AMD Athlon 64 3200+ CPU Review

by Ilya Gavrichenkov
10/20/2003 | 11:01 AM

About a mot h ago AMD announced its new processor families based on AMD64 architecture. They were Athlon 64 FX and Athlon 64. At launch we posted an article where we took a closer look at the features and performance of the eldest desktop processor aka Athlon 64 FX-51 (see our AMD Athlon 64 FX-51 vs. Intel Pentium 4 Extreme Edition 3.2GHz: Clash of Strong Wills). However, at that time we promised that we will definitely return to this topic and test Athlon 64 3200+ targeted for a more mass market. Today we are happy to put our promise into life. We would like to offer you a new review of this particular processor. In this article we are going to talk in detail about the feature of AMD Athlon 64 3200+, its overclocking potential, and Cool’n’Quiet technology, which is typical only of this CPU.

For our review we purchased in a retail store a new boxed Athlon 64 3200+, so this article is not just an engineering sample review, but it looks at a real mass piece. I would like to stress that these processors are currently available in many stores for about $400-$450. Since the Socket 754 mainboards now cost around $150, it appears that an AMD Athlon 64 3200+ based system can really be  regarded as a High-End mass platform. Well, we are going to look at our today’s hero from this particular viewpoint.

Closer Look: AMD Athlon 64 3200+

First of all let me point out that Athlon 64 processor is exactly that particular 64bit processor for desktop systems, which AMD was going to introduce from the very beginning. Later on, when Intel announced faster Pentium 4 processors and provided them with a faster 800MHz bus and Hyper-Threading technology support, AMD had to quickly re-position a one-way Opteron processor for the desktop market having hidden it behind the Athlon 64 FX trade mark. However, the server origin of Athlon 64 FX will prevent this solution from spreading really widely and getting less expensive so far. So, it is exactly Athlon 64 that is intended to really promote the 64bit architecture into the market.

So far AMD has released only one processor model from the new Athlon 64 family with the 3200+ performance rating. This CPU works at the actual 2GHz frequency, which is 200MHz lower than Athlon 64 FX-51 supports. Later on AMD will also release a faster modification of its Athlon 64 CPU with 3400+ performance rating and 2.2GHz actual core clock frequency. However, this can hardly be expected to happen soon. Why so, you will understand later on.

Since we have already discussed in detail all the architectural peculiarities of AMD Athlon 64 in our article called AMD Athlon 64 FX-51 vs. Intel Pentium 4 Extreme Edition 3.2GHz: Clash of Strong Wills I would like just to briefly remind you of the basic things in a table below. The table you see contains specifications for Athlon 64 3200+ compared with those for Athlon 64 FX-51 and Athlon XP 3200+:

	Athlon 64 FX-51	Athlon 64 3200+	Athlon XP 3200+
Packaging	Socket 940	Socket 754	Socket 462
Frequency	2.2GHz	2.0GHz	2.2GHz
Manufacturing technology	0.13micron, SOI	0.13micron, SOI	0.13micron
Number of transistors	105.9 mln	105.9 mln	54.3 mln
Die size	193 sq.mm	193 sq.mm	101 sq.mm
Nominal Vcore	1.5V	1.5V	1.65V
Integrated memory controller	Dual-channel, 128bit	Single-channel, 64bit.	None
Supported memory types	Registered DDR400/ DDR333/ DDR266 SDRAM*	DDR400/ DDR333/ DDR266 SDRAM	-
ECC support	+	+	-
L1 cache	128KB (64KB for data and 64KB for instructions)	128KB (64KB for data and 64KB for instructions)	128KB (64KB for data and 64KB for instructions)
L2 cache	1024KB (exclusive)	1024KB (exclusive)	512KB (exclusive)
SIMD instructions support	SSE2/SSE/3DNow!	SSE2/SSE/3DNow!	SSE/3DNow!
AMD64 technology support	+	+	-

* - Note that since the memory in Athlon 64 and Athlon 64 FX is clocked relative to the core frequency, the actual memory working frequencies equal 100, 129.4, 157.1 and 200MHz.

