Valve Hardware Day 2006 - Multithreaded Edition
by Jarred Walton on November 7, 2006 6:00 AM EST- Posted in
- Trade Shows
Particle Systems Benchmark
The more meaningful of the two benchmarks in terms of end users is going to be the particle simulation benchmark, as this has the potential to actually impact gameplay. The only problem is that the map is a contrived situation with four rooms each showing different particle system simulations. As proof that simulating particle systems can require a lot of CPU processing power, and that Valve can multithread the algorithms, the benchmark is meaningful. How it will actually impact future gaming performance is more difficult to determine. Also note that particle systems are only one aspect of game engine performance that can use more processing cores; artificial intelligence, physics, animation, and other tasks can benefit as well, and we look forward to the day when we have a full gaming benchmark that can simulate all of these areas rather than just particle systems. For now, here's a quick look at the particle system performance results.
There are several interesting things we get from the particle simulation benchmark. First, it scales almost linearly with the number of processor cores, so the Core 2 Quad system ends up being twice as fast as the Core 2 Duo system when running at the same clock speed. We will take a look at how CPU cache and memory bandwidth affects performance in the future, but at present it's pretty clear that Core 2 once again holds a commanding performance lead over AMD's Athlon 64/X2 processors. As for Pentium D, we repeatedly got a program crash when trying to run it, even with several different graphics cards. There's no reason to assume it would be faster than Athlon X2, though, and we did get results with Pentium D on the other test.
Athlon X2 performed the same, more or less, whether running on 939 or AM2 - even with high-end DDR2-800 memory. Our E6700 test system generated inconsistent results when overclocked, likely due to limitations with the nForce 570 SLI chipset. For most of the platforms, the 20% overclock brought on average a 20% performance increase, showing again that we are essentially completely CPU limited. The lack of granularity makes the scores vary slightly from 20% but it's close enough for now. Finally, taking a look at Athlon 64 vs. X2 on socket 939, the second CPU core improves performance by ~90%
VRAD Map Compilation Benchmark
As more of a developer/content creation benchmark, the results of the VRAD benchmark are not likely to be as interesting to a lot of people. However, keep in mind that better performance in this area can lead to more productive employees, so hopefully that means better games sooner. (Or maybe it just means more stress for the content developers?)
The results we got on the map compilation benchmark support Valve's own research and help to explain why they would be very interested in getting more Core 2 Quad systems into their offices. We don't have a single core Pentium 4 processor represented, but even a Pentium D 920 still ends up taking more than twice as long as a Core 2 Duo E6700 system, and about four times as long as Core 2 Quad. Looking at the CPU speed scaling, a 20% higher clock speed with the Pentium D resulted in 19% higher performance. If Intel had tried to stick with the NetBurst architecture, they would need dual core Pentium D processors running at more than 6.0 GHz in order to match the performance offered by the E6700. We won't even get into discussions about how much power such a CPU would require.
Performance scales almost linearly with clock speed once again, improving by 20% with the overclocking. Moving from single to dual core Athlon chips improves performance by about 92%. Going from a Core 2 Duo to a Core 2 Quad on the other hand improves performance by "only" 84%. It is not too surprising to find that moving to four cores doesn't show scaling equal to that of the single to dual move, but an 84% increase is still very good, roughly equal to what we see in 3D rendering applications.
The more meaningful of the two benchmarks in terms of end users is going to be the particle simulation benchmark, as this has the potential to actually impact gameplay. The only problem is that the map is a contrived situation with four rooms each showing different particle system simulations. As proof that simulating particle systems can require a lot of CPU processing power, and that Valve can multithread the algorithms, the benchmark is meaningful. How it will actually impact future gaming performance is more difficult to determine. Also note that particle systems are only one aspect of game engine performance that can use more processing cores; artificial intelligence, physics, animation, and other tasks can benefit as well, and we look forward to the day when we have a full gaming benchmark that can simulate all of these areas rather than just particle systems. For now, here's a quick look at the particle system performance results.
There are several interesting things we get from the particle simulation benchmark. First, it scales almost linearly with the number of processor cores, so the Core 2 Quad system ends up being twice as fast as the Core 2 Duo system when running at the same clock speed. We will take a look at how CPU cache and memory bandwidth affects performance in the future, but at present it's pretty clear that Core 2 once again holds a commanding performance lead over AMD's Athlon 64/X2 processors. As for Pentium D, we repeatedly got a program crash when trying to run it, even with several different graphics cards. There's no reason to assume it would be faster than Athlon X2, though, and we did get results with Pentium D on the other test.
Athlon X2 performed the same, more or less, whether running on 939 or AM2 - even with high-end DDR2-800 memory. Our E6700 test system generated inconsistent results when overclocked, likely due to limitations with the nForce 570 SLI chipset. For most of the platforms, the 20% overclock brought on average a 20% performance increase, showing again that we are essentially completely CPU limited. The lack of granularity makes the scores vary slightly from 20% but it's close enough for now. Finally, taking a look at Athlon 64 vs. X2 on socket 939, the second CPU core improves performance by ~90%
VRAD Map Compilation Benchmark
As more of a developer/content creation benchmark, the results of the VRAD benchmark are not likely to be as interesting to a lot of people. However, keep in mind that better performance in this area can lead to more productive employees, so hopefully that means better games sooner. (Or maybe it just means more stress for the content developers?)
