The Eternal Struggle: AMD vs. Intel
The competition raging between archrivals AMD and Intel has just increased a notch. It’s a bit of a stretch to consider this competition as a special form of class struggle from a business standpoint. Critics with philosophical leanings interpret a “class struggle” as a society split into two castes, one of which continually gives while the other continually takes. Although this might be the case in some areas – it doesn’t hold water in the processor business. After all, long gone are the days when AMD played the role of the outsider, and when innumerable publications all over the world endeavored to emphasize the positive sides of the assumed “weaker and vulnerable” AMD.
All AMD CPUs before the lab test.
Our last test focused exclusively on comparing the top models from AMD and Intel. A company’s top-of-the-line models are, after all, its flagship – this is as true for the Pentium 4/2000 as it is for the Athlon XP 1800+. But only a handful of users actually buy these expensive products – ultimately, a medium-range processor costs considerably less than its faster big brother, and provides almost the same performance. For this reason, we’ve lined up all the Athlon XP processors available against comparable Pentium 4 CPUs. AMD currently offers the Athlon XP in four performance levels: the XP 1500+ (1333 MHz), the XP 1600+ (1400 MHz), the XP 1700+ (1467 MHz) and the XP 1800+ (1533 MHz), and another model will soon be introduced. The Intel Pentium 4 comes in just as many flavors, running the gamut from 1400 MHz to 1500 MHz, 1600 MHz, 1700 MHz, 1800 MHz, 1900 MHz, all the way to the top-of-the-line 2000 MHz model. We tested a total of 12 processors in this comparison.
All Intel CPUs before the lab test.
THG Upgrades Benchmark Suite for Windows XP
We’ve introduced a slew of new benchmarks in this test. Firstly, we wanted to cover an even wider range of practical applications in our tests; secondly, Windows XP is becoming increasingly important. Windows XP comes with some SSE optmizations, which benefits both Pentium 4 and Athlon XP, and it’s slightly faster than Windows 2000. Benchmarks under the older Windows 98 SE have been removed because Microsoft is planning to freeze support for this operating system.
And that’s not all – we’ve also introduced a value indicator that determines which processor is the most reasonably priced for the performance it offers. This indicator alone sheds light on some interesting aspects and might help ambitious users decide which product to buy. We’ve often received mail from readers complaining, with just cause, that we only evaluate the benchmark performance of the CPUs. In this comparison, there is also a winner for best value.
Athlon XP Overclocking: Multiplier In The L3/L4 Bridge
These pictures show clearly, that the Athlon XP CPUs were produced on different production lines. The burning power of the lasers, used to cut the bridges, varies. The Athlon XP 1600+ you see above was cut with a higher powerd laser than the 1500+. The burnt spots (dark brown color) are therefore ‘bolder’. A burnt spot marks a severed bridge (open), while intact bridges are closed.
Overclocking the AMD Athlon XP is a hot topic for speed freaks, particularly for overclockers. Because AMD codes its multiplier using the L3/L4 bridges at the fab, any modifications have to be made on these contacts. Unlike its predecessor, the AMD with a Thunderbird core, on which the bridges could be connected simply by using a pencil, the “Palomino” is set up somewhat differently. If you look closely, you can see we tried the lead pencil-trick in a futile attempt on the Athlon XP 1800+ (second from above). If someone knows how to manipulate those L bridges properly, don’t hesitate to write to comment@tomshardware.com. We’re especially interested in simple, cheap and effective solutions. AMD severs the contact using a laser in accordance with the coding and then visibly strips away material (organic substrate), and so the bridge has to be closed again using a suitable process.
THG Value Indicator: Bang for Your Bucks Results
In our observations for this test, we’ve included a “value” criterion for the very first time. Bang for bucks. AMD’s XP branding campaign has changed the way that people now perceive and compare Intel and AMD CPUs. In fact, some might say it has muddied the waters. So, we thought it might make sense to come up with a comparison that reflective of actual bang for bucks, or in other words, how much performance for what cost.
These three diagrams spell out in black and white which CPU offers the most performance for the least amount of money. Our calculation is based on the current street prices for all CPUs (as of 30 October 2001) as well as the Cinebench 4D results. We chose this benchmark, which analyzes the rendering performance of a processor, because AMD and Intel CPUs can be better compared head-to-head with this benchmark.
For more information on this benchmark you can visit www.maxon.com. It gives us a starting point for our analysis, but we know that we will probably be debating the true value of a CPU for years to come as AMD and Intel slug it out with marketing and technology.
Intel: Pentium 4/1400 Best Value
This index is calculated with the price divided by the score of Cinebench. The score and fps benchmark results can be found further down in this article.
