Home AMD’s Athlon 64 3500+ processor
Reviews

AMD’s Athlon 64 3500+ processor

Geoff Gasior
Disclosure
Disclosure
In our content, we occasionally include affiliate links. Should you click on these links, we may earn a commission, though this incurs no additional cost to you. Your use of this website signifies your acceptance of our terms and conditions as well as our privacy policy.

LAST WEEK, AMD slashed the price of its Athlon 64 3500+ processor by 30%, making the cheapest 939-pin chip more affordable for those saving pennies for a Doom 3-inspired upgrade. There’s more to this price cut than the hefty discount, though. Because the 3500+ is AMD’s least expensive 939-pin chip, the cut effectively lowers the price of entry for the entire Socket 939 platform.

Is the Athlon 64 3500+ the chip to have for budget-conscious enthusiasts? Is its 3500+ performance rating well-deserved? Read on to find out.

A closer look at the 3500+
Before we dive into the benchmarks, let’s have a look at the 3500+ up close.


The Athlon 64 3500+


939 pins

AMD’s current Athlon 64 lineup is made up of chips with different clock speeds, memory interfaces, and L2 cache sizes, making assigning arbitrary performance ratings a dubious task. With a little help from our handy CPU Decoder Ring, here’s a cheat sheet that sums up where the 3500+’s clock speed, memory interface, and L2 cache size fit in with the rest of the Athlon 64 family.

Socket Memory interface L2 cache 1.8GHz 2.0GHz 2.2GHz 2.4GHz
754 single-channel 512KB 2800+ 3000+ 3200+ 3400+
1MB   3200+ 3400+ 3700+
939 dual-channel 512KB     3500+ 3800+
1MB       FX-53

As you can see, the Athlon 64 3500+ actually shares the same 2.2GHz clock speed as the 3400+. So why the higher performance rating for the 3500+? Because the 3400+’s on-die memory controller can only address a single channel of DDR memory. The 3500+’s memory controller is a dual-channel design that offers twice the peak theoretical memory bandwidth of the 3400+, provided you use two DIMMs. So with clock speeds being equal, the 3500+’s extra memory channel earns it a higher performance rating. Simple, right?

Not so fast. The Athlon 64 3500+ and 3400+ also differ in L2 cache size, and when it comes to cache, the 3400+ is better-equipped. The 3400+ is based on an older Clawhammer Athlon 64 core that AMD endowed with 1MB of L2 cache, while the 3500+ is based on a more recent Newcastle core with only 512KB of L2. The 3500+ may have twice the memory channels of the 3400+, but it only has half as much L2 cache.

Based on their Athlon 64 performance ratings, AMD apparently feels that the 3500+’s dual-channel memory controller can more than compensate for the chip’s smaller L2 cache. They may be right, but it’s hard to believe that the 3400+’s larger cache won’t give it the upper hand in at least a couple of tests. We have plenty of benchmark results to sift through to find out.

 

Our testing methods
As ever, we did our best to deliver clean benchmark numbers. Tests were run at least twice, and the results were averaged.

Our test systems were configured like so:

