Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Wow, the PM 2.5GHz G5 was spanking those little suckers around. Even with a new Opteron and whatever, the PM should still be beating the dead horse in pretty much every category.

*cue evil laugh*

And what's the cost difference on these machines? Dare I say that the PM may actually be cheaper than some of the competition?
 
The DP2.5Ghz G5 will be cheaper that the DP3.06GHz Xeon, and the DP2GHz Opteron (when bought from a company, not home builds) I'd have thought.

These benches show the new top-of-the-line in a very good light indeed :)
 
hcuar said:
Yeah... but... They didn't bench mark the 3.6 Ghz P4??? Only seems fair. :cool:

well the test skiped a lot of the higher end chips from intel and AMD. Then eniter FX line was skiped along with the extrem edition.
 
Unfortunately, the benchmarks don't include the latest parts from Intel or AMD. There's a 90nm Xeon MP at 3.6ghz now and a 90nm 2.5ghz Opteron is either released or about to come to market. Compare that to the fact that the tested machines were 3.06ghz and 2.0ghz, respectively.

Also, the chipset used could make a huge difference. The G5 is fast, but we really do need to keep up on the cutting edge if these kinds of claims are going to be made.
 
jared_kipe said:
Can we start airing the "world's fastest, most powerful PC" advertisements now??

Apple was in hot water with this before. By the time they get a new ad out like this, newer and faster PCs will be out. That's what happened last time. When the dual 2.0Ghz was introduced, I am certain that it was the fastest PC. However by the time it shipped and the ads for it played, newer procs had hit the market.
 
Timelessblur said:
well the test skiped a lot of the higher end chips from intel and AMD. Then eniter FX line was skiped along with the extrem edition.

Both the Athlon FX and the P4EE (which is discontinued, incidentally) get slaughtered by DP systems with good chipsets. If you're going to do a comparison, at least be fair and let the x86 side take advantage of having more than one processor, too.

Hell, you can build a pretty nice dual-Opteron rig for the cost of using a single Athlon FX.
 
thatwendigo said:
Unfortunately, the benchmarks don't include the latest parts from Intel or AMD. There's a 90nm Xeon MP at 3.6ghz now

I looked around and no one I can find sells the 3.6Ghz Xeon yet, and I have no doubt that it would beat the 2.5Ghz G5. But by the time it comes out, Apple will have certianly upgraded the G5. Comparing Apple's current offerings to stuff that isn't out yet isn't fair.

thatwendigo said:
and a 90nm 2.5ghz Opteron is either released or about to come to market. Compare that to the fact that the tested machines were 3.06ghz and 2.0ghz, respectively.

According to pricewatch, the 2.4Ghz Opteron is over $1400 (for just one chip!)
The two chips would cost about as much as a complete Dual 2.5Ghz G5 system...
 
the other hard part is there is no real good software out there to compare them since it either opimized for one of the other. None of those test I have ever seen be used to compared there diffence x86 processors but I have seen in the past on Apple computers.

If apple made a bold claim there would be lawsuits dealing with false advertiment because they pull out some test that are not good for the Apple computers but good for PCs.
 
Dippo said:
I looked around and no one I can find sells the 3.6Ghz Xeon yet, and I have no doubt that it would beat the 2.5Ghz G5. But by the time it comes out, Apple will have certianly upgraded the G5. Comparing Apple's current offerings to stuff that isn't out yet isn't fair.

Yes, but the hardware benchmark sites are getting their hands on both parts and putting them in test systems, showing what they're capable of. The parts may not be mainstreamed into the OEMs just yet, but that doesn't change the fact that they're coming soon and the dual 2.5ghz G5 needs to be compared to them, too.

Believe me, I would love to have Apple possess the speed crown, but we need to keep things honest and open.

According to pricewatch, the 2.4Ghz Opteron is over $1400 (for just one chip!)
The two chips would cost about as much as a complete Dual 2.5Ghz G5 system...

Yeah, top of the line Opterons aren't cheap. Then again, the dual 2.0ghz part that's tested by Barefeats is $350+ a pop, which means it could be built into a professional system for (quite possibly) less than a G5.

