Sign in with
Sign up | Sign in
Your question

AMD XP 3000+ vs AMD 64 3200+

Last response: in CPUs
Share
July 28, 2004 2:34:23 PM

Hi, I currently have a AMD 2200+, Radeon 9700 Pro, 1 gig 2700 Ram and a Soyo mobo with Via KT400 chipset. I am trying to decide if i want to just go for the new 3000+ xp for less than 150 bucks or buy a new mobo and an amd 64 3200 and use the current RAM i have. It is mainly for gaming and I dont know if the extra money is really worth upgrading to an AMD 64. I figure that when 64 bit windows and applications come out ill have to upgrade again anyways but that could be awhile in the future. So i just want to know if performance wise for games the amd 64 is much better than an xp. Thanks for your help.

More about : amd 3000 amd 3200

July 28, 2004 3:35:10 PM

Quote:
So i just want to know if performance wise for games the amd 64 is much better than an xp.

AMD64 3200+ would really kick the butt of any Athlon XP system. Athlon XP 3200+ can be compared to Athlon 64 2800+. So an Athlon XP 3000+ would be equivalent in performance to an Athlon 64 2600+/2700+ (by the way these processor doestn't exist).

My personnal recommendation : keep your AXP and overclock it if it's not locked (change the multiplier in the BIOS) or buy the A64 2800+/3000+ they have a better price/performance value than A64 3200+.

--
Asus A7N8X / <font color=green><b>AMD Sempron 2800+</b></font color=green> (tbred @ 167x12)
Kingston DDR333 2x256Megs
<font color=red>Built by ATI Radeon 8500LE 128Megs</font color=red> @ C:275/M:290
July 28, 2004 3:42:09 PM

Thanks for your help. I havn't done any overclocking just read up on how to do it. My CPU temperature already seems fairly hot and my computer is in a cramped space so air flow isn't that great so i probably won't overclock anytime soon. Is the 64 3000+ a better buy cause it can be overclocked and outperform the 64 3200+. Or is it that the 3200+ just isn't that much anyways without overclocking.
Related resources
July 28, 2004 3:53:23 PM

Most Athlon 64 will run at +200Mhz from their original speed on air cooling. So getting a 2800+ and running it at 3000+ is an easily reachable target. So getting a 3200+ and overclocking it to 3400+ is possible too.

It mostly depend on your budget. If you have the money to buy an Athlon 3200+, go for it! If you are a bit tight on budget and to have a 64bit ready system, the best choice is the 2800+. And you may heard that AMD have just launched Sempron CPU's to replace Duron and Athlon XP. the Sempron 3100+ will be quite interesting if you don't want a 64bit ready system yet. The performance of the Sempron 3100+ is nearly equal to Athlon 64 2800+ in most situation. So, if you want to give 64bit support, the Sempron 3100+ is cheaper than the Athlon 64 2800+ so it has a better price/performance value.

By the way, Sempron 3100+ is a Socket754 CPU, so you would be able to switch to an Athlon 64 in the future.

But, the best choice is still the Athlon 64 2800+ for a "budget" mid/high-end gaming system.

--
Asus A7N8X / <font color=green><b>AMD Sempron 2800+</b></font color=green> (tbred @ 167x12)
Kingston DDR333 2x256Megs
<font color=red>Built by ATI Radeon 8500LE 128Megs</font color=red> @ C:275/M:290
July 28, 2004 4:00:47 PM

Thanks for the help again. I believe i will go for the 64 3000+ and a soyo mobo i saw that seems to work well with what I have. I have 3 ata hard drives (2 in raid) plus 2 cd drives so i need a mobo that has at least 3 ata133 slots with one being raid capable. If anyone has any recommendations they would like to share for motherboards or even a good fan since i'll have to get one of those too I'd like to hear them. Thanks.
July 28, 2004 4:30:15 PM

My expeierence has been sort of opposite. I have a computer with a 2800+ XP processor, and one with a 3000+ 64-bit processor. Using PCMark 2002, the 2800 scores close to 900 points higher than the 64-bit chip.
July 28, 2004 7:24:57 PM

Was the rest of the system exactly the same besides processor? That's kind of wierd.......I guess I'll try to do some more research/see what other peoples experiences are before I purchase anything. Thanks for the input.
July 28, 2004 7:53:58 PM

No, the two systems are different, but these scores are in PCMark which supposedly tests just the processor.

I checked out the PCMark FAQ, and they say the CPU test is "somewhat dependent on the memory speed." The 2800+ XP system is using 400 mhz RAM that is 2x512 chips running dual channel. The machine running the 3000+ 64-Bit chip is running 333 mhz RAM and is only 2x256 chips. Both boards have nForce hardware.

