Aces High Bulletin Board
General Forums => Hardware and Software => Topic started by: Vipermann on November 19, 2003, 11:55:33 AM
-
An AMD Athlon Barton 3000XP 400FSB
or a
P4 3.0 ghz 800FSB
Which would you choose for gaming?
-
The latter. 800 FSB is the kicker.
-
Overall I'd say the P4 is a slightly faster CPU, but the XP 3000+ is quite a bit cheaper. For gaming though, they are very close and by far your main bottleneck with either of these CPUs is going to be the video card. They both have the necessary power to push a Radeon 9800 Pro or GeForce FX 5900 card near its limits. There are certain games that are faster on P4s (Quake 3, Ghost Recon, Jedi Knight 2) and others that are faster on Athlons (Serious Sam 1 & 2, Unreal Tournament 2003), but overall those two CPUs are close for gaming.
You are going to need DDR400 (PC3200) memory for both for best performance. AMD recently adjusted the speed of the 3000+ model to better compete with the 'C' (800 MHz FSB) P4s. The original 3000+ was a 2167 MHz 333 MHz FSB part, the new 3000+ is a 2100 MHz 400 MHz FSB part and is significantly faster than the old 3000+ overall.
-
Either of these will be paired with
Radeon 9800XT
and
1GB OCZ Dual Channel DDR4200 RAM
I'm just having trouble deciding which CPU to go with. I have both here, the AMD with an Nforce2 board and the P4 with the Canterwood 875P.
-
Tough call from a gaming standpoint alone. I'd say if you do a lot of video encoding, go with the P4. Engineering/scientific apps, go with the Athlon.
-
If you're going for a top-end CPU, you should consider the AMD Opteron or FX chips.
-
Guys...the measly 400 FSB on the Athlon will make a big difference. It might be cheaper, but it's going to be slower for high-powered games like AH. The video card is not always the bottleneck.
The Athlon FX seems OK, but too expensive. Once they drop the price it will become more interesting.
The best bang for the buck right now is an overclocked 2.4C P4, with PC4200 RAM. The 2.4C will easily clock up to 3.2+, and you will get a 1066+ FSB with that. That will beat or at least match a stock Athlon FX, for games.
-
Ok i have a 2.4c cpu it overclocks stable to 3.0G and by no means will every 2.4c "easily" overclock to 3.2g.
I am at 1:1 ratio at 1004 mhz FSB and memory speed.
However having said that if you check the AH2 beta 4 thread i think you will see i am getting the best frame rates by a large margin, using a 9800 pro.
However i do not know whether anybody using a 3000 XP has posted any results.
http://www.hitechcreations.com/forums/showthread.php?s=&threadid=101278
-
You may need to raise your Vcore or DRAM voltage, or you may not have adequate cooling, or your RAM ma not be able to keep up. Try a 3:2 OC, and see how far you can go. Or perhaps you got unlucky, but given the ad hoc stats I'd say that was unlikely. It is well known that almost all recent canterwood P4s clock up to 3.5 or 3.6 GHz; hence my comment that it easily clocks up to 3.2+.
-
214--
I see you have a Max3 board. I have the same board. That board has a well known bug which prevents proper operation when you set Vdimm > 2.8. This can limit how much your memory overclocks. If you set the ratio to 3:2, you'll probably find that the processor has a lot more headroom.
Go to the abit forums (abit-usa.com) to see more about the max3 problem.
-
I used the Abit max3 board and with proper cooling you can see the results
(http://home.comcast.net/~c.hambleton/wsb/media/134502/site1042.JPG)
-
Cool. What RAM and Vdimm settings, mrblack?
-
Mushkin pc3500 with the BH-5 chips.
Ram settings where at spd voltage at 2.8
I was using the prometiea phase change cooling system
http://www.chip-con.com
-
Originally posted by mold
Guys...the measly 400 FSB on the Athlon will make a big difference. It might be cheaper, but it's going to be slower for high-powered games like AH. The video card is not always the bottleneck.
you should read THIS thread:http://www.hitechcreations.com/forums/showthread.php?s=&threadid=99359
the 400 vs 800 thing is better explained.
800fsb doesn't always mean faster if the architecture is different.
-
Originally posted by mold
214--
I see you have a Max3 board. I have the same board. That board has a well known bug which prevents proper operation when you set Vdimm > 2.8. This can limit how much your memory overclocks. If you set the ratio to 3:2, you'll probably find that the processor has a lot more headroom.
Go to the abit forums (abit-usa.com) to see more about the max3 problem.
I heard about that mold. is that all the max3 boards or just some of them. I thought that there were 'broken' oards and good ones. I havnt testem mine yet, but I sure hope I didnt get a lemon?:mad:
-
It's not at all valid to compare Intel and AMD CPUs by their front side bus speeds alone. They use totally different bus protocols and have very different architectures which put different types of stress on their main memory bus. (Check out my PC Architecture thread.)