I need to wake up more before I try to talk about this. :D
 
thatwendigo said:
Unfortunately, the benchmarks don't include the latest parts from Intel or AMD. There's a 90nm Xeon MP at 3.6ghz now and a 90nm 2.5ghz Opteron is either released or about to come to market. Compare that to the fact that the tested machines were 3.06ghz and 2.0ghz, respectively.

Also, the chipset used could make a huge difference. The G5 is fast, but we really do need to keep up on the cutting edge if these kinds of claims are going to be made.

Barefeats should have testing for at least the XEON MP @3.06Ghz ready in two weeks as well as testing for the Opteron 250. The G5 2.5Ghz MP is allready leading by 10-50% faster in all tests compared to the current Xeon MP 2.4Ghz and Opteron 246 systems so I think it will fair well against Opteron 248 and 250 as well as any Xeon system that is currently available. Would be interestin to see some Game scores especially with the new Geforce 6800 Ultra.
 
In the first test, they use a Xeon 3.06Ghz, but in the remaining tests they use a Xeon 2.4Ghz. In any test, you should retain the same system configuration so you can take all the results in and compare evenly.

I wonder why they changed systems in the middle of testing.
 
missing a lot of newer cpus from the pc world. the real answer to G5 2.5 performance will be what does Doom3 look like on it. will just have to wait on the mac version but i do think that macs are great for work applications(boring) but for gaming they have allways come up a little short.
 
tomf87 said:
In the first test, they use a Xeon 3.06Ghz, but in the remaining tests they use a Xeon 2.4Ghz. In any test, you should retain the same system configuration so you can take all the results in and compare evenly.

I wonder why they changed systems in the middle of testing.

This is a work in Progress. Barefeats will add test results for other systems as they become available. Rob Art Morgan does not own all those machines so he tests when he has an oppurtunity in many cases that means that not all tests can be done at once. Read the entire Page. Xeon 3.06 and others will be posted as figures become available.
 
Little Endian said:
This is a work in Progress. Barefeats will add test results for other systems as they become available. Rob Art Morgan does not own all those machines so he tests when he has an oppurtunity in many cases that means that not all tests can be done at once. Read the entire Page. Xeon 3.06 and others will be posted as figures become available.

Considering a 3Ghz processor is the norm currently, I have to disregard most of the tests until the latest, or near latest, proc is used in all tests. Why not compare a 2.5 G5 to a Pentium 4 2Ghz? Because it doesn't represent a comparable platform, and that's my basis for disregarding the other tests as well.
 
tomf87 said:
Considering a 3Ghz processor is the norm currently, I have to disregard most of the tests until the latest, or near latest, proc is used in all tests. Why not compare a 2.5 G5 to a Pentium 4 2Ghz? Because it doesn't represent a comparable platform, and that's my basis for disregarding the other tests as well.
I concur P4s @ 4.0 are on the way, so are the faster Athlons. I remember seeing a Macworld test(thanks Rdowns)and a 2.2 athlon fx51 was sometimes faster then the dual G5 at 2.0. it depends on applications you are running but if into gaming you are then a PC is pretty fast for many reasons. though its a week old this Alienware has showed some awesome gaming performance in Ut2004 & Doom3 & Far Cry. :)
 
Dont Hurt Me said:
missing a lot of newer cpus from the pc world. the real answer to G5 2.5 performance will be what does Doom3 look like on it. will just have to wait on the mac version but i do think that macs are great for work applications(boring) but for gaming they have allways come up a little short.

I don't think the real answer to the G5's performance will come from the results of a Doom3 timedemo. Modern games are heavily optimized for x86 and all the associated multimedia instruction sets. Code Doom3 from the ground up with the G5 and OS X in mind, and then we might have the real answer to the G5's performance.
 
frozenstar said:
I don't think the real answer to the G5's performance will come from the results of a Doom3 timedemo. Modern games are heavily optimized for x86 and all the associated multimedia instruction sets. Code Doom3 from the ground up with the G5 and OS X in mind, and then we might have the real answer to the G5's performance.

Uh I don't think "optimized for x86" is correct, but you're right--the code is different.

OS X is not very game friendly. Probably through no fault of its own, the video drivers are usually written by Apple and optimized for Apple's applications/OS.

Also there's no real direct-x (for obvious reasons). The mac version of DX is like 7 or 8 equivalent.

This is why SPEC benchmarks are performed on the same OS.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.