Processor Athlon XP-M ‘Barton’ 2500+ 2.4GHz Athlon XP ‘Barton’ 3200+ 2.2GHz Athlon XP ‘Barton’ 3000+ 2.167GHz Athlon 64 3000+ 2.0GHz
Athlon 64 3200+ 2.0GHz
Athlon 64 3400+ 2.2GHz
Opteron 146 2.0GHz
Opteron 148 2.2GHz
Opteron 150 2.4GHz
Athlon 64 3800+ 2.4GHz
Athlon 64 FX-53 2.4GHz
Athlon 64 3500+ 2.2GHz Pentium 4 2.8’C’GHz
Pentium 4 3.2GHz
Pentium 4 3.4GHz
Pentium 4 3.2GHz Extreme Edition
Pentium 4 3.4GHz Extreme Edition
Pentium 4 2.8’E’GHz
Pentium 4 3.0’E’GHz
Pentium 4 3.2’E’GHz
Pentium 4 3.4’E’GHz
Front-side bus 400MHz (200MHz DDR) 400MHz (200MHz DDR) 333MHz (166MHz DDR) HT 16-bit/800MHz downstream
HT 16-bit/800MHz upstream
HT 16-bit/800MHz downstream
HT 16-bit/800MHz upstream
HT 16-bit/1GHz downstream
HT 16-bit/1GHz upstream
HT 16-bit/1GHz downstream
HT 16-bit/1GHz upstream
800MHz (200MHz quad-pumped)
Motherboard Abit AN7 Asus A7N8X Deluxe v2.0 Asus A7N8X Deluxe v2.0 MSI K8T Neo MSI 9130 MSI MS-6702E Abit AV8 Abit IC7-G
BIOS revision 1.4 C1007 C1007 1.1 1.31 3.0B10 v12 IC7_21.B00
North bridge nForce2 SPP nForce2 SPP nForce2 SPP K8T800 K8T800 K8T800 Pro K8T800 Pro 82875P MCH
South bridge nForce2 MCP-T nForce2 MCP-T nForce2 MCP-T VT8237 VT8237 VT8237 VT8237 82801ER ICH5R
Chipset drivers ForceWare 3.13 ForceWare 3.13 ForceWare 3.13 4-in-1 v.4.51
ATA 5.1.2600.220
4-in-1 v.4.51
ATA 5.1.2600.220
4-in-1 v.4.51
ATA 5.1.2600.220
4-in-1 v.4.51
ATA 5.1.2600.220
INF Update 5.1.1002
Memory size 1GB (2 DIMMs) 1GB (2 DIMMs) 1GB (2 DIMMs) 1GB (2 DIMMs) 1GB (2 DIMMs) 1GB (2 DIMMs) 1GB (2 DIMMs) 1GB (2 DIMMs)
Memory type Corsair TwinX XMS4000 DDR SDRAM at 400MHz Corsair TwinX XMS4000 DDR SDRAM at 400MHz Corsair TwinX XMS4000 DDR SDRAM at 333MHz Corsair TwinX XMS4000 DDR SDRAM at 400MHz Corsair CMX512RE-3200LL PC3200 registered DDR SDRAM at 400MHz Corsair TwinX XMS3200LL DDR SDRAM at 400MHz Corsair XMS3500C2PT DDR SDRAM at 400MHz Corsair TwinX XMS4000 DDR SDRAM at 400MHz
CAS latency 2 2 2 2 2 2 2 2
Cycle time 6 6 5 5 6 6 6 6
RAS to CAS delay 3 3 3 3 3 3 3 4
RAS precharge 3 2 2 3 2 3 3 4
Hard drive Seagate Barracuda V 120GB ATA/100 Seagate Barracuda V 120GB ATA/100 Seagate Barracuda V 120GB ATA/100 Seagate Barracuda V 120GB SATA 150 Seagate Barracuda V 120GB SATA 150 Seagate Barracuda V 120GB SATA 150 Seagate Barracuda V 120GB SATA 150 Seagate Barracuda V 120GB SATA 150
Audio Creative SoundBlaster Live!
Graphics Radeon 9800 Pro 256MB with CATALYST 4.1 drivers
OS Microsoft Windows XP Professional
OS updates Service Pack 1, DirectX 9.0b

I should point out that our 3500+ test platform uses a different motherboard and memory than our other 939-pin chips. Both platforms use the same VIA K8T800 Pro chipset and identical memory timings and BIOS settings, so results should be comparable.

Since I don’t have a Radeon 9800 Pro 256MB in the Benchmarking Sweatshop, I used an underclocked Radeon 9800 XT 256MB running Pro speeds. OVERDRIVE was disabled on the underclocked 9800 XT, so the card’s performance should match that of the Pro.

All tests on the Intel systems were run with Hyper-Threading enabled.

Thanks to Corsair for providing us with memory for our testing. If you’re looking to tweak out your system to the max and maybe overclock it a little, Corsair’s RAM is definitely worth considering.