But in my opinion these systems are that drastically different. The 2800+ XP chip has the 333 FSB, so the 400 mhz RAM should be limited to the 333(someone correct me if I am wrong). Therefore these machines shouldn't be too drastically different, especially to do a CPU test.

I don't know what PCMark means by "somewhat." Obvisuly benchmarking scores aren't everything, becuase all chips have strengths and weaknesses.

I decided to do another benchmark trying a different program, to see if my results were any different. I use PC Wizard 2004, here are the results:

3000-64 2800-XP
Add/Multi SSE 2211 2525 MFlops
Div SSE 776 903 MFlops
SqRoot SSE 659 691 MFlops
Add/Multi 3dN 2486 3135 MFlops
Div 3dN 356 409 MFlops
SqRoot 3dN 241 283 MFlops
Add/Multi FPU 2080 2422 MFlops
Div FPU 109 139 MFlops
SqRoot FPU 56 63 MFlops
Whetstone 644 774 KWPS

Was kind of suprised my XP chip took the 64-bit chip in every catagory. Which makes me wonder why?
July 28, 2004 8:02:42 PM

Wierd.....maybe someone can clue us in as to maybe why that is so. Also, will it matter all that much if i go with an nforce3 250 or K8T800 chipset. I read some people praising nforce3 250 but they didn't say exactly why its good.
July 29, 2004 6:30:54 AM

Well I did some more checking and my 64-bit chip is the mobile version, which is a little different than the desktop version. The mobile version has more L2 cache, and generally includes cool'n'quiet to control CPU speed. I disabled cool'n'quiet and my PCMark02 scores did go up a little, but the 3000 64-bit mobile chip is still 570 points behind my 2800+ XP chip.
July 29, 2004 7:20:46 AM

Don't put any faith in simple synthetic benches like PCMark. Just about any real world app should run significantly faster on the A64, especially games (provided you use similar videocards obviously).

= The views stated herein are my personal views, and not necessarily the views of my wife. =
July 29, 2004 1:17:04 PM

Mr5oh if you had any newer games would you mind running playing them and maybe compare FPS for me. If not that's okay. Still not quite sure what I'm gonna do...... Keep posting guys it's all good information.
July 29, 2004 2:21:30 PM

Yeah I can do that, I will have to find some sort of tool to measure FPS, any suggestions?

The games I have been playing recently (on both machines): America's Army, Enemy Territory, Need For Speed Unground, and Red Faction (older game but still one of my favorites). All these games I don't really notice any difference between the two machines, both running 1280x1024 at full graphics, I don't notice any lag or anything, everything runs smoothly. I will try and get some actual numbers for you though (Maybe some numbers for Half-Life, and UnReal).


http://Mr5oh.tripod.com/pc.html" target="_blank">My Desktop PC
July 29, 2004 2:28:20 PM

Never used it myself, but <A HREF="http://www.fraps.com/download.htm" target="_new">FRAPS</A> is what you're after, I think.



---
Epox 8RDA+ V1.1 w/ Custom NB HS
XP1700+ @200x10 (~2Ghz), 1.4 Vcore
2x256Mb Corsair PC3200LL/1x512Mb Corsair XMS PC4000 2.5-3-3-7
Sapphire 9800Pro @412/740
July 29, 2004 2:50:40 PM

In americas army open up the console (~ button i think) and type show fps or show fps 1 and it will show you the fps your getting. I dont know anything for the others though. Thanks for the help, I appreciate it.

Also, try to use the same server if you are going to test it. The server you are at and amount of people playing will cause the fps to change.

<P ID="edit"><FONT SIZE=-1><EM>Edited by RaI3id_Squirrel on 07/29/04 10:58 AM.</EM></FONT></P>
July 29, 2004 3:27:49 PM

in AA i think that the console command is "stat fps"
July 29, 2004 3:37:55 PM

Yeah thats right, i must be confusing it with another game. So in console type stat fps and it will show you fps.
July 29, 2004 5:53:44 PM

here is a quick and easy fps check.
Download UTdemo348.exe from
http://www.unrealtournament.com/utgoty/downloads.php
i have the best luck with the FileFront link.
install, pick your 3d mode of choice and run the game.
from the top menu bar choose "Tools" and then "TimeDemo Statistics."
run a match from there and you can see your fps in the upper left
hand corner. Its very detailed and a great DX7 test.
Then under the menu "multiplayer" select "open location".
type 64.237.35.199:7778 and get your NOOB-arsse whipped.
=D Its fun! lol. I play as "THE~[]\/[][][]D[]-[]~"


