***News*** Starcraft II beta benchmarked

Starcraft II beta benchmarked

Core i5/i7 ideal, Phenom II X2 545 recommended

As the gaming community keeps their eyes open for Starcraft II related news, PC Games Hardware has taken 13 CPUs and put them through their paces in Starcraft II beta benchmark.

Note that for reasons of convenience, the benchmark was ran in Starcraft II's replay feature, which takes AI calculations out of the test.

Starcraft II uses only two cores, but quad core Q6600 ends up slightly faster than dual core E6600. The game apparently likes additional cache, which works well for Intel's E8400 and two Lynnfield CPUs. Phenom CPUs also did pretty well, although they're behind i5 and i7 offerings.

Based on these results in this particular scene, Core i5 and i7 CPUs would be ideal CPUs for the game, but Phenom X2 545 at 3.0 GHz is obviously a much wiser investment as it performes a bit slower but comes at a significantly lower price point.

You can find benchmark results and check out the video here.

Fudzilla

SC2 means hardware upgrade time for me :)

*edit, for some reason i cant link the page where the video comes from, so here i will put it unlinked.

http://www.pcgameshardware.com/aid,705392/Starcraft-2-Beta-CPU-benchmarks-x-Core-i5/i7-leading/Practice/

Im not getting this right, these are poor results, most of them :(GazaAli

It was a setup benchmark so this is most likely a worst case scenario!

EDIT: I found the video of the test and yes, its just on big battle! VIDEO

Also, why did they turn off turbo mode on the i5? Surely the FPS would have been even better with it on... :/

EDIT: Just fond this on there: "The reason we disable the turbo mode for benchmarks is, that it works dynamically depending on workload and temperature and can even vary from mainboard to mainboard. Plus according to our test results, the turbo mode leads to a higher power consumption, but almost no gain performance-wise. Nevertheless in print tests we include TM sometimes for comparison reasons, but don`t use it as our standard setting."

An RTS usually doesn't require too high framerates. Some were locked to go no higher than 30 fps such as Generals and Tiberium Wars. I think my Phenom II X3 720 BE with HD 5770 should be fine.

An RTS usually doesn't require too high framerates. Some were locked to go no higher than 30 fps such as Generals and Tiberium Wars. I think my Phenom II X3 720 BE with HD 5770 should be fine.

jun_aka_pekto

Why would you lock a game to 30fps? Seems like quite a random number to go for, why not make it 60fps?Daytona_178

Beats me. Ask Westwood. They're the ones who made Generals and Tiberium wars. I noticed both games were locked at 30 fps when using FRAPS.

[QUOTE="Daytona_178"]Why would you lock a game to 30fps? Seems like quite a random number to go for, why not make it 60fps?jun_aka_pekto

Beats me. Ask Westwood. They're the ones who made Generals and Tiberium wars. I noticed both games were locked at 30 fps when using FRAPS.

I suppose it makes sense. When a game is locked at 30 fps, more resources can be dedicated to maintaining that frame rate rather than having it unlimited with the resulting big highs and lows.

It's more evident in flight sims. A sustained/locked 24 fps rarely has stutters while alternating between high and low frame rates often resulted in micro-pauses.

Haha please don't tell me that game, looking how it does. Only hits 40-45 FPS on a Zerg rush @ 16x resolution and 1xaa with a 5870, and high end quad cpu. I would expect 200fps to be honest, the polycount looks so low.

Haha please don't tell me that game, looking how it does. Only hits 40-45 FPS on a Zerg rush @ 16x resolution and 1xaa with a 5870, and high end quad cpu. I would expect 200fps to be honest, the polycount looks so low.

Solar-X

Haha please don't tell me that game, looking how it does. Only hits 40-45 FPS on a Zerg rush @ 16x resolution and 1xaa with a 5870, and high end quad cpu. I would expect 200fps to be honest, the polycount looks so low.

Solar-X

http://www.legionhardware.com/articles_pages/starcraft_ii_wings_of_liberty_beta_performance

Better?????

Poor optimization as a way to use more resource T.T the end of pc gaming!!! The console people are right! jk jk. To be fair is not an fps, we will just have to see it for ourselves.

the test is obviously flawed...

From everyone who played it, the performance is good...

And look at my link above...

the test is obviously flawed...

From everyone who played it, the performance is good...

And look at my link above...

Idontremember

Obviously you overlooked the fact that they used an i7 965 @ 3.7 GHz.

Console gaming should rot in hell...developers aren't even putting the effort into optimization anymore.

[QUOTE="Idontremember"]

the test is obviously flawed...

