Mac Pro 3D test scores revisited
(Credit:
Sarah Tew/CNET)
We've received a few e-mails since our post last week on Apple's new Mac Pro and its higher-end 3D graphics card option, AMD's Radeon HD 4870. Echoing the comments in the original post, some folks wanted to know how we tested. Others wanted us to test the Mac Pro in Boot Camp. Also, AMD wanted us to show actual test results to back up our claim that the combination of the Mac Pro and its Radeon HD 4870 card was not great for gaming, especially compared with competing Windows desktops.
We're happy to oblige everyone.
First, how we tested. We used the same Call of Duty 4 timedemo files Apple uses for its own performance assessments. At least, that's what Apple said it sent to us. We also followed Apple's exact recommended test procedure and settings, which were even more aggressive than we originally intended. Here's the full breakdown, which we applied to all four systems in the charts below:
Graphics options:- Video mode: 1,680x1,050
- Sync every frame: No
- Aspect ratio: Auto
- Anti-aliasing: 4x
- Every frame: No
- Shadows: Yes
- Specular map: Yes
- Depth of field: Yes
- Glow: Yes
- Number of dynamic lights: Normal
- Soften smoke edges: Yes
- Ragdoll: Yes
- Bullet impacts: Yes
- Model detail: Normal
- Water detail: Normal
- Texture filtering: Trilinear
- Anisotropic filtering: Maximum (drag slider completely to the right)
- Texture quality: Manual
- Texture resolution: Extra
- Normal map resolution: Extra
- Specular map resolution: Extra
- Enable console: Yes
The numbers below incorporate the Mac OS X-based scores we ran last week, as well as the Windows Vista 64-bit scores we generated today. We made sure all relevant software was updated to its must current version in both cases, including each OS, graphics driver, and version of Call of Duty 4. Fortunately, Apple's timedemo files also worked in Vista. The software playing field is as level as we can get it between the two operating systems.
(Longer bars indicate better performance)
 |
Pipeline | |
Bog | |
Ambush |
The Mac Pro gains a step in Boot Camp, but the sub-$1,000 Velocity Micro Edge Z5 is the clear winner on Apple's own test. The specifications on that Velocity Micro system include a factory-overclocked 2.93GHz Intel Core 2 Quad Q9400, 4GB 800MHz DDR2 SDRAM, a vendor-specific 512MB ATI Radeon HD 4850 3D card (with faster core and memory clocks than standard), a 750GB 7,200rpm hard drive, and Windows Vista 64-bit. The review will hopefully post tomorrow.
Before anyone jumps all over us, we're absolutely not arguing that a $999 Windows PC is a better all-around computer than the Mac Pro. In the course of our regular benchmarking, the new Mac Pro blasted through our multitasking test in 133 seconds, a new record, and posted a second best (behind only an $8,800 Maingear) 21,204 score on our multicore Cinebench test. The Velocity Micro system took 440 seconds for multitasking, and managed only a 12,931 on Cinebench. If we had to buy a system for digital media production, we'd pick up the Mac Pro in a heartbeat.
Also to the Mac Pro's credit, its gaming profile improves considerably in Vista via Boot Camp compared with what we saw in OS X. What we want to make crystal clear, however, is that with two quad-core Nehalem-based Intel chips and a well-regarded 512MB ATI Radeon HD4870 graphics card, the Mac Pro might sound tempting as a serious gaming system. It's certainly a more competent gamer than older models, especially if you rely on Boot Camp. But from the standpoint of pure 3D bang-for-your-buck, the Windows gaming ecosystem holds a clear advantage over the Mac Pro, with faster gaming PCs available for $2,500 to $2,600 less.
Rich Brown reviews desktops and various other components and peripherals for CNET. E-mail Rich. 
all of us over at the pc gaming camp knew that for ages
I'm surprised that the Velocity Micro machine came out on top,the 4850 is still slower than the 4870 any way you put it. I guess Intel is right, the CPU(even clock speed) still matters in gaming.
Adobe all around is pretty much bloat ware in my opinion. But in either case, for video compositing and what not, all I ever use from Adobe is After Effects. Everything else is Final Cut Studio (although Motion is starting to become a quicker replacement for basic things that I don't need After Effects for). Even most production houses use Avid.
There's a joke between me and my video friends that Premier Pro is for weddings and grandma's. You're either using FCP or Avid for video production, which is what seamonkey420 was referring to, not content creation.