In fact, Athlon 64 differs from its elder brother, Athlon 64 FX, only by the memory controller (and the packaging form-factor, of course). Although despite that, both CPUs are designed from one and the same silicon die. The memory controller of Athlon 64 is a single-channel one, and this is its weak point and at the same time an advantage compared with the Athlon 64 FX. The bad thing about a single-channel memory controller in Athlon 64 is evident: it is the lower theoretical memory bus bandwidth. Keeping in mind that Athlon 64 can work with DDR400 memory, the maximum bandwidth of the memory controller integrated into the CPU equals 3.2GB/s. It is twice as little as this parameter value by Athlon 64 FX. The advantage of the single-channel memory controller of Athlon 64 implies that unlike the controller of Athlon 64 FX, this one supports regular non-registered memory modules. These memory modules are certainly cheaper than the Registered ones, feature more aggressive timing settings and work faster than the Registered ones even if the timing settings are the same in both cases. So, if we put it into more formal words, despite the lower bandwidth provided by the Athlon 64 memory controller, the memory subsystem using it boasts lower latencies, which we will illustrate with actual examples later in this article.

Athlon 64 3200+ processors selling in stores right now are shipped in a very original package:

Inside this plastic box you can find a CPU, of course, a cooler, and cooler retention mechanism, an installation guide and a sticker with the CPU logo.

I would definitely like to say a few words about the cooler provided by AMD for its Athlon 64 processors.

This cooler recommended by AMD for use with AMD Athlon 64 CPUs consists of a heatsink and a fan. The heatsink footing is made of a copper alloy (high percentage of copper) and has thin aluminum ribs glued to it. The fan features a ball bearing motor and an integrated thermal diode. The fan rotation speed changes depending on the temperature and at about 32^oC reaches 3,050rpm, and at 42^oC rises up to 6,000rpm. The cooler is supplied together with a Shin Etsu G751 thermal pad. Here I have to draw your attention to the fact that the construction of this cooler ensures very low level of noise produced by the fan, especially compared to the pretty noisy coolers coming with the boxed Intel CPUs. Despite that, the cooling efficiency of this solution is very high.

The AMD Athlon 64 CPU looks very much like an Opteron or Athlon 64 FX:

The difference can be only noticed in the marking of this processor and in the smaller number of pins on the reverse side of the CPU, because Athlon 64 processors fit into Socket754 mainboards and are incompatible with Socket940 mainboards designed for Athlon 64 FX and Opteron processors.

Cool’n’Quiet Technology

Besides the above listed peculiarities, Athlon 64 boast one more remarkable feature. These CPUs support Cool’n’Quiet technology, which has in fact been borrowed from the mobile CPU versions. Actually, Cool’n’Quiet is somewhat similar to PowerNow! power saving technology, which has long been used in mobile CPUs from AMD. Now this technology has finally come into desktops as well. Cool’n’Quiet support is another big advantage of AMD Athlon 64 over Athlon 64 FX and Opteron CPUs, which do not (yet?) have anything like that.

AMD Company has been paying due attention to lowering the heat dissipation of its desktop CPUs for a long time already. I should say that here they have long surpassed Intel: the fastest AMD processors dissipate considerably less heat under maximum workload than the top Pentium 4 models. Also, these CPUs use special technologies reducing the heat dissipation significantly under low workloads as well. Besides, AMD Athlon 64 processor know to switch to the Halt/Stop Grant mode when the HALT command is performed. As a result, the processor temperature gets dramatically lower if it is not loaded 100%. However, now AMD has made another big move forward. The new Athlon 64 processors feature an even more intellectual algorithm for lowering the heat dissipation.