From everyone who played it, the performance is good...

And look at my link above...

kaitanuvax

Obviously you overlooked the fact that they used an i7 965 @ 3.7 GHz.

Console gaming should rot in hell...developers aren't even putting the effort into optimization anymore.

Obviously you forgot to look at the other graph on the next page with the cpu scaling for an i7...

Obviously you forgot to look at the other graph on the next page with the cpu scaling for an i7...

Idontremember

And what does that prove? You cant expect to compare a downclocked top-of-the-line CPU to lesser CPUs.

[QUOTE="Idontremember"]

Obviously you forgot to look at the other graph on the next page with the cpu scaling for an i7...

kaitanuvax

And what does that prove? You cant expect to compare a downclocked top-of-the-line CPU to lesser CPUs.

It's playable with a 1,7 Ghz I7, It might be an I7, but it's still clocked super low, the performance are gonna be mutch lower compare to any current average processor.

If 1,7GHz can run it, then there is no reason to say an average core 2 duo or core 2 quad, or a phenom would fail to perform like shown in the Graph on page 1.

[QUOTE="GazaAli"]Im not getting this right, these are poor results, most of them :(Daytona_178

It was a setup benchmark so this is most likely a worst case scenario!

EDIT: I found the video of the test and yes, its just on big battle! VIDEO

They're probably still working on the pathfinding though, which can absolutely kill a CPU.

[QUOTE="kaitanuvax"]

[QUOTE="Idontremember"]

Obviously you forgot to look at the other graph on the next page with the cpu scaling for an i7...

Idontremember

And what does that prove? You cant expect to compare a downclocked top-of-the-line CPU to lesser CPUs.

It's playable with a 1,7 Ghz I7, It might be an I7, but it's still clocked super low, the performance are gonna be mutch lower compare to any current average processor.

If 1,7GHz can run it, then there is no reason to say an average core 2 duo or core 2 quad, or a phenom would fail to perform like shown in the Graph on page 1.

That's exactly the issue. The CPU performances in the first graph are horrific to say the least. If it takes a $1000 CPU (at stock) and a $400 GPU just to break 70 FPS, then we seriously have a problem here.

[QUOTE="Idontremember"]

[QUOTE="kaitanuvax"]

And what does that prove? You cant expect to compare a downclocked top-of-the-line CPU to lesser CPUs.

kaitanuvax

It's playable with a 1,7 Ghz I7, It might be an I7, but it's still clocked super low, the performance are gonna be mutch lower compare to any current average processor.

If 1,7GHz can run it, then there is no reason to say an average core 2 duo or core 2 quad, or a phenom would fail to perform like shown in the Graph on page 1.

That's exactly the issue. The CPU performances in the first graph are horrific to say the least. If it takes a $1000 CPU (at stock) and a $400 GPU just to break 70 FPS, then we seriously have a problem here.

If the Graph was accurate, then there would be a load of complaint on every forum around... It cant be accurate

[QUOTE="kaitanuvax"]

[QUOTE="Idontremember"]

It's playable with a 1,7 Ghz I7, It might be an I7, but it's still clocked super low, the performance are gonna be mutch lower compare to any current average processor.

If 1,7GHz can run it, then there is no reason to say an average core 2 duo or core 2 quad, or a phenom would fail to perform like shown in the Graph on page 1.

Idontremember

That's exactly the issue. The CPU performances in the first graph are horrific to say the least. If it takes a $1000 CPU (at stock) and a $400 GPU just to break 70 FPS, then we seriously have a problem here.

If the Graph was accurate, then there would be a load of complaint on every forum around... It cant be accurate

[QUOTE="Idontremember"]

[QUOTE="kaitanuvax"]

That's exactly the issue. The CPU performances in the first graph are horrific to say the least. If it takes a $1000 CPU (at stock) and a $400 GPU just to break 70 FPS, then we seriously have a problem here.

Swiftstrike5

If the Graph was accurate, then there would be a load of complaint on every forum around... It cant be accurate

Well, people with a blizzard account who registered for the beta could have had a chance to participate...

[QUOTE="Swiftstrike5"][QUOTE="Idontremember"]

If the Graph was accurate, then there would be a load of complaint on every forum around... It cant be accurate

Idontremember

Well, people with a blizzard account who registered for the beta could have had a chance to participate...

[QUOTE="Solar-X"]

Haha please don't tell me that game, looking how it does. Only hits 40-45 FPS on a Zerg rush @ 16x resolution and 1xaa with a 5870, and high end quad cpu. I would expect 200fps to be honest, the polycount looks so low.

Daytona_178