By the way, I'm tired of people's misspellings or adding 420 in their screen names. It really ruins credibility.
No way, Final Cut Studio 2 wipes the floor with Premier. Final Cut on it's own mops up Premier. Apple doesn't have any products competing against Photoshop, Flash, Dreamweaver, etc so that's not even an issue but for video, audio and DVD production, Adobe is lame. Sure, Premier has power and some good plugin support (After effects is pretty sweet on it's own, that's for sure!) but it just doesn't handle EDLs or a modular workflow well at all and that's a pretty major issue in my book... Anyhow, just my $0.02 worth...
I don't think your statement is accurate. That may have been true 10 years ago but not today. Many professional firms now use Windows for their video production. I worked for several years at a magazine publisher and we made the Mac to Vista jump. The higher end 64-bit Windows machines really outpace the highend Macs. I don't care one way or another. However, I predict that when Win 7 ships that we will see even more pros shift to the Windows platform.
Yes, Adobe is bloat, but it is what most people use.
Well unless you changed OS and switched to an AMD opteron 8 way motherboard. 40 cores, now, how many cores does the pro have? ah yes, 8, or 16 logical ones.
I kid, I kid.
I have more fun with my Mac Pro and Photoshop than I do with ETQW. Macs are a hoot.
Maybe I should install Spore in Parallels and see if it works any better...
At least no trees died in bringing us this information.
The rest just clearly shows that the game is not optimized for OS X / nix, and that isn't really Apple's fault, it's the game company.
It's also interesting that a 2.26 Quad Core Xeon with non-overclocked graphics comes so close to the 2.93 Core 2 Quad and the overclocked otherwise identical video card. This despite the heavy correlation between single core speed and gaming speed...
:)
Rather than try for the highest fps possible, if my computer takes a game above 60 fps (which it does for all but the newest games), I just keep cranking up the settings until runs between 40 and 60 fps and looks as good as possible.
The argument that Mac Pros are the lesser because they can't do games as fast as a cheap PC is like arguing lolcats is just as important to science as the study of predators on the African savannah.
http://www.nvidia.com/object/product_geforce_gt_120_us.html
There's the link to the nVidia site on it if you would like...
To me its strictly for creative work.
Id rather use a game console to escape too ;)
1) Why would you want a Mac or even a PC to play games when you can use a much cheaper console (XBOX or PS3)? I can only think of a few games that are not on consoles.
2) Comparing a Mac Pro and a PC is almost like comparing a truck to a sports car. The truck is very powerful but not that fast, the sports car is fast, but cannot do much more than that. The point is, they are for different use. Does it matter if a Mac Pro plays game at best speed? I don't think so.
To the defense of this article, this is part of the Crave section, which is not about making economical sense.
Just a thought: as this column likes senseless comparison, next time less compare the performance of the computer to the consoles, using similar textures, resolutions, etc.
BTW I just got the new 2x2.93ghx 12GB RAM ATI4870 4TB HDs for $7,000. I don't intend to use this to play CoD4 on Windows64. I have my good old XBOX360 for that ($300). And it's great.
2. I agree, the Mac Pro and a gaming PC will hardly be used for the same purpose. Though, I do occasionally edit HD video on my gaming PC and my Q6600 does decently enough in converting the raw .avi files into videos at 1280*720.
In either case, it would be nice to find a benchmark-worthy game that's more recent. Keeping my eye on Rage for the future.
I'm not saying that the results we're seeing here aren't real, I'm just saying that citing hardware differences as the source of this result makes almost no sense. (I strongly suspect that there are some drivers issues with the bootcamp configuration.)
- by sting7k March 25, 2009 10:26 AM PDT
- I'm kind of surprised at the gaming scores for the Mac Pro. With those processors and all it should in theory be able to beat a Core 2 Duo system. More surprising is that the scores increase when you are using the exact same setup but running Windows instead.
- Like this Reply to this comment
-
-
- by 1363nd0f1337 March 25, 2009 10:40 AM PDT
- I doubt they do put as much effort into OS X performance as the user base for OS X gaming is much smaller than Windows gaming. Either that, or it is harder to get the game to interface with OS X. Honestly, most of my buddies that do run Macs usually end up playing all of their games on Windows machines, save for one who plays WoW on his MacBook in class.
- Like this
-
(37 Comments)So is it that OS X doesn't play nice with games or that game developers just don't put the effort into OS X optimizing as they do Windows?