Axiom Home Page Forums General Discussion Stereo A musical revelation

In reply to:

---

The PowerPC G5 can pump through more than 200 in-flight instructions at a time, a whopping 71% more than the 32-bit Pentium 4

---

To my knowledge that is only because that is 32 bit, and because the P4 only does 6 IPC (instructions per clock) compared to a 32 bit aXP's 9; compare it to a 'real' chip - ie the AMD Opteron (sledgehammer same as fx-51).

And I'm pretty sure you are going to say that Mac's already have their 64 bit jaguar os x out, and the version of WinXP for the Sledge/Clawhammer is still in late beta stage. Well, even in 32 bit mode the Sledge/Clawhammer is faster than the G5. And when the 64 Bit winXP comes out, it will be fully 64 Bit, unlike the "64 bit" os x which is not fully 64 Bit.

Just noticed that I forgot to point out that the better of the two Mac setups included has two 2.0 GHz G5 chips, not one; whereas the 2.0 GHz single Opteron processor (Sledgehammer core based) obliterates it on all counts.

The AMD is not king of the PC. Intel is. If basing on 1 or 2 apps fine. When based on overall power the Intel P4-3.2 Extreme Edition does a better job.
http://www6.tomshardware.com/cpu/20030923/index.html

Here are some benches on PHotoshop stuff.
http://www.barefeats.com/pentium4.html

Apple marketing may have something to do with this so I take these results with a grain of salt.
http://www.apple.com/powermac/performance/
http://www.apple.com/powermac/

Yes you are right about the OS issues. A CPU can only be as fast as the OS and its components. One thing to note I do not think or running the G5 as a Apple machine but use it as a Unix server.

And yes I an wrong about the G5. The fastest chip in a multi CPU configuaration to serve the most processes at any one time and the quickest is the Itanium 2

I did find the add the the AMD kicked the G5
http://www.pcworld.com/news/article/0,aid,112749,pg,8,00.asp

But I personally would still use a IBM multiprocessor RISC box (Power PC chips) running Unix to handle 40,000-100,000 credit card transactions per day which is stored in oracle database than have a multiprocessor AMD running Win2003 server with Oracle handling those same transaction information.

Wow - I never would have thought the P4 EE was better - but i guess that huge 'L3' cache has to be good for something. (Hasn't become an ATI vs nVidia where an underdog becomes the faster one)

Itanium 2 - Heh I also forgot about this - How many MHz are they up to now? Last I remember was seeing an 800 MHz Itanium that just beat almost everything else.

Yes, I am one of them. I loves me my Mac.

I also use and appreciate PCs. I like them a bunch, but I can't say I love them.

Let me rephrase that. I prefer the Mac "user experience" to the PC (Windows).

Indeed, but I was only joking . A few of my friends are MAC users, and it's almost like an inside joke with us that we each point out and laugh when the other one crashes or doesn't do something it's supposed to.

I definately agree that the Mac experience is potentially better to many people because there's just less potential problem areas to mess around with. I personally like to tweak my system (and overclock it) to oblivion, so I just need direct access to everything heh.

[/me checks my temps and rejoices that they are still under 45 C, while my poor aXP 2000+ Palomino core is at 1.95 vCore instead of 1.75 stock running at an estimated 2300+]

Hope your running a Thermaltake or a Zalman. If your using stock your chip will start flaking out around 6 months or so. Used to overclock everything but with todays technology its really all in the videocard...not so much in the CPU. Yes a faster CPU does help but I found over the years that if you overclock the CPU vs videocard. The CPU get flaky first. I find that with todays chip they did not increase dramatically in speed in comparison to older chips. ie when I o'clock a P700 to 1066 or something like that with a GForce2 GTX and I would get 7000+ on 3dmark. The newer chips only go up to 8,000 and 9,000 and those are chips in the 1.6 and 1.8 with Gforce4 cards. Youd think it would jump exponentially but it doesnt. It do not know with the 2.2 or higher if the score are much higher...ops looked it up and they were in the 17,000-20,000...wow the Athlon 64 3200+(2572) was in the 30k..phenominal.

Swiftech MCX-462 with a Delta 80 MM, but it is controlled by a rheostat, so it's not too loud, unless I want to benchmark and overclock to its fullest.

Vid Card: 9800 Pro (but I leave at stock, as this is my second 9800 Pro because my last one's RAM went bad after about 2 months of OCing - and it had ramsinks on it )

My 3Dmark score isn't too high right now for 2k1 (only about 14k - but that's severely CPU limitedby my 'old' 2000+). In 2k3, I get 5600 (because I refuse to OC my 9800 Pro now), and with my 9700 Pro (RIP from volt modding), I would get 5000. I wish I could get that 30k in 2001, but my CPU now is low end, and as most games and apps are CPU limited like that (once you have a 9800/9700/NV30/NV35), it would be practical speed. But I guess I make up for it by running with all the eye candy enabled (4x Anti Aliasing and 8x Anisotropic Filtering). Makes my video card the bottleneck in the end, instead of the CPU.

In reply to:

---

when I o'clock a P700 to 1066 or something like that with a GForce2 GTX and I would get 7000+ on 3dmark.

---

That is incredible! Especially for a card that cannot do the Nature test (Game Test 4). That would be another few thousand right there. From direct-x 7 compliant GPU's such as that GF2 to a DX8 Compliant one, there is such a gain in 3DMark score (for 2k1), but after that it's kind of not important, as there are no tests in 3DM2k1 that it cant do (3DM 03 is another story ).

That is why I mentioned that the CPU in between 1gz-2gz did not step up the speed for 3d games (3d mark test) I was totaly floored when the good ole o'clocked 700 was doing so well. They were built like tanks. They can take the abuse. the new CPUs, RAM and bus are a whole new ballpark. Now they are where they are supposed to be.

As with your Radeon. It is the few cards that can run that fast by having 8x anti-aliasing and 8x Anisotropic Filtering as of 4 months ago where I was looking into vid card...hence my Radeon 9700Pro purchase. 4 months is long time and Nvidia might or might not have a Radeon killer by now. But since my main purpose of using my HTPC is for video processing...hands down the Radeon series are steps ahead of Nvidea. I just wish my 9700Pro didnt have that extra piggy back power drain. It doesn't work too well if your building a HTPC for low noise and low power.