This diagram illustrates clearly that the low end Pentium 4 is the best value performer for Intel. At a price/performance factor of almost 9 it delivers the same performance as the P4/2000 at almost tripple the cost. If you are a budget P4 buyer then the 1400 is your port of call.
This diagram also shows that, for Pentium 4 processors, price increases exponentially to performance! In other words, every little extra boost in performance costs disproportionately more.
AMD: Athlon XP 1500+ Best Value
This index is calculated with the price in US$ divided by the score of Cinebench.
An entirely different pattern is shown in the AMD diagram. Here you’ll also see that we’ve included the older top-of-the-heap Athlon 1400 – with the Thunderbird core. Compared to its performance, this Athlon 1400 is much too pricey. The best value is offered by the Athlon XP 1500+, which runs at a real clock speed of 1333 MHz. Compared to the Athlon 1400 (1400 MHz) it’s a much better pick, particularly if you look at how it implements SSE commands. Even the Athlon XP 1600+, clocked at 1400 MHz, boasts better value compared to the Athlon 1400. The bottom line – stay away from the “old” Athlon 1400! The worst value is the Athlon XP 1800+, whose additional performance costs a pretty penny. Unlike the Intel processors, price increases linearly to performance, while the Pentium 4’s bars show exponential leaps.
Athlon XP 1500+ Is The King Of The Hill!
This index is calculated with the price divided by the score of Cinebench.
The comparison of all the processors lays it all out – in terms of value, the AMD Athlon XP 1500+ can’t be beat. Also interesting is a direct comparison of medium range Intel and AMD CPUs. Pentium 4/1700 and Athlon XP 1800+ are still good value, whereas the Pentium 4/1800 to 2000 are definitely too expensive for the performance you get. It offers much less value than any of the other processors.
However, you mustn’t forget that if you’re in the market for an AMD processor, be prepared to shell out at least 50 dollars for a powerful cooler. Cheap run-of-the-mill coolers costing as little as 15 dollars can only be used for speeds up to 1200 MHz.
Testing Procedure Peculiarities
| Intel Hardware Socket 478 |
|
| Processor | Intel Pentium 4/2000 MHz (400 MHz QDR FSB) Intel Pentium 4/1900 MHz (400 MHz QDR FSB) Intel Pentium 4/1800 MHz (400 MHz QDR FSB) Intel Pentium 4/1700 MHz (400 MHz QDR FSB) Intel Pentium 4/1600 MHz (400 MHz QDR FSB) Intel Pentium 4/1500 MHz (400 MHz QDR FSB) Intel Pentium 4/1400 MHz (400 MHz QDR FSB) |
| Motherboard | ASUS P4T-E (I850) Revision: 1.00 |
| Memory | 2 x 128 MB, RDRAM, 400 MHz, Viking |
| AMD Hardware Socket 462 |
|
| Processor | AMD Athlon XP 1800+ MHz (1533/266 MHZ DDR) AMD Athlon XP 1700+ MHz (1467/266 MHZ DDR) AMD Athlon XP 1600+ MHz (1400/266 MHZ DDR) AMD Athlon XP 1500+ MHz (1333/266 MHZ DDR) AMD Athlon 1400 MHz (1400/266 MHZ DDR) |
| Motherboard | Epox EP-8KHA+ (VIA KT266A) Revision: 2.0 |
| Memory | 256 MB DDR-SDRAM, CL2, PC2100, Micron |
| General Hardware | |
| Graphics card | GeForce 3 Memory: 64 MB DDR-SDRAM Memory clock: 400 MHz Chip speed: 250 MHz |
| Hard drive | 40 GB, 5T040H4, Maxtor UDMA100 7200 rpm 2 MB Cache |
| Drivers & Software | |
| Graphics card driver | Detonator 4 Serie V21.85 |
| DirectX version | 8.1 |
| Operating system | Windows XP Final, Build 2600 (Englisch) |
| Benchmarks & Settings | |
| Quake III Arena | Retail Version 1.16 command line = +set cd_nocd 1 +set s_initsound 0 Graphics detail set to ‘Normal’ Benchmark using ‘Q3DEMO1’ |
| 3DMark2000 | Version 1.1 Build 340 – default Benchmark |
| 3DMark2001 | Build 200 – default Benchmark |
| SiSoft Sandra 2001 | Professional Version 2001.3.7.50 |
| CINEMA 4D | CineBench 6.103 |
| mpeg4 encoding | Flask V0.6 (MPEG 3) DivX 4.02b codec Compression: 100 Data Rate: 1500 Kbit 720×480 Pixel, 25 fps no Audio |
| Studio 7 | Version 7.02.7 (MPEG 2) |
| Sysmark 2001 | Patch 3 |
| Lame | Lame 3.89 MMX, SSE, SSE 2, 3DNow |
| WinACE | 2.04, 178 MB Wave file, best compression, Dictonary 4096 KB |
| Cinema 4D XL R6 | CineBench |
| Suse Linux 7.3 | Kernel 2.4.13 Compiling |
Testing Procedure: Peculiarities, Continued
The picture shows the test system set-up for all the AMD Athlon XP processors.