Beside the already mentioned Halt/Stop Grant, AMD Athlon 64 knows to reduce the working frequency and the Vcore in order to lower the heat dissipation even more. When this technology is enabled the processor driver manages the CPU clock frequency, by either reducing it or increasing according to the recent information about the CPU workload at that particular moment of time. Of course, if the CPU copes with the work successfully, and its workload is below 100%, its clock frequency can be reduced without affecting the overall system performance. For example, when the system is idle, when you work in office applications, watch video or perform disk defragmentation and the like, the processor doesn’t use all its power. This is exactly the case when the processor driver transfers Athlon 64 to a lower working frequency. As soon as full processor power needs to be involved, for instance in games, for complex calculations or data encoding, the processor clock frequency is returned back to the nominal. This is how Cool’n’Quiet technology actually works.

Now let’s see how the whole thing looks in real life. In common conditions with the minimal CPU workload, the processor driver drops the Athlon 64 3200+ frequency from the nominal 2GHz down to 800MHz.

The processor Vcore in this case drops to 1.3V. As you see, the clock frequency is reduced by lowering the CPU clock multiplier to 4x. This also explains why Athlon 64 are supplied with an unlocked clock frequency multiplier. The CPU keeps working in this mode until its workload exceeds 70%-80%. In particular, we managed to perform disk defragmentation together with MP3 track and MPEG-4 video playback, which the CPU still kept working at 800MHz core clock.

When the workload of the Athlon 64 processor working at 800MHz exceeds the allowed maximum, the CPU is transferred by the processor driver in the following mode:

In this mode the frequency of Athlon 64 3200+ makes 1.8GHz, and the Vcore – 1.4V. This is again achieved by reducing the CPU clock multiplier to 9x this time. And only if the processor workload again appears too high, the driver pushes the CPU back to the nominal settings: 2GHz core clock and 1.5V Vcore.

I believe there is no need ton explain that in workmodes with lower Vcore and working frequency, the Athlon 64 3200+ heat dissipation gets significantly lower. For a better comparison, take a look at the table below showing the heat dissipation of our today’s hero in major work modes:

This way, the use of Cool’n’Quiet technology allows to significantly reduce the CPU temperature not only when the system is idling, but also when the system is working on the tasks, which do not load the CPU that much. It is also very important that the performance in CPU-hungry applications doesn’t get affected by the Cool’n’Quiet technology at all. As a result, if the cooling solution in your system features a fan with varying rotation speed, the Cool’n’Quiet technology may significantly reduce the noise level of your system. And beside that, we save a lot on the CPU power consumption.

To be able to use Cool’n’Quiet technology, you will need the following:

1. AMD Athlon 64 3200+ processor supporting Cool’n’Quiet. These CPUs marking ends up with “AP”.



2. Mainboard supporting Cool’n’Quiet technology.



3. Operation system supporting ACPI 2.0 specification.

If you have all that at hand, you should update the processor driver (you can download the driver from AMD site, just click here).

Then you should enable any of the power consumption management algorithms in your OS other than Home/Office Desk and Always On:

Of course, like any other new technology Cool’n’Quiet is not absolutely bugless yet. For instance, AMD warns that with enabled Cool’n’Quiet your CPU may sometimes fail to return to the nominal settings if the processor workload dashes up too rapidly. However, they claim that the problem will be completely solved in the new upcoming driver versions.

Testbed and Methods

The major goal of our test session will be to find out how fast the new AMD Athlon 64 3200+ can actually be. For this purpose we compared the performance of our processor with that of Intel Pentium 4 3.2GHz, Intel Pentium 4 Extreme Edition 3.2GHz, AMD Athlon XP 3200+ and AMD Athlon 64 FX-51. Besides, we also added the results for AMD Opteron 146 processor working at the same 2GHz frequency as Athlon 64 3200+, but featuring the dual-channel memory controller similar to that used in Athlon 64 FX. It will allow us to evaluate how big is the contribution of the dual-channel memory controller to the overall performance of Athlon 64 FX.