The test systems’ Windows desktops were set at 1152×864 in 32-bit color at an 85Hz screen refresh rate. Vertical refresh sync (vsync) was disabled for all tests.

We used the following versions of our test applications:

The tests and methods we employ are generally publicly available and reproducible. If you have questions about our methods, hit our forums to talk with us about them.

 
Memory performance

Our memory bandwidth results make a good case for the 3500+’s higher performance rating. The 3500+’s dual-channel memory controller offers nearly double the bandwidth of the 3400+ in Sandra, but its advantage is less pronounced in Cachemem’s bandwidth tests.

The 3500+’s Linpack performance nicely highlights how the chip differs from the 3400+. Notice how the 3500+’s performance begins to trail off as matrix sizes approach 512KB. The 3400+ doesn’t take a dive until matrix sizes flirt with 1MB, but it’s a sharp dive. The 3500+’s dual-channel memory controller more gracefully shoulders the burden of larger matrices.

Cachemem’s latency tests make a compelling case for on-die memory controllers. The Athlon 64 family as a whole rules this test, and the 3500+ is right up there.

 
Memory performance (continued)
Not only are our 3D graphs indulgent, but they’re useful, too. I’ve arranged them manually in a very rough order from worst to best, for what it’s worth. I’ve also colored the data series according to how they correspond to different parts of the memory subsystem. Yellow is L1 cache, light orange is L2 cache, and orange is main memory. The red series, if present, represents L3 cache. Of course, caches sometimes overlap, so the colors are just an interesting visual guide.

Although it can’t quite match the latencies of AMD’s 2.4GHz Athlon 64 chips, the 3500+ isn’t far behind. Note the 3400+’s lower latencies at the 1024KB block size when compared with the 3500+, that’s the 3400+’s larger L2 in action.

 

Unreal Tournament 2003

The 3400+ performs better than the 3500+ in Unreal Tournament 2003, proving that a dual-channel memory controller can’t always overcome an L2 cache deficiency.

Quake III Arena

Quake III Arena also favors the 3400+ over the 3500+…

Wolfenstein: Enemy Territory

But Wolfenstein: Enemy Territory does not.

 

Comanche 4

The 3500+ just edges out the 3400+ in Comanche 4.

Serious Sam SE

However, the 3400+ flexes its L2 cache in Serious Sam SE and comes out ahead by over five frames per second.

3DMark03

3DMark03’s overall score isn’t supposed to be CPU-bound, but our 3500+ system leads the way. I suspect that this bit of funny business has something to do with our underclocked Radeon 9800 XT, which according to Rage3D’s Radeon overclocking tools, was running with the same core and memory speeds as a Radeon 9800 Pro 256MB.

3DMark03’s CPU tests are split; test 1 favors the 3400+’s larger cache, while test 2 prefers the 3500+’s memory bandwidth.

 

Sphinx speech recognition
Ricky Houghton first brought us the Sphinx benchmark through his association with speech recognition efforts at Carnegie Mellon University. Sphinx is a high-quality speech recognition routine that needs the latest computer hardware to run at speeds close to real-time processing. We use two different versions, built with two different compilers, in an attempt to ensure we’re getting the best possible performance.

There are two goals with Sphinx. The first is to run it faster than real time, so real-time speech recognition is possible. The second, more ambitious goal is to run it at about 0.8 times real time, where additional CPU overhead is available for other sorts of processing, enabling Sphinx-driven real-time applications.

Sphinx loves memory bandwidth, so the 3500+ does well. The chip can’t quite catch Intel’s fastest Prescotts, but it’s close.

LAME MP3 encoding
We used LAME to encode a 101MB 16-bit, 44KHz audio file into a very high-quality MP3. The exact command-line options we used were:

lame –alt-preset extreme file.wav file.mp3

DivX video encoding
This new version of XMPEG includes a benchmark feature, so we’re reporting scores in frames per second now.

If the 3500+ has a weakness, it’s media encoding. The chip can only manage a middle-of-the-pack performance with LAME MP3 encoding, and there’s no touching Intel’s Pentium 4s when it comes to DivX video encoding.