The picture shows the test system set-up for all the Intel Pentium 4 processors.
Benchmarks: 19 Rigorous Tests
| OpenGL Performance | Quake 3 Arena “Demo 1” and “NV15 Demo” |
| Direct3D Performance | 3D Mark 2000 and 3D Mark 2001 |
| 3D Rendering | Cinema 4D XL R6 “Shading” |
| 3D Rendering | SPECviewperf “Lightscape” |
| MP3 Audio Encoding | Lame MP3 Encoder |
| MPEG-2 Video Encoding | Pinnacle Studio 7 |
| MPEG-4 Video Encoding | Flask Mpeg 0.6 and DivX 4.02 codec |
| Office Performance | Sysmark 2001 |
| Archiving | WinACE 2.04 |
| Compiling Linux Kernel | Suse Linux 7.3 (Kernel 2.4.13) |
| SiSoft Sandra 2001 | CPU and Multimedia Bench |
Since we want to convey the widest spectrum possible of practical benchmarks, we performed all the tests under Windows XP and completely re-configured the tests. In total, we used 19 different benchmarks in order to obtain the most complete, the most well-balanced view of how each individal AMD and Intel processor performs. We continued to determine OpenGL performance using four different Quake tests – Direct3D performance from the DirectX package is determined using the 3D Mark 2000 (Direct X 7) and the 3D Mark 2001 (DirectX8). The different MPEG-encoding benchmarks portray a comprehensive testing scenario – the Lame MP3 Encoder was used to encode a 178 MB Wav file into “MPEG-1 Layer 3 Format.” Still a classic, our MPEG-4 test converts a file from a DVD film into MPEG-4 format using Flask Mpeg and DivX codec. A new addition to our benchmark suite is MPEG-2 file encoding using the video editing software “Pinnacle Studio 7.”
OpenGL Performance: Quake 3 Arena
At the low resolution of 640 x 480, the Pentium 4/2000 is in the lead, while the AMD Athlon XP 1800+, boosted by the enhanced Nvidia driver, takes the lead at the high resolution.
Direct3D Performance – DirectX 7: 3D Mark 2000
3D Mark 2000 determines DirectX 7’s Direct3D performance under Windows XP. The diagram shows that all AMD processors come out better than the Intel CPUs. Interestingly, the older Athlon 1400 comes out ahead of the Athlon XP 1500+. The front-runner is the Athlon XP 1800+.
Direct3D Performance – DirectX 8: 3D Mark 2001
3D Mark 2001 determines DirectX 8’s Direct3D performance under Windows XP. Once again, the Athlon XP 1800+ is the king of the hill. The Pentium 4/2000 doesn’t even enter the picture until fourth place.
MP3 Audio Encoding: Lame MP3
The Lame MP3 Encoder under Windows XP is used to convert a 178 MB sound file from a WAV format to a “MPEG-1 Layer 3” format. In this discipline, the Intel Pentium 4/2000 makes a clean sweep of the competition. Nonetheless, the new Athlon XP CPUs don’t cut a bad figure, either.
MPEG-4 Video Encoding: Flask Mpeg And DivX
The new Divx 4.02 codec, used together with Flask Mpeg 0.6, caused quite a stir – the AMD Athlon XP 1800+ beats the Intel Pentium 4/2000, knocking the ex-gold medallist out of its top position.
This category of benchmarks is subject to change. Tom’s Hardware has knowledge that the new DivX codec 4.5 is almost ready for release and Flask Mpeg is also working diligently on a new version of its code. A new version of this software can easily change the outcome for both CPUs. It’s likely that more performance can be squeezed out of either processor by optimization.
SiSoft Sandra Benchmarks: CPU And Multimedia
In the SiSoft Sandra Benchmark 2001, the overall performance is uneven – both Athlon XP and Pentium 4 switch places several times.
3D-Rendering: Cinema 4D XL R6
In both Cinema 4D tests, AMD and Intel switch placings. In “normal” shading, the Intel Pentium 4/2000 is in the lead, while the AMD Athlon XP 1800+ takes the gold in “Shading Single CPU”.
Download of Cinebench 4D benchmark
Office-Performance: Sysmark 2001
The Pentium 4/2000 takes first place in the “unofficial score” category, while all AMD CPUs come in ahead of the Intel competition in the normal test.
A comment on Sysmark 2001 – despite the fact that AMD sent us a patch activating the recognition of the SSE command set in Media Encoder 7.1, which is included in the benchmark suite, we didn’t use it. The reason for this was that users can’t update the Media Encoder in reality. There is no official update from BAPCo yet.