AMD Athlon XP 2100+-=-1Gb PC2700-=-40Gb WDD HDD-=-
16x DvD-=-32x12x40 Lite on CD/RW-=-Floppy, "Who needs a damned floppy!"-=-ECS K7S5A Pro-=- Geforce 4 Ti 4400 128Mb 4x-=-
July 29, 2004 6:29:35 PM

Mr5oh, let me get this straight. You're comparing a desktop PC with an AXP2800+ and 1GB of dual-channel DDR400 to a laptop with an A643000+ and 512MB of single-channel DDR333 and you're even slightly amazed that the AXP in a perfect memory setup is performing better than the only slightly higher rated A64 that's being strangled by its platform?

<pre><b><font color=red>"Build a man a fire and he's warm for the rest of the evening.
Set a man on fire and he's warm for the rest of his life." - Steve Taylor</font color=red></b></pre><p>
July 29, 2004 7:27:19 PM

How much is the laptop setup really affecting benchmarking though? Obvisuly memory scores will be high on the dual channel setup? But as far as CPU testing I don't see how the chip is really being affected. It is my understanding that 64-bit chip has a far superior bus system, and in general* only the processor is being tested in CPU tests? So, yeah, I am wondering why there is a big difference.

My Desktop: <A HREF="http://Mr5oh.tripod.com/pc.html" target="_new">http://Mr5oh.tripod.com/pc.html&lt;/A>
July 29, 2004 8:08:03 PM

See, here's the thing. A processor can only process data if it <i>has</i> the data. If the memory system in the PC isn't able to provide data quickly, the processor will waste a lot of time waiting for those bytes. In real-world usage memory <i>greatly</i> impacts a CPU's performance. The laptop platform you have not only is single-channel instead of dual-channel (which cuts both the bandwidth in half as well as greatly increases latencies right there) but it's at a slower clock rate as well. Plus I'd be willing to bet that your AXP's memory latency settings are much tighter than your A64's, if for no other reason than because it's a laptop, and most laptop manufactuers pad those settings to ensure stability.

So what difference there is in raw CPU power between the two systems is being mitigated by the fact that the A64's memory system isn't allowing the CPU to run optimally where as the AXP's memory system <i>is</i> allowing the CPU to run flat out to the absolute best of its abilities.

It's like making one marathon runner run freely while his competition has to run through jello. Even if the second runner is faster in a flat out fair race, because of the situation this isn't a flat out fair race.

<pre><b><font color=red>"Build a man a fire and he's warm for the rest of the evening.
Set a man on fire and he's warm for the rest of his life." - Steve Taylor</font color=red></b></pre><p>
July 29, 2004 8:23:08 PM

I removed the extra 512 from the PC so it is no longer in dual channel. Here are the timings for both computers (Both are how they came, I never changed them). Should I even be trying to compare them.

Laptop - 2.5,3,3,7
PC - 2.5,3,3,8

I know you keep saying there is a huge difference, but the stats really seem to be pretty close, especially when I take away a stick of RAM in the desktop. The 64-bit processor is running 1600 mhz bus (correct?), while the XP chip is only 333 FSB. Is 400 mhz memory in the desktop be limited to the 333 of the processor?


My Desktop: <A HREF="http://Mr5oh.tripod.com/pc.html" target="_new">http://Mr5oh.tripod.com/pc.html&lt;/A>
July 29, 2004 8:55:17 PM

Well I have looked into things a little more. Since I am trying to decide 3000xp or 300064 this is what I'm thinking. Anyone correct me if I'm wrong. Since the xp has a higher clock rate it will probably perform in certain areas better that are dependant on it. However the 64 has a larger bus giving it access to more information at a time thus making it better in certain areas? Now I guess i need to decide what exactly those areas are and if that's what im looking for.
July 29, 2004 9:29:36 PM

I think for real world stuff you will notice the extra bus speed. I believe there is only like 100 mhz clock speed difference between the two chips, but the 64-bit chip has 4 times the bus speed. Figure there is roughly $60-70 USD between the two chips. Plus a little additional cost for the 64-bit's motherboard.

I believe it is going to come down to how much is the extra performance worth to you? Probably at least $100 extra to go the 64-bit route.

My Desktop: <A HREF="http://Mr5oh.tripod.com/pc.html" target="_new">http://Mr5oh.tripod.com/pc.html&lt;/A>
July 30, 2004 12:25:35 AM

Alright well I think I have made my decision. I can sell my current mobo and processor to someone I know for a good price so im gonna go the 64 route. Thanks again for all the input it wasn't an easy choice.
!