So, our testbeds were built with the following equipment:

• CPUs: AMD Athlon 64 FX-51 (2.2GHz), AMD Athlon 64 3200+ (2.0GHz), AMD Athlon XP 3200+ (2.2GHz, 400MHz FSB), Intel Pentium 4 3.2GHz (800MHz FSB), Pentium 4 Extreme Edition 3.2GHz (800MHz FSB);
• Mainboards: ASUS P4C800 Deluxe (i875P), ASUS SK8N (NVIDIA nForce3 Pro 150), ASUS K8V Deluxe (VIA K8T800), ABIT NF7-S 2.0 (NVIDIA nForce2 Ultra 400);
• Memory: 1024MB DDR400 SDRAM (2 x 512MB, 2-2-2-5), 1024MB Registered DDR400 SDRAM (2 x 512MB, 2.5-3-3-5);
• NVIDIA GeForce FX 5900 Ultra graphics card (Detonator 45.23);
• Western Digital WD400JB HDD.

Notes:

• In all cases the memory worked in the fastest mode with the minimal timings possible for each particular configuration. This way, the systems requiring Registered memory worked with DDR400 with 2-2-2-5 timings, and the timings for Athlon 64 FX based system were set to 2.5-3-3-5, because faster memory modules are not available in the market yet.
• The tests were run in Windows XP SP1 with DirectX 9.0b installed, unless specified otherwise.

Memory Subsystem Performance

First of all, we decided to study the performance of the memory controller integrated into Athlon 64 3200+. We have already mentioned above that low latency is one of the theoretical advantages of this solution, while the lower bandwidth than that of Athlon 64 FX is a theoretical drawback. Therefore, I decided to test the Athlon 64 3200+ memory controller and compare it to the dual-channel controller of the Athlon 64 FX/Opteron CPUs. In order to obtain adequate results, we compared the memory performance in an Athlon 64 3200+ based system with that in an Opteron 146 based system. Both these processors work at the same clock frequency and in fact differ only by the memory subsystem implementation. In an Opteron 146 based system we used a single-channel and dual-channel registered DDR400 with the timings set as 2.5-3-3-5, and in Athlon 64 3200+ based system – single-channel non-registered DDR400 with 2-2-2-5 and 2.5-3-4-5 timings. This way, we will be able to estimate not only the practical correlation between the performance of the memory subsystems in Athlon 64 3200+ and Athlon 64 FX/Opteron, but also the performance differences between the registered and non-registered memory.

So, let’s pass over to the practical experiments. We used the already familiar Cache Burst 32 utility, which is a worthy successor of the good old Cachemem:

Well, it was quite predictable that Athlon 64 FX/Opteron based system with a dual-channel memory controller will win the memory bandwidth test. There is another interesting thing here. If we compare the results for single-channel configurations, then we will notice that the memory controller of Athlon 64 ensures higher memory bandwidth, even if the memory configuration is absolutely identical to that of the registered memory in Athlon 64 FX/Opteron. This is exactly the price you pay for the use of Registered memory in Socket940 systems. Moreover, non-registered memory modules allows using much more aggressive timings and in this case the advantage of Athlon 64 (compared with the Athlon 64 FX/Opteron with a single memory channel) appears even greater.

The latency tests once again prove how slow the registered memory is. The memory controller of Athlon 64 working with the non-registered memory modules, outperforms the memory controller of Athlon 64 FX/Opteron even if the timing settings are the same in both cases. If you set the minimal timings possible for the Athlon 64 based system, then the its memory latency will appear 25% lower than that of the Athlon 64 FX/Opteron based system.

