all four CPUs as supporting 8 threads at the same time so the playing field is as level as it’s going to get comparing games.

Chyanne Waters.8719:

I am not asking that I am asking because AMD uses hypertransport can the fps numbers be inaccurate? Since Guild wars 2 seems to prefer hyperthreading

Hypertransport is a bus the ties the CPU to the motherboard northbridge chip on an AMD platform. Hyperthreading is the ability of an Intel core to run two threads at the same time on one core.

Yes, you had “Zombies.” But this is “Zombie Redneck Torture Family.” Entirely separate thing. It’s like the difference between an elephant and an elephant seal.

Anyways it’s just the opposite with Hyperthreading. An Intel core can run single thread much faster, nearly double, where AMD having a core per thread simply can’t double it’s clock speed. So in an environment when there aren’t a lot of threads running, using up a lot of “cores” (from the OS’s point of view), Intel simply out runs it.

Chyanne, to put it simply, forget about Hyperthreading or “HyperTransport” for a second. Pretend neither exists, look at the raw single threaded performance, Intel nearly doubles AMD’s performance.

This does not mean that they are bad CPUs, almost any CPU you can get today is going to be fairly powerful, some are just better for their price than others. How good of a choice a CPU is really depends on your needs and the applications you intend to run. So if you are just for example, like me and run a test bench for peak performance numbers for review products, or for my main rig with gaming, encoding, CAD work, Folding@Home, etc. The Intel is the obvious choice as it reduces the amount of time to perform certain tasks.
(Or in the case of this game, better framerates)

You can bend this so many different ways it’s crazy.
Example lets look and another game and a simple i3 2120 beats the 8350
http://www.tomshardware.com/reviews/gaming-processor-frame-rate-performance,3427-6.html

Something simple to look at is single thread performance which nobody will say a AMD can match a Intel. 8350 vs a simple i3 2100 the 2100 has a 30% slower clock speed but is still 4% faster at single core performance. http://cpuboss.com/cpus/Intel-Core-i3-2100-vs-AMD-FX-8350

Now against a 4670K where the prices are similar http://cpuboss.com/cpus/Intel-Core-i5-4670K-vs-AMD-FX-8350
For overall performance in gaming the Intel is the way to go now if you hand pick a game or 2 you could say the 8350 is the way to go but who only plays one or 2 games and don’t switch a year or so down the road?

As long as most games are more single core dependent as are 99.% of them, meaning a single core loaded 90% or so and the rest less than 60% again Intel is the way to go.

Their not really going to start making games that will put them out of a lot of their customers by demanding you have 8 cores which in the case of the 8350 is not a true 8 core since 2 cores share the same resources.

If it came down to that everybody would need to buy a $500 six core processor from Intel to get the best performance.

This is in no way a hit on AMD they had good stuff in the day but have since pretty much just gave up trying to compete and are now looking at their APU processors to try and take a different market. Look at their A10’s for about $500 you can build a decent gaming PC , or buy a 750K and pair it with a 7850 video card and play most games with very good performance.

This is my take on 99% of all the new games that will be released.
We have millions of dollars in this already lets release it and get the money back (coming from the investors) The developers say the game is not ready is poorly optimized and has to many bugs. Investors say roll it out we will fix the bugs later and then optimize. So after that you get a game that has so many bugs they chase their tells fixing this and that and breaking something along the way so the game never gets optimized.

So not that AMD multi core CPUS are in all next gen consoles (xbox, PS4 and WII U), will game be better optimized for AMD?

That’s difficult to answer, as there are both nvidia and amd titles that perform better on the competitor’s cards. As far as what the future holds, imo at least I think ports from console to pc will be overall more refined with amd helming the console market. As far as anything else, we will just have to wait and see.

Consoles are always easier to program than PCs, advantage of dedicated hardware and knowing the limitations up front. Everyone has the same CPU and GPU.

Again it’s not an AMD thing. It’s a multi-threaded thing. To scale with number of cores you need to be able to split up the workload evenly to get the scaling from additional cores. Easy for a ray trace, scientific or video compression apps, not so easy for everything else and that includes gaming.

Nikola…

Consols games dont have to worry about API bottlenecks. The games interact almost directly to the hardware, meaning they can get alot more performance out of them vs PC’s, that have an API like DirectX or OpenGL in the way.
Also, those AMD cpu’s are only 1.75ghz at best per core, they are very week, so games will be coded to use more cores, not more AMD cores, just mroe cores in general.
if said games get ported over ot the PC with the same core optimisations, then both AMD AND Intel will benefit, that said those PC ports will once again be limited by the API.

however…again… there is yet another thing to take into account. AMD has their new Mantle API coming out, which may or may not show huge performance gains over DirectX, and it only works with AMD GPU’s. Said API would use the GPU more and so CPU load ‘could’ be reduced even more so.

Anythign and everythign about performance on PC and games vs next gen consols is all theory atm, we just dont know…what we do know is it will be interesting to find out. 8 core consols and a New APi could mean great things ..or it could mean absolutly nothing.

Anyway, people often say that its impossible to rewrite engine, too expensive etc, but one just have to look at LOTRO to find out thats not true at all…they improved engine several times during game life, going from dx9 to dx10 to dx11, each time with visible improvements