 

3ds max rendering
We begin our 3D rendering tests with Discreet’s 3ds max, one of the best known 3D animation tools around. 3ds max is both multithreaded and optimized for SSE2. We rendered a couple of different scenes at 1024×751 resolution, including the Island scene shown below. Our testing techniques were very similar to those described in this article by Greg Hess. In all cases, the “Enable SSE” box was checked in the application’s render dialog.

The 3500+ sits in the middle of the field in 3D Studio Max. Scores are close between the 3400+ 3500+, but the 3400+ sneaks in a half-second victory with the Islands scene.

 

Lightwave rendering
NewTek’s Lightwave is another popular 3D animation package that includes support for multiple processors and is highly optimized for SSE2. Lightwave can render very complex scenes with realism, as you can see from the sample scene, “A5 Concept,” below.

We’ve tested the Hyper-Threaded processors with one, two, and four rendering threads. For non-Hyper-Threaded processors, we just tested with one and two threads. For the line graphs, we’ve tried to pick results from the optimal number of threads to represent each processor.

The 3500+ is marginally faster than the 3400+ in Lightwave rendering.

 

POV-Ray rendering
POV-Ray is the granddaddy of PC ray-tracing renderers, and it’s not multithreaded in the least, because it’s designed to be a cross-platform application. POV-Ray also relies more heavily on x87 FPU instructions to do its work, and it contains only minor SIMD optimizations.

POV-Ray loves the Athlon 64, and the 3500+ comes in just behind AMD’s 2.4GHz chips.

 

Cinebench 2003 rendering and shading
Cinebench is based on Maxon’s Cinema 4D modeling, rendering, and animation app. This revision of Cinebench measures performance in a number of ways, including 3D rendering, software shading, and OpenGL shading with and without hardware acceleration. Cinema 4D’s renderer is multithreaded, so it takes advantage of Hyper-Threading, as you can see in the results.

The Pentium 4 gets a nice boost from Hyper-Threading in Cinebench’s rendering test, making it tough for the 3500+ to compete.

However, the 3500+ is all over the rest of the Cinebench suite, beating the 3400+ and even Intel’s Extreme Editions in each test.

 

SPECviewperf workstation graphics
SPECviewperf simulates the graphics loads generated by various professional design, modeling, and engineering applications.

Viewperf doesn’t seem to mind that the 3500+ has only 512KB of L2 cache. The 3500+ bests the 3400+ throughout the suite.

 

ScienceMark
ScienceMark is optimized for SSE, SSE2, 3DNow! and is multithreaded, as well. In the interest of full disclosure, I should mention that Tim Wilkens, one of the originators of ScienceMark, now works at AMD. However, Tim has sought to keep ScienceMark independent by diversifying the development team and by publishing much of the source code for the benchmarks at the ScienceMark website. We are sufficiently satisfied with his efforts, and impressed with the enhancements to the 2.0 beta revision of the application, to continue using ScienceMark in our testing.

The 3500+ performs extremely well in ScienceMark, but the 3400+ manages to steal a little glory in the Molecular Dynamics test.

The Athlon 64 offers more balanced performance across different code paths than the Pentium 4, and the 3500+ is no exception. Here, the 3500+ is neck and neck with the 3400+.

 

picCOLOR image analysis
We thank Dr. Reinert Muller with the FIBUS Institute for pointing us toward his picCOLOR benchmark. This image analysis and processing tool is partially multithreaded, and it shows us the results of a number of simple image manipulation calculations.

The 3500+ can’t quite catch Intel’s 3.2 and 3.4GHz Prescotts in picCOLOR’s image analysis tests, but it’s not far off the pace. Let’s break things down to individual tests to get a better idea of where the 3500+ excels and where the chip stumbles.

The 3400+ and 3500+ battle back and forth in picCOLOR’s individual tests, but when the 3500+ comes out on top, it tends to do so by a bigger margin.