All this indicates that even though Athlon 64 FX boasts a dual-channel memory controller, the competition with the new Athlon 64  working at the same clock frequency is not settled once and for all. Due to the fact that Athlon 64 supports unregistered memory, the systems based on it can be even faster in some applications than Socket940 systems with dual-channel memory subsystem using registered modules. This is probably why the currently in production Athlon 64 FX-51 model works at a higher clock frequency than the top Athlon 64 processor. Otherwise, it would be simply impossible to claim that Athlon 64 FX is definitely faster than the regular Athlon 64 .

 In this respect the upcoming Athlon 64 FX processors designed for Socket939 seem to me very interesting. These processors are due next year and they are expected to be free from the major drawback of the today’s Athlon 64 FX: they will support unregistered memory modules in dual-channel DDR SDRAM configurations.

Performance in Office and Content Creation Applications

The testing methodology remained unchanged. As usual, we used Winstone benchmarks here:

Business Winstone 2002 emulates the everyday work of an office computer. This benchmark measures the performance in multi-task mode, when you simultaneously run a word processor, work with tabled data, create presentations, check your e-mail, run antivirus software, etc. In this type of tasks AMD processors have always been very successful, and this time was no exception, of course. Although Athlon 64 3200+ fell a little bit behind Athlon 64 FX-51 and even Athlon XP 3200+, because these two processors work at higher clock frequency, and faster memory subsystem doesn’t actually matter that much in this sort of applications.

The second test, Multimedia Content Creation Winstone 2003 emulates workstation mode for audio, video and image processing. Here the leadership belongs to Pentium 4 CPUs, because their architecture is better suitable for streaming data processing. As for Athlon 64 3200+, it is considerably ahead of the Athlon XP 3200+ due to faster memory subsystem, SSE2 instructions support and larger L2 cache. However, it falls behind AMD CPUs with a dual-channel memory controller, namely Athlon 64 FX-51 and Opteron 146.

Performance in data Encoding Applications

Data compression with WinRAR archiving utility requires high-speed memory subsystem. Therefore, Athlon 64 3200+ featuring a single-channel memory controller manages to outpace only Athlon XP 3200+ here, which performance is severely limited by a pretty narrow processor bus.

As we have already seen several times, WAV-files compression into mp3 format loads the computational units of the CPU in the first place. Besides the codec we are using supports Hyper-Threading, which allows Pentium 4 processors to perform very well in this test. As for Athlon 64 3200+, it doesn’t run very fast because of the pretty low clock frequency of only 2GHz. Moreover, I would also like to draw your attention to the fact that we see Athlon 64 3200+ outperforming Opteron 146 here, even though the latter features dual-channel memory subsystem. It is exactly what I have already warned you about: due to lower memory latency Athlon 64 manages to outperform Opteron working at the same clock frequency.

During MPEG-2 encoding in Canopus ProCoder 1.5 AMD processors show their best. Athlon 64 3200+ is no exception again. It manages to be one of the fastest here. It even outpaces Opteron 146!

The tests of MPEG-4 encoding speed name Pentium 4 CPUs as winners. As for Athlon 64 3200+, it runs almost as fast as Opteron 146, being just a little behind Athlon 64 FX-51 and a little ahead of Athlon XP 3200+.

A similar picture can be observed in case of Windows Media Encoder version 9.

Performance in Gaming Applications

We paid special attention to testing AMD Athlon 64 3200+ in gaming applications. The biggest part of all AMD CPU fans is the computer enthusiasts, and gamers definitely belong here, too.

In a pretty old but still very popular test set, such as 3DMark 2001 SE, Athlon 64 3200+ shows very good results outperforming even Pentium 4 3.2GHz. However, the situation changes drastically in 3DMasrk03, where Pentium 4 processors beat the Athlon family. Nevertheless, according to the CPU score index, Athlon 64 3200+ outperforms Pentium 4 3.2GHz, probably due to larger L2 cache in the first place.

The picture we see in Aquamark3 is very similar to what we have just seen in 3DMark03. Intel Pentium 4 processors are pretty strong here. Also note that just like in 3DMark03, Athlon 64 3200+ performs nearly as fast in Aquamark3 as Opteron 146. since both these processors work at the same clock frequencies, it means that low latency of Athlon 64 memory subsystem is capable of making up for the lower bandwidth.

