Homeworld Remastered - FX gets hammered

[Ramses]

This is another one of those games that I'd read the thread, go get/buy/demo the game, it'd run fine, and I'd be baffled wtf the problem is. But I don't have an FX anymore so someone else has to do that.

edit: looking at vids on youtube, do you really need more than like, 20fps on this thing? lol

 

[SlowSpyder]

This game runs on OpenGL and DirectX? If that's the case, why wouldn't I just run it in DX where I could disable half my cores, drop my clockspeed by a full 1/3 from factory, and still have three times my monitor's refresh rate at minimum?

 

[HurleyBird]

It makes very little sense for this game to be so CPU sensitive given that the gameplay has been drastically simplified versus the 1999 original. The remaster uses dice rolls for hits while the original featured honest to goodness bullet physics, and AI certainly hasn't been noticeably improved.

 

[Makaveli]

lol from those graphs you can tell which site actually did benchmarks once the game was at peak load and who did benchmarks at the start of the game when everyone is just building and nothing is going on.

And yes AMD chips get killed because this game doesn't seem to use more than 2 cores.

IPC and core efficiency will matter more.

 

[soccerballtux]

In defense of your position I believe that reality (i.e the FPS) of the CPU situation is what matters, not theory-crafting, compiler/code bias, or lack of optimization. However...

Something is wrong (with the game code) when an 3.0GHz 8-core HW-E leads the charts over a 3.6GHz 6-core HW-E (suggesting the game likes a combination of high IPC and multiple threads) but a 2 core Pentium beats a 8350. The 3GHz Pentium may still have the single thread advantage due to IPC over the 4GHz 8350, but not enough to counter the 2 vs 8 thread difference.

yeah. there's no way the FX is THAT bad.

 

[Sequences]

What are these numbers? Minimum? Average? Median? Max? These review charts have poor descriptive statistics.

 

[soccerballtux]

Possibly the game really benefits from the larger caches on LGA2011 chips? If it was all about single-thread performance, you'd expect the Phenom II to be beating up on the lower-end FX chips, but it isn't.

FX-8350 has 16MB cache

 

[soccerballtux]

Its probably bits of old code combined with single threaded prowess that trips up FX but Intel shines as usual. Point is FX hasn't gotten better. It isn't "future-proof".

EDIT: Techspot is testing OpenGL.

no. those of us supporting it consider it future proof for 21st century game engines that actually make proper use of higher level paradigms that 20 years of experience with software engineer has netted us.

 

[soccerballtux]

That's the AMD mantra...just wait and things will be better later. The problem is later never arrives.

considering we have yet to see many properly designed, efficient game engines besides UE3, yes, it's never going to arrive.

It's up the studios to make a non-crap architecture

At the end of the day I can encode BigBuckBunny on my FX-8310 at 264.5 FPS on the "Android" profile.
The best the Core i3 can pull is 8F5PS or so, 110-115 with overclocking.

That's all the proof we need that these older game engines are completely crap. There's a guy that runs a full-time LN2 rig at 5-something Ghz on his core i7 for his professional Starcraft 2 matches and even he dips to 15-20fps on the 8v8s

 

[Shivansps]

considering we have yet to see many properly designed, efficient game engines besides UE3, yes, it's never going to arrive.

It's up the studios to make a non-crap architecture

At the end of the day I can encode BigBuckBunny on my FX-8310 at 264.5 FPS on the "Android" profile.
The best the Core i3 can pull is 8F5PS or so, 110-115 with overclocking.

That's all the proof we need that these older game engines are completely crap. There's a guy that runs a full-time LN2 rig at 5-something Ghz on his core i7 for his professional Starcraft 2 matches and even he dips to 15-20fps on the 8v8s

Sure, enconding videos is the exact same thing that running AI...

Im gona be very simple here... once the DX botteneck is gone we just gona run in intro another, AI is where my money is, and thats a insteresting point because Starm Swarm does not have real AI.