The AMD 400 MHz FSB and Intel 800 MHz FSB are actually more alike than you would think. Both offer the same latency, because the fundamental clock is 200 MHz for both (DDR vs QDR signaling for 400 and 800 MHz effective). The Athlon (K7 & K8) architecture depends a lot more on low latencies (6 max instructions in flight - lower clockspeed), versus the P4 which depends more on high bandwidth for best performance (4 max instructions in flight - higher clockspeed). I'm not even going to get into the Athlon 64 and Athlon 64 FX, because they don't even have a traditional front side bus - they have an on-die memory controller. The system Hypertransport link runs at 800 MHz DDR for 1600 MHz effective. Again, check out my PC Architecture thread for more on this.
I wish it was a simple topic to discuss, but giving advice to buy one CPU over another because of its front side bus speed is somewhat like recommending one car over another because of the maximum RPM their engines can turn alone.
I guess it's worth talking about the Athlon 64 3200+ or the Athlon 64 FX Model 51, as these are without a doubt faster than the P4 3.2 'C', but they are probably out of the price range and are in very short supply.
Mold, if there's one thing the Athlon 64 does well, it's gaming. There's no way a 3.2 GHz P4 (even with a 1066 MHz FSB) will beat a stock Athlon 64 FX 51 for that. Gaming IS primarily limited by the video card, but beyond that memory latency and bandwidth are key. The Athlon 64 series CPUs have an on-die memory controller, which offers dramatically reduced latencies and also improves bandwidth. The closest thing they have to a tradidtional FSB to main memory is part of the CPU itself and runs at the processor clockspeed. The system link to the video card and rest of the system is 800 MHz DDR - 1600 MHz effective.
-
Originally posted by bloom25
Mold, if there's one thing the Athlon 64 does well, it's gaming. There's no way a 3.2 GHz P4 (even with a 1066 MHz FSB) will beat a stock Athlon 64 FX 51 for that.
While I think all the stuff you say is interesting and has merit, what it comes down to for me is price/performance. Not theoretical performance, but actual. I don't mean to belittle all the info you posted, but I have seen some game benchmark comparisons between stock Athlons 64's and stock 3.2 P4's, and the P4's are very close. They even beat the 64 in some cases. And when you OC the P4, it just sweetens the deal. I still stand by my statement that the best deal now is an OC'ed 2.4C, and that this will give you essentially top-of-the-line performance today.
Furthermore, I believe the effects you discuss between the differing architechtures and how they utilize RAM are second-order effects. And I think the test data bears me out in this. Game benchmarks I find to be a realistic test of system performance, if suitably randomized--since they fully measure actual usage.
Edit: However, I think you are correct that my focusing on just the 400 vs 800 is the wrong approach.
-
Originally posted by WhiteHawk
I heard about that mold. is that all the max3 boards or just some of them. I thought that there were 'broken' oards and good ones. I havnt testem mine yet, but I sure hope I didnt get a lemon?:mad:
It looks like it's all of them. In the abit forums, there hasn't been even one case of a board working properly above 2.8 vdimm. The good news is, it is looking like Abit is going to solve the problem and replace our boards under warranty.
-
The 2.4 'C' type P4 is definately a very good deal and an excellent overclocker. I've done quite a few of these on Asus P4P800s. If you plan to overclock, it is certainly the best bang/buck ratio CPU.
The 3.2 GHz 'C' type P4 is well matched with the Athlon 64 3200+ (single channel) in a lot of gaming tests, but you said the Athlon 64 FX (dual channel memory and 200 MHz faster) in your post. The Athlon 64 FX is much faster (and a LOT more expensive) for gaming than the standard Athlon 64 3200+.
Utilization of memory is hardly a "second order" effect. Modern CPUs spend well almost 90% of their time waiting for data and/or instructions from DRAM. Memory latency (or the amount of time the CPU spends waiting for data from either DRAM or the hard drive once it is requested - measured in nanoseconds or CPU clockcycles) is BY FAR the number one most critical bottleneck on system performance. Bandwidth, or the maximum amount of data that can be transferred per given unit of time, is secondary to latency for almost all applications and certainly nearly every game. Video encoding is probably the one application where bandwidth is more critical than latency, as would be most vector operation intensive code.