 
Overclocking
I was able to get the 3500+ stable at 2.37GHz on a 215MHz system bus in the steamy confines of the Benchmarking Sweatshop. Unfortunately, I had to increase the motherboard’s memory command rate from 1T to 2T to get the system stable at that speed, but that’s not the processor’s fault. With a 1T command rate, the highest I was able to overclock the 3500+ was 2.31GHz. That’s a repsectable boost, especially considering that I only had to raise the processor’s core voltage to 1.6V.

Because raising the system’s command rate to 2T results in lower performance, I’ve left the 2.37GHz scores out of the graphs below. Frame rates at 2.37GHZ and 2T were lower than 2.31GHz and 1T, anyway.

A 130MHz clock speed boost is good for a few extra frames per second in Unreal Tournament 2003, but it’s not enough to catch chips clocked at 2.4GHz. As is always the case with overclocking, your mileage may vary. Overclocking success is never guaranteed and can often depend as much on individual system components as pure luck.

 
Conclusions
Despite its 512KB L2 cache, the Athlon 64 3500+’s performance rating is well-deserved. The chip’s dual-channel memory controller more than makes up for the smaller cache, but it’s not a clean sweep. In seven of our tests, mostly games, 1MB of L2 cache helps the Athlon 64 3400+ overcome its single-channel memory controller and come out ahead of the 3500+. If you’re primarily a gamer, that’s a pretty compelling selling point for the 3400+, which can be had for $60 cheaper than the 3500+.

Of course, the 3400+ doesn’t come without baggage. AMD has indicated that Socket 939 is the future of the Athlon 64 platform, and there’s no telling how long it will keep Socket 754 around or how often it will release faster processors for the platform. Socket 754’s processor upgrade potential is questionable at best, especially if your starting point is a 3400+.

Comparing the 3500+ to an equivalent Pentium 4 from Intel is a little more difficult. The 3500+’s $345 street price is sandwiched between the Pentium 4 3.2GHz at $251 and the 3.4GHz at $397. That doesn’t take into account Intel’s latest LGA775 chips, but those require virtually a system-wide upgrade. However, LGA775 or not, it’s tough to recommend a Pentium 4 of any speed unless you spend the majority of your time encoding video or using multi-threaded apps that take advantage of Hyper-Threading’s second virtual processor. Prescott’s toasty tendencies don’t help matters either, making the Athlon 64 a safer bet that could become even more compelling when Windows XP for 64-bit Extended Systems is—barring further delays—finally released next year.

At the end of the day, the 3500+’s solid performance and reasonable price tag make it an excellent choice for gamers and enthusiasts looking for a processor platform with guaranteed upgrade potential. Single-channel Socket 754 chips like the 3400+ may be cheaper today, but the 3500+ is definitely a more sound investment for tomorrow. 

Latest News

IPTV
Streaming News & Events

Operator of Illegal IPTV Streams Sentenced to Five Years in Jail

White House Announces New Set of Rules for Federal Agencies Using AI
News

White House Announces New Set of Rules for Federal Agencies Using AI

The US government has announced that federal agencies using AI tools will be required to follow new safeguards by December 1. The announcement comes from the Office of Management and...

4 Canadian School Boards Sue Three Social Media Giants
News

Four Canadian School Boards Have Sued Social Media Giants for Sabotaging Young Minds

Four of the largest school boards in Canada have filed a lawsuit against social media giants for being addictive, disrupting student learning, and harming their mental health. The lawsuit seeks...

Slothana goes parabolic
Crypto News

Traders Transfer $2.2 Million in Solana to Emerging Meme Cryptocurrency Slothana

Reddit Shares Fall 16% In A Day After Promoters Sell
News

Reddit Shares Fall 16% in a Day after Promoters Sell One Million Shares

Gold Miner Nilam Resources Shares Surge 22x Amidst Bitcoin Buying Announcement
Crypto News

Gold Miner Nilam Resources Shares Surge 22x Amidst Bitcoin Buying Announcement

BlackRock CEO Goes Bullish on BTC as Spot Bitcoin ETF Crosses $17 Billion
Crypto News

BlackRock CEO Goes Bullish on BTC as Spot Bitcoin ETF Crosses $17 Billion