The problem with these older games is that they where created before the multicore era, they are in no way crap, they are a product of their time, and you known what i kinda impressed by how much could be done back them in a single core.
And even with a modern engine, you just cant solve the AI problem that a game like Homerworld likely has.

 

[dave_the_nerd]

And even with a modern engine, you just cant solve the AI problem that a game like Homerworld likely has.

The AI _is_ pretty stupid.

 

[Shivansps]

The AI _is_ pretty stupid.

Yes but a game like that, with high number of individual AI, sends everything to hell, Sins of a Solar Empire Rebellion has the same problem, it starts perfectly, but once the total unit number gets too high it just kills any cpu, and its not a DX problem.

 

[Phynaz]

considering we have yet to see many properly designed, efficient game engines besides UE3, yes, it's never going to arrive.

It's up the studios to make a non-crap architecture

At the end of the day I can encode BigBuckBunny on my FX-8310 at 264.5 FPS on the "Android" profile.
The best the Core i3 can pull is 8F5PS or so, 110-115 with overclocking.

That's all the proof we need that these older game engines are completely crap. There's a guy that runs a full-time LN2 rig at 5-something Ghz on his core i7 for his professional Starcraft 2 matches and even he dips to 15-20fps on the 8v8s

You feel it's the software devs job to make their software run well on AMD CPUs.

I feel it's AMD's job to make their CPUs run software well.

 

[escrow4]

yeah. there's no way the FX is THAT bad.

It is in OpenGL, and looking online the original engine is OpenGL and AMD fails so so badly.

Straight from Techspot:

"A well designed first-person shooter should offer fairly consistent frame rates throughout a level. RTS games on the other hand typically start off with extremely high frame rates as there are few units for the GPU to render and for the CPU to calculate. As the unit count increases, the frame rate tends to drop and once a major battle takes place this is typically when you will see the lowest frame rates.

Homeworld Remastered was tested in a similar manner to StarCraft II. We played a skirmish with seven AI-controlled players in a 4v4 match. The resources were set to their maximum value and we went with a 15 minute build time so players could reach critical mass. Just moments before the first major battle was set to take place we made a save which could be loaded repeatedly.

As the test starts over 200 of my own ships engage the enemy and all hell breaks loose for two minutes. With the GTX 980 frame rates were initially around 140fps at 2560x1600 but that dipped down to 50fps as the ships began to engage each other.

So if we were to benchmark the game in the building stage, gathering resources and what not, then frame rates would be around three times higher than what we are going to show during our massive battle scene, so keep that in mind."

Real gameplay, who knows what GameGPU tested.

http://www.techspot.com/review/970-homeworld-remastered-benchmarks/

 

[Ramses]

You feel it's the software devs job to make their software run well on AMD CPUs.

I feel it's AMD's job to make their CPUs run software well.

Seems like software can/does change a lot more often than CPU guts do.
I can't help but think it's easier to write software than to design/build a CPU, no? What is the AMD cpu now, four year old, five?

Furthering that, the more I think about it the more I have a real problem with the difference between testing software and testing hardware. One could design an amazing CPU, one with features unheard of, doing things previously thought impossible, an engineering triumph. But you test it with software that is totally unaware of these amazing advances, and it bombs.

Is the software crap, or is the CPU crap?

Maybe I'm a product of my era but I have a hard time blaming the hardware. What if dx12 comes along and does what is promised, what if you could apply it to the oft mentioned "64player MP BF4" that is always trotted out as being a bad spot for an FX chip, what if it suddenly allowed the FX to perform acceptably? Is the CPU suddenly better? It didn't change, how can it? Does software = hardware? Does current software dictate the perceived quality/value of hardware?

There are IMO some philosophical channels here that I really wish were not a part of this hobby. It's hard enough to determine what is good hardware and what is best, let along if it depends on what damn software you are running. I know it's not likely to change, and I know the easy answer is you judge hardware by the software you have at hand, but that's a shallow answer. I'm not into shallow. Especially not in technical hobbies.

 

[Idontcare]

Seems like software can/does change a lot more often than CPU guts do.
I can't help but think it's easier to write software than to design/build a CPU, no? What is the AMD cpu now, four year old, five?