Lets go into more detail here because I think this will be helpful for others to read later:
The bandwidth available from a single DDR400 memory channel during a burst mode transfer is 3200 MB/second and a single channel is 64 bits wide. Dual channel DDR400 offers a 128 bit wide memory path and 6400 MB/second maximum bandwidth. The P4 'C' type "800 MHz" effective FSB can offer 6400 MB/sec bandwidth and the Athlon XPs 400 MHz FSB can offer 3200 MB/sec bandwidth - BUT the latency is the same because both use a 200 MHz fundamental clock. Why is this important? If an Athlon and P4 issue a memory read request that request must travel over the FSB to the chipset Northbridge. That request is then processed by the Northbridge and translated to the physical location in main memory. The command for a read is sent to DRAM, the row is activated (RAS), and finally the column (CAS) is addressed and the data from memory (64 bits, 1 row, at a time from a single memory module) is placed into a Northbridge buffer. The contents of that buffer are then sent back over the FSB again to the CPU and placed in the CPUs L2 cache (usually) and finally the CPU can directly work with that data. Obviously this takes a considerable amount of time. How long? Depending on the data requested (32 bit, 64 bit doubles, 128 quad doubleword, or 256 bit vector) between 50 (32/64 bit - 32 bit is transfered as 64 bits and the top 32 bits are simply discarded) and 300 nanoseconds (256 bit). That's up to 900 clock cycles on a 3 GHz CPU for 256bit SSE2 vector data! Now, consider this, the P4 can issue 4 instructions per clock - 2 memory read/writes PER clock cycle and 2 more integer/floating point executes. The Athlon/Athlon 64 can do 6 per clock - 3 read/writes and 3 int/fp instruction. So, a 3 GHz P4 that issues 2 256 bit SSE2 vector data read requests could waste up to 1800 useful clockcycles doing essentially nothing. For the Athlon 64 at 3 GHz (doesn't exist yet) it could waste up to 2700 clock cycles if latency was the same for its main memory. Obviously memory latency is absolutely critical, and even more so for the Athlon. This is why the Athlon 64 (K8) CPUs have an on-die memory controller, because it offers vastly reduced latencies. What about bandwidth, well I kind of answered that already when I said the CPU can request 32 bit up to 256 bit (for SSE2) pieces of data. A single DDR400 channel can transfer a whole 64 bits per clock and a dual channel setup can send 128 bits per clock. Most memory accesses are 32 bit or 64 bit. That's why having dual channel memory doesn't always result in greatly improved performance. Dealing with dual channels can also slightly increase latencies. From what I said previously, I think you can now see why AMDs "400 MHz" Athlon XP FSB and the P4 'C's "800 MHz" FSB offering the same latencies are important. If you are dealing with 32 or 64 bit reads/writes they both have sufficient bandwidth and the extra bandwidth offered on the 800 MHz FSB is wasted.
You guys might be interested in this page of Aces Hardware P4 EE, Athlon 64 review: http://www.aceshardware.com/read.jsp?id=60000258 Just look at how much the on-die memory controller helped the Athlon 64 CPUs in both latency and to a lesser extent bandwidth. This is the sole reason why the Athlon 64 series CPUs are so good for gaming, the reduced latency. It also explains why Quake 3 favored the P4 so much, look at the 64 bit latencies of the 'C' type P4 on i875 Canterwood versus the Athlon XP 3200+ on nForce 2. It's plainly obvious that the memory controller on the i875 is better than the memory controller on the nForce 2 chipset.
-
Many are talking RAM, I am using:
Corsair 1GB TWINX 1024-4000PRO COR RT
BOY talk about performance!!
Anyone else using this type?
-
I have no problem with either CPU, but lately the boards that I have with Intel-based chipsets have been more reliable over time (years) than Via, SiS, and other chipsets (No NForce chipsets in the mix, however).
That alone makes me lean more towards an Intel solution.
-Llama
-
Mold i am well aware of the Abit bbs and all the problems reported with the Max3.
I am using PC4000 memory.
I have given the cpu as much juice as i care too with air cooling 1.65 volts bearing in mind i do not want it to crap out next week.
In fact it is not prime stable even at 3g unless i up the cpu voltage.
It will post and load windows at 3.1 but nothing will make it prime stable.
After early experiments it was clear to me that any memory voltage above 2.8v is unusable due to wild fluctuations and i again do not care to try running 3.2v when memory is rated for 2.8
Using the earlier bios 1.1 i think, it would not even post at 5:4 divider at any bus speed, and at 3:2 max fsb was only 225 when i could get 250 fsb at 1:1
1.2 bios was no improvement at all.
Now with 1.3 some things are better, i can now boot at 5:4 but only to the same fsb of 250 ( although to be honest it craps out somewhere between 250 and 258 at 1:1 and 5:4) I just cant be bothered to find the exact setting for the gain involved.
Bigtoes 1.2 modded bios was a waste of time, sure he has uncovered some extra features but so did I when i modded my own version of 1.3
Anyway point is all 2.4c's are not virtually gauaranteed to be hitting 3.4 to 3.6 and the sites i am reading in the uk seem to be suggesting 3g is becoming the norm.
Although there is some talk of a new stepping hitting 3.6 on std volts, they seem to think they are 3.2 EE's that failed the grade.
So thats where i stand, not sure if its the board, the memory, the cpu or even a combination of them all holding me back.
And i do not have money to speculate and go buying different memory or another cpu.
Still i gotta be happy bought a 2.4 and getting 3.0 performance totally stable, others may have better but its the luck of the draw.
-
Originally posted by mold
It looks like it's all of them. In the abit forums, there hasn't been even one case of a board working properly above 2.8 vdimm. The good news is, it is looking like Abit is going to solve the problem and replace our boards under warranty.
o sht.. i damm near had a anyerism installing the one I got and it dodnt blow up when i turned on the power. But keep posting, i'll certainly get the new one.
-
bloom... my eyes... ouch ><
Ill take yer word for it :D