If Athlon 64 FX-51 has finally managed to defeat Pentium 4 3.2GHz in Quake3, then Athlon 64 3200+ will not be able to repeat this success. Although it is only 2.5% behind Pentium 4 3.2GHz. I would also like to point out that due to larger L2 cache and integrated memory controller Athlon 64 3200+ outperforms Athlon XP 3200+ by about 20%, even though Athlon XP is 200MHz faster.

In Unreal Tournament 2003 the situation is just the opposite: AMD CPUs are ahead here. Note that Athlon 64 3200+ is again faster than Opteron 146.

The same is true for Serious Sam: The Second Encounter and Splinter Cell.

The new DirectX9 benchmark based on X2 – The Threat game gives the victory to Pentium 4 Extreme Edition. However, the remaining participants seem to be doing just the same way as in the previous benchmarks. Again lower memory subsystem latency appears more important that high bandwidth that is why Athlon 64 3200+ outperforms Opteron 146, Pentium 4 3.2GHz and Athlon XP 3200+.

Here Athlon 64 3200+ still keeps pleasing us with the results. However, in Halo this CPU is still a little behind Opteron 146 (for the first time throughout the last 6 tests) and Pentium 4 Extreme Edition with 2MB L3 cache.

The results in the new Tomb Raider series appears a surprise for us. Here Athlon 64 3200+ is an indisputable winner, leaving behind even Athlon 64 FX-51. low memory subsystem latency can sometimes work wonders, as we have just seen.

Performance in Professional OpenGL Applications and Photoshop

Here we are going to check how fast our platforms are in CINEMA4D viewports and a popular Photoshop graphics editor.

The performance of Athlon 64 3200+ in CINEMA4D can hardly be called impressive. It must be the pretty low working frequency that tells here.

Now let’s see how well it copes with Photoshop tasks (as usual we will use PSBench7 test processing a 50MB image).

In this popular graphics package the memory subsystem performance matters a lot. However, Intel CPUs with Hyper-Threading technology manage to stay ahead. As for Athlon 64 3200+, it is definitely better than the predecessor, Athlon XP 3200+. Moreover, Athlon 64 3200+ again outpaces Opteron 146 with its dual-channel memory subsystem.

Now let’s go a little bit more into detail regarding the performance of our testing participants in Photoshop.

Performance in 3D Rendering Applications

AMD processors could do the rendering in 3ds max 5 much faster, if they supported Hyper-Threading technology, which has already helped Intel Pentium 4 processor an immense lot here. As for Athlon 64 3200+, it doesn’t tend to prove anything here, but still defeats Opteron 146 with dual-channel memory subsystem and Athlon XP 3200+ with higher clock frequency.

Faster memory subsystem and SSE2 support have had a very positive effect on the CPUs with AMD64 architecture in Lightwave test. However, Athlon 64 3200+ still doesn’t manage to outperform Pentium 4 3.2GHz.

Note that in all 3D rendering applications low memory subsystem latency appears more important for the performance than high memory bus bandwidth. In all the three tests: 3ds max 5, Lightwave and CINEMA4D, Athlon 64 3200+ outperforms Opteron 146.

Performance in Scientific Applications

Here we are going to use a beta-version of ScienceMark 2.0 test package, which measures how fast the tested platforms solve math1ematical modeling tasks.

One of the big advantages of Athlon architecture is a powerful FPU, which is exactly what works the most in applications like that. Therefore, no wonder that even Athlon XP 3200+ is so successful in ScienceMark 2.0. The performance of Athlon 64 FX-51 is even higher. However, the working frequency of Athlon 64 3200+ is lower than that of the Athlon 64 FX-51 or Athlon XP 3200+. That is why our hero didn’t make it this time.