Furthering that, the more I think about it the more I have a real problem with the difference between testing software and testing hardware. One could design an amazing CPU, one with features unheard of, doing things previously thought impossible, an engineering triumph. But you test it with software that is totally unaware of these amazing advances, and it bombs.

Is the software crap, or is the CPU crap?

Maybe I'm a product of my era but I have a hard time blaming the hardware. What if dx12 comes along and does what is promised, what if you could apply it to the oft mentioned "64player MP BF4" that is always trotted out as being a bad spot for an FX chip, what if it suddenly allowed the FX to perform acceptably? Is the CPU suddenly better? It didn't change, how can it? Does software = hardware? Does current software dictate the perceived quality/value of hardware?

There are IMO some philosophical channels here that I really wish were not a part of this hobby. It's hard enough to determine what is good hardware and what is best, let along if it depends on what damn software you are running. I know it's not likely to change, and I know the easy answer is you judge hardware by the software you have at hand, but that's a shallow answer. I'm not into shallow. Especially not in technical hobbies.

This is an odd conversation in its own right because the definition of success appears to be "winning the academic argument".

The real answer should be "it (be it software or hardware) is crap if it fails to generate the sales and revenue that it was expected to generate for the business owners and shareholders of the business that invested in creating said hardware or software".

The hardware is crap if its market viability squarely depends on unaffiliated businesses electing to create enabling software for said hardware, only the software then never shows up. Put a fork in the hardware, its done.

But you know what else was crap in that equation? The business leader who made the decision at the hardware company to create a product that would flop in the market if not for a software company or three who were needed to swoop in (time the market) and release the software that would create a viable market for the hardware.

If you are in the business of creating hardware that is going to live or die on the basis of software being available to make the hardware desirable, then you need to be in the business of ALSO making the software.

Like Apple did. Or IBM, or HP, or Cray, or just about any other well run business back in their day of owning and rolling niche hardware features.

 

[Ramses]

You aren't wrong. At all.

But I refuse to be a part of the business/financial end of things. This is a hobby, and the stuff I outlined above seems to be a relatively recent development, least in my world, and it's bumming me out.
I'm interested in hardware for the sake of the hardware. The same reason I like a nice pocket knife or a zippo lighter or my fifty year old adjustable crescent wrench. The same reason I spend many times the value on old cars. The same reason I rebuild old abandoned household stuff with new guts. They are quality hardware, and it's nothing to do with anything remotely associated with business or finances. I get enough of that crap at work.

The only other thing I can say is I've seen some outstanding devices fail miserably financially. That didn't make them crap imo. Bad business decisions associated with them, sure, but the hardware is pure. It does not care how many it sold or how it was reviewed, it just is. And I strive to evaluate and value such things on their own merits, independent of profit or marketing or competition(to a degree). It's no different than the way I judge a forty year old car, they are universally piles of trash compared to anything and everything current in every measurable way, but there is more to the machine than that.

 

[soccerballtux]

You feel it's the software devs job to make their software run well on AMD CPUs.

I feel it's AMD's job to make their CPUs run software well.

to make effective use of multiple threads. how did you get anything but this out of what I said?

 

[soccerballtux]

Seems like software can/does change a lot more often than CPU guts do.
I can't help but think it's easier to write software than to design/build a CPU, no? What is the AMD cpu now, four year old, five?

Furthering that, the more I think about it the more I have a real problem with the difference between testing software and testing hardware. One could design an amazing CPU, one with features unheard of, doing things previously thought impossible, an engineering triumph. But you test it with software that is totally unaware of these amazing advances, and it bombs.

Is the software crap, or is the CPU crap?

Maybe I'm a product of my era but I have a hard time blaming the hardware. What if dx12 comes along and does what is promised, what if you could apply it to the oft mentioned "64player MP BF4" that is always trotted out as being a bad spot for an FX chip, what if it suddenly allowed the FX to perform acceptably? Is the CPU suddenly better? It didn't change, how can it? Does software = hardware? Does current software dictate the perceived quality/value of hardware?