The results we got on the map compilation benchmark support Valve's own research and help to explain why they would be very interested in getting more Core 2 Quad systems into their offices. We don't have a single core Pentium 4 processor represented, but even a Pentium D 920 still ends up taking more than twice as long as a Core 2 Duo E6700 system, and about four times as long as Core 2 Quad. Looking at the CPU speed scaling, a 20% higher clock speed with the Pentium D resulted in 19% higher performance. If Intel had tried to stick with the NetBurst architecture, they would need dual core Pentium D processors running at more than 6.0 GHz in order to match the performance offered by the E6700. We won't even get into discussions about how much power such a CPU would require.
Performance scales almost linearly with clock speed once again, improving by 20% with the overclocking. Moving from single to dual core Athlon chips improves performance by about 92%. Going from a Core 2 Duo to a Core 2 Quad on the other hand improves performance by "only" 84%. It is not too surprising to find that moving to four cores doesn't show scaling equal to that of the single to dual move, but an 84% increase is still very good, roughly equal to what we see in 3D rendering applications.
Facebook
Twitter
RSS
55 Comments
View All Comments
primer - Tuesday, November 7, 2006 - link
it's a nice article. i'm anxious to see what develops out of Valve in the near future.how about using some Athlon FX, Opteron 100/1200 series or higher speed Athlon X2s for crying out loud. i know that here is a performance gap even with the higher AMD models currently, but please show us an attempt at not being as one-sided.
jm20 - Tuesday, November 7, 2006 - link
The tests at 2.0 and 2.4 Ghz for the AMD chips are there, you can plot how it 'should' improve. Here are my scaling projections based on the data given for the 939 platform.AMD Speed single dual
2.0 14.0 27.0
2.2 15.5 29.5
2.4 17.0 32.0
2.6 18.8 35.0
2.8 20.8 38.2
3.0 23.1 41.7
I just hope this will show up correctly :/
Very interesting article, I'm very excited to see an almost linear improvement from single to dual core.
yacoub - Tuesday, November 7, 2006 - link
Your Tested Systems details on page 8 - the last three show a CPU of a Venice 3200+ oc'd to 2.4GHz (hey that's what I run too!), but the headings suggest they should really show Intel chips. :)yacoub - Tuesday, November 7, 2006 - link
Yes please make it only visual and not actually a part of the game, because much like trying too hard to make graphics look realistic, it really just adds more frustration for the player than anything else. It's already annoying in HL2 getting caught up on the edge of prop_physics related objects. The last thing we need is also getting caught or bumped by another player ;)
JarredWalton - Tuesday, November 7, 2006 - link
If the world is more interactive, there might not actually be a need for "prop physics" objects. There's also the potential to make a game with different gameplay mechanics depending on how the character interacts with the world. It sounds like initially Valve will make it a visual enhancement, to make entities look more lifelike in their behavior, but down the road others could potentially do more. Like most of the enhancements, what we want to see is how they can actually change and improve gameplay beyond just being visual.yacoub - Tuesday, November 7, 2006 - link
"You're doing a disservice to the customer if you're not using all of the CPU power."okay so that alone shows the need for dual-core so the customer can offload background tasks to a second core so the game can truly get 100% of a core. ;)
timmiser - Tuesday, November 7, 2006 - link
Background tasks?? I agree with Gabe that I want all of my CPU working the game. I'm sitting here with an Athlon X2 4800+ and I can't run Flight Sim X at an acceptable frame rate all the while I watch the graph that shows CPU#2 at idle while CPU#1 tries to run the entire program by itself!On another note, how about Microsoft develope some type of new API (DirecThread?) that automatically takes care of utilizing multiple cores so that game companies like Valve don't have to employ entire multi-threaded R&D divisions just so their games use both cores.
It seems like we are going back to the caveman days before DirectInput, DirectSound, & DirectX etc. Remember when you had to choose your sound card, joystick, and video card from a list within the game?? Let's not go there again!!
-Tim
JarredWalton - Tuesday, November 7, 2006 - link
That was a direct quote from... Gabe I think. Others might disagree, but I thought it was an interesting take on things. As for background tasks, they usually only need a bit of CPU time (less than 5% in most cases), so unless you really want to encode videos and play games at the same time....yacoub - Tuesday, November 7, 2006 - link
What a worthless attempt at sarcasm:Why not just say something normal like:
It's not omg-witty, but at least that sentence doesn't end with a preposition. @___@
JarredWalton - Tuesday, November 7, 2006 - link
Which obviously ruined the whole article. Seriously yacoub, a little tact with posting would be nice. Rather than stating:You could have just said something normal like:
We are real people, and derogatory adjectives like "worthless" tend to irritate more than help. You may not intend it that way, but imagine for a second someone talking about what you worked on for a couple days last week and describing it as "absolute garbage". The key word in "constructive criticism" is to actually make it constructive.
Now, thanks for the suggestion, and I'm more than happy to change things a bit to appeas people. Like most people, I just appreciate a bit more consideration, even if I'm just doing my job here. :) As the old cliche goes, you can catch more flies with honey than with vinegar. (Well, my grandpa used to say that anyway.)