Overclocking

Overclocking Athlon 64 3200+ is definitely a worthy experience. Keeping in mind that this CPU is made of the same dies as Athlon 64 FX-51, which works at a higher clock frequency, we dare suppose that it is quite possible to overclock Athlon 64 3200+ over the nominal frequency. The major question is how high will we be able to raise its frequency then? In fact, we do not expect this processor to work wonders. If it had had a big clock frequency potential, AMD would definitely have taken advantage of it and released faster models as well, especially now that the competition with Intel has become really hard for AMD. However, this hasn’t yet happened. It means that AMD has almost exhausted the potential of the 0.13micron SOI production technology used for Athlon 64 3200+ manufacturing, so Athlon 64 is very unlikely to acquire much higher core clocks unless the 90nm core comes out.

So now let’s pass over to a practical aspect of this test. For overclocking we used ASUS K8V mainboard. It allows increasing processor Vcore and manage the bus frequency accordingly. Unfortunately, the mainboard does not allow changing the CPU clock frequency multiplier, although the multiplier of Athlon 64 3200+ is unlocked. That is why we overclocked the processor by raising the bus frequency. The Vcore was increased by 0.15V above the nominal value.

Our suppositions proved true: Athlon 64 3200+ doesn’t boast an impressive overclocking potential. The maximum core clock, which we managed to achieve without losing stability (with the default cooling solution) made only 2.34GHz:

This way, we managed to speed up the CPU by about 17%, which is actually not that bad at all. The performance of the Athlon 64 3200+ CPU overclocked to 2.34GHz will definitely be better than that of Athlon 64 FX-51, which may inspire you to take the risks.

I also have to say that although we are talking about the FSB frequency, this is not quite correct when we refer it to Athlon 64. In fact, Athlon 64 has no front side bus at all. The memory controller is integrated into the CPU and works at the full core frequency. The chipset is connected with the CPU via the Hyper-Transport bus working at 800MHz. As for the parameter initially set to 200MHz, which we increased during CPU overclocking, this frequency is generated by the mainboard and serves as a basis for forming the CPU clock rate.

Conclusion

In this article we have just discussed a new AMD CPU targeted for the mass market aka Athlon 64 3200+. Unlike Athlon 64 FX-51, this processor features a single-channel memory controller supporting unregistered memory modules and doesn’t require expensive 6-layer mainboard PCBs. All this may make Athlon 64 3200+ a more popular solution that Athlon 64 FX-51.

I have to say that the performance of Athlon 64 3200+ is slightly lower than that of Athlon 64 FX-51. Mostly it is not because of the memory controller with lower bandwidth, but because of the lower core clock frequency. Since the memory subsystem latency of Athlon 64 3200+ is considerably lower than that of Athlon 64 FX/Opteron processors, the performance of the two is about the same on average. That is exactly why AMD hasn’t yet released Athlon 64 with 2.2GHz core clock. For the same reason AMD will not roll out faster Athlon 64 models that soon, because Athlon 64 FX family positioned as High-End stuff, will have to be faster than Athlon 64.

Also I have to stress that Athlon 64 3200+ is much faster than Athlon XP 3200+ in most benchmarks, even though both these processors have similar rating. In fact, Athlon 64 3200+ can easily compete with Intel Pentium 4 3.2GHz: the tests prove it hundred percent. Of course, in some applications NetBurst architecture appears more suitable, but there definitely is a number of tasks where Athlon 64 3200+ is an indisputable performance leader.

All in all, Athlon 64 3200+ can be considered a good choice for those users who prefer AMD products. The only thing I have to mention though in conclusion is a questionable situation with the Socket754 platforms. Since the CPUs with a dual-channel memory controller cannot be used in this socket, I expect all Socket754 CPUs to be soon moved into the Budget segment. Therefore, you should consider Socket754 as a solution allowing further upgrade only if you realize that this system will not be able to set any performance records later on.