Benchmark Request

[damric]

W1zzard,

I get a lot of people asking me about Minecraft performance. Would it be possible to add that into your benchmark suite for graphics cards? Also for CPUs.

I hate asking because I think the game is so stupid, but it gets brought up a lot on other forums and I always point people to your reviews.

Thanks,

damric

 

[Luka KLLP]

Isn't RAM the most important for Minecraft? Especially when you have 200+ mods

 

[RCoon]

Nobody but Dave does reviews of CPUs and he's not done that for a long time now. As above RAM does more for mine craft. Also, if he did it for all his GPU's his benchmark results would be in the thousands of frames per second because that's what you get with high end systems. It would make the charts look ridiculous.

 

[Aquinus]

Nobody but Dave does reviews of CPUs and he's not done that for a long time now. As above RAM does more for mine craft. Also, if he did it for all his GPU's his benchmark results would be in the thousands of frames per second because that's what you get with high end systems. It would make the charts look ridiculous.

MineCraft definitely does not run at "thousands of frames per second" on any machine. I suspect a lot of the work Minecraft needs to do is computational work, in which case a JVM-based benchmark might partially reflect such performance.

 

[buildzoid]

Nobody but Dave does reviews of CPUs and he's not done that for a long time now. As above RAM does more for mine craft. Also, if he did it for all his GPU's his benchmark results would be in the thousands of frames per second because that's what you get with high end systems. It would make the charts look ridiculous.

My R9 290 + FX 6350 system does about 100ish FPS without any explosions on max graphics. After putting a bit of TNT down I was at 45FPS so if someone wanted to benchmark minecraft they would just need to make map that has a 5x5x5 cube of TNT detonate it and measure FPS while it's exploding. On weak CPUs it'll choke completely.

 

[Aquinus]

My R9 290 + FX 6350 system does about 100ish FPS without any explosions on max graphics. After putting a bit of TNT down I was at 45FPS so if someone wanted to benchmark minecraft they would just need to make map that has a 5x5x5 cube of TNT detonate it and measure FPS while it's exploding. On weak CPUs it'll choke completely.

Optifine makes a huge difference in performance and/or quality depending on your goal. I found with the right modifications that MineCraft will run silky smooth and look gorgeous at the same time. I've had my fill of MineCraft though, I think. It's been quite a while since I've played it.

 

[RCoon]

MineCraft definitely does not run at "thousands of frames per second" on any machine. I suspect a lot of the work Minecraft needs to do is computational work, in which case a JVM-based benchmark might partially reflect such performance.

I haven't played it in years, but I recall the Base game getting hundreds of fps on a standard map. The other issue would be deciding what mods to apply to the game for a suitable average benchmark figure. Not everybody uses the same mods, so the benchmark figures wouldn't be representative for a lot of people.

 

[buildzoid]

I haven't played it in years, but I recall the Base game getting hundreds of fps on a standard map. The other issue would be deciding what mods to apply to the game for a suitable average benchmark figure. Not everybody uses the same mods, so the benchmark figures wouldn't be representative for a lot of people.

No need for mods since those just need XXXX MB of RAM to run(my friend once went insane with mods and built a pack that needs at least 10GB of RAM just to launch) really to test the performance of something in mine craft just put some TNT down. Also the performance of many mods is tied to what they do. For example many of the mods with pipes in them make no impact on performance until there are XXX items moving around in the pipes. So a modded minecraft benchmark is impossible to do properly. IMO just measuring FPS in forest while XXX blocks of TNT detonate would be the most effective since everyone can relate to that.

 

[newtekie1]

Slowdown by TNT is all CPU, it is from the physics calculations. So that wouldn't be a worthwhile test for GPUs.

 

[buildzoid]

Slowdown by TNT is all CPU, it is from the physics calculations. So that wouldn't be a worthwhile test for GPUs.

well minecraft isn't GPU bound. And there's not much you can do to change that. Also when people complain that MC is running slow it's always the CPU being too weak because that's just how minecraft is. Really a minecraft CPU benchmark makes sense a minecraft GPU benchmark is stupid because that just means you get some shader mods and photo realistic 512x512 textures(I doubt that would do much to the FPS anyway) and people who run modded minecraft don't use those because minecraft mods needcustom 512x512 textures and those are either hard too find or non-existent. But all mods that deal with explosives or large quantities of entities moving around the map will hammer the CPU. So a CPU test is useful.

 

[damric]

Thanks for all your inputs.

Like I said, I don't play it. My 4 year old does, but he plays it on a smart phone.

So, when people ask for specs for a minecraft rig, what should I reply? Are we talking needing an i7 and 32GB RAM or can this be done on an AMD APU, or what?

 

[Luka KLLP]

Well I can play Minecraft with quite a bit of mods with my current PC, so unless you go absolutely crazy with mods (like the YogsCast Complete Pack haha) you need almost nothing