There are IMO some philosophical channels here that I really wish were not a part of this hobby. It's hard enough to determine what is good hardware and what is best, let along if it depends on what damn software you are running. I know it's not likely to change, and I know the easy answer is you judge hardware by the software you have at hand, but that's a shallow answer. I'm not into shallow. Especially not in technical hobbies.

once we know the execution resources, IMO, measuring branch-prediction accuracy is the only other variable in the environment

 

[soccerballtux]

This is an odd conversation in its own right because the definition of success appears to be "winning the academic argument".

The real answer should be "it (be it software or hardware) is crap if it fails to generate the sales and revenue that it was expected to generate for the business owners and shareholders of the business that invested in creating said hardware or software".

The hardware is crap if its market viability squarely depends on unaffiliated businesses electing to create enabling software for said hardware, only the software then never shows up. Put a fork in the hardware, its done.

But you know what else was crap in that equation? The business leader who made the decision at the hardware company to create a product that would flop in the market if not for a software company or three who were needed to swoop in (time the market) and release the software that would create a viable market for the hardware.

If you are in the business of creating hardware that is going to live or die on the basis of software being available to make the hardware desirable, then you need to be in the business of ALSO making the software.

Like Apple did. Or IBM, or HP, or Cray, or just about any other well run business back in their day of owning and rolling niche hardware features.

is it POSSIBLE for AMD to create a much-better-single-threaded-performance CPU core?

 

[Ramses]

is it POSSIBLE for AMD to create a much-better-single-threaded-performance CPU core?

It would depend on ones definition of possible.
I estimate I have twenty good years left before my physical and/or cognitive functions degrade a noticeable amount, so I could learn what I'd need to know to develop one myself in the time remaining. But it's not very likely.

A better way to approach it may well be asking if there is any reason they can not or will not develop a cpu with better single core performance. I lack data to speculate if I was so inclined though.

 

[Fox5]

How can such an old game run so poorly on modern cpus?

Homeworld 2 came out in 2003. At best, the fastest single core cpu was half as fast as a single core of a modern AMD cpu.

 

[wright.pwa]

You feel it's the software devs job to make their software run well on AMD CPUs.

I feel it's AMD's job to make their CPUs run software well.

Well said!

This game runs on OpenGL and DirectX? If that's the case, why wouldn't I just run it in DX where I could disable half my cores, drop my clockspeed by a full 1/3 from factory, and still have three times my monitor's refresh rate at minimum?

The game engine only supports OpenGL so I am guessing that's why they didn't test using DX

What are these numbers? Minimum? Average? Median? Max? These review charts have poor descriptive statistics.

The graphs also don't say who created them so I guess you have to read the actual review to get the info.

How can such an old game run so poorly on modern cpus?

Homeworld 2 came out in 2003. At best, the fastest single core cpu was half as fast as a single core of a modern AMD cpu.

Unit caps.

lol from those graphs you can tell which site actually did benchmarks once the game was at peak load and who did benchmarks at the start of the game when everyone is just building and nothing is going on.

And yes AMD chips get killed because this game doesn't seem to use more than 2 cores.

IPC and core efficiency will matter more.

This is the gamegpu.ru test :S

https://www.youtube.com/watch?v=AV8iQICQ3VI

 

[Ramses]

How can such an old game run so poorly on modern cpus?

Homeworld 2 came out in 2003. At best, the fastest single core cpu was half as fast as a single core of a modern AMD cpu.

That's another facet to my gripe against software.
Just because you build a computer and it runs a given game great, does not mean you can revisit that game five years later and have it run just as or more amazingly.

 

[wright.pwa]

That's another facet to my gripe against software.
Just because you build a computer and it runs a given game great, does not mean you can revisit that game five years later and have it run just as or more amazingly.

True but if you play the original version of Homeworld which is included in the remastered package you will find the game runs much better. The TechSpot test isn't even possible in the original version due to the unit cap. There is no way CPU's back in 2003 could handle hundreds of ships from just one player, let alone eight.