Compiling Linux: Suse Linux 7.3 / Kernel 2.4.13
When it comes to compiling the latest Linux kernel, the AMD Athlon XP 1800+ is the fastest CPU. It needs a mere 223 seconds – whereas the Intel Pentium 4/2000 needs 264 seconds to compile the same kernel. Bringing up the rear, the Pentium 4 at 1400 MHz needs 329 seconds.
Archiving: WinACE 2.04
Archiving is a very practical application. WinACE 2.04 was used under Windows XP to archive a 178 MB file while the clock was running. In this discipline, the Intel Pentium 4/2000 is head and shoulders above the rest. The AMD Athlon XP, on the other hand, can’t quite seem to keep up – all its processors lagged behind the Intel CPUs. The worst performance came from the older AMD Athlon 1400.
3D Rendering Performance: SPECviewperf “Lightscape”
The “Lightscape” benchmark produced very clear results – the Pentium 4 CPUs, with their faster clock speeds, were clearly in the lead, while the fastest AMD processor, the Athlon XP 1800+, took fifth in the rankings. The big loser is the Intel Pentium 4/1400.
MPEG-2 Video Encoding: Pinnacle Studio 7
The Intel Pentium 4/2000 was the fastest at creating an MPEG-2 film using Pinnacle Studio 7. The Athlon XP 1800+ took second place – and the remaining rankings were split between Intel and AMD CPUs.
Asus: Athlon XP Support For Old Boards
On its homepage, Asus has listed all the boards that run with the new AMD Athlon XP processors.
Conclusion: Athlon XP Has The Lead Over Pentium 4
We’ve got our performance winner in this extensive CPU test – the AMD Athlon XP 1800+ tops the Intel Pentium 4/2000 in most of the applications benchmarks we selected. The Athlon XP’s strengths really lie with 3D games that use DirectX 7 or DirectX 8.
Another one of the reasons is Windows XP, being forced as the successor to Windows 98/ME and Windows 2000 by Microsoft. In comparison to the past, the Athlon XP can profit from Windows XP, probably more than Intel CPUs. It seems that AMD helped Microsoft get the optimum out of its new CPUs.
Even in MP3 audio encoding, the fastest Athlon XP was a nose faster that the faster clocked Pentium 4 processors. We were in for a surprise with the new MPEG-4-Codec DivX 4.02.1: when used together with Flask Mpeg 0.6, the Athlon XP 1800+ (clocked at 1533 MHZ) knocked the stuffing out of the Pentium 4, clocked at 2000 MHz. This is a watershed in CPU comparisons between AMD and Intel. Although the benchmarks favor AMD over Intel presently things can, and will change. A new version of the DivX codec or Flask Mpeg, can change the picture again. Higher performance gains can be had on either processor.
The picture is similar in 3D rendering (OpenGL) – the AMD Athlon XP’s three FPU units helped it to outstrip the Pentium 4, with 2 FPU units. Ideally, you can employ the following equation:
Performance = Clock Speed x Operations/Cycle
This equation helps explain the theory behind why the AMD Athlon XP, although clocked at a lower speed, is able to reach the same performance than a faster-clocked Intel Pentium 4. But one thing should be made clear here – since Intel introduced its 0.13-micron processor, the Northwood Pentium 4, it can turn the MHz dial up higher than AMD can. High clock speeds are a real burr under AMD’s saddle, as has been shown by the developments of the past several months. While Intel has already cleared the 2000 MHz hurdle, AMD has barely scraped over the 1533 MHZ one from 1400 MHz.
In this comparison, we’ve introduced a new evaluation criterion that may be of particular interest for home users – the value rating for each processor states which processor offers the most performance at the cheapest price. According to our calculations, the AMD Athlon XP 1500+ currently offers the best value among all CPUs.
Another factor is the stability and product quality of a system: while all Athlon processors suffered from occasional instability in our tests, the Pentium 4 platform ran without a glitch. Reasons for this behaviour might not lie in the processor itself, but rather in the motherboard design and the chipset used. Future driver updates might not just improve performance but also stability of a platform. And of course, every user knows that the lightweight price tags on Athlon XP processors may have a downside compared to their more expensive Pentium 4 rivals.
However, we still believe that AMD is offering its processors much too cheaply compared to Intel’s prices. A basic law of marketing should be that cheap doesn’t necessarily sound better. Otherwise, the customer will begin to think that the product offered is of inferior quality. This is precisely the area that AMD management needs to work on and get some measure of self-confidence.
The Intel Pentium 4 is supposed to benefit more from application software updates than the AMD Athlon XP, since developers are continuing to add SSE2 enhancements. Today’s crop of 3D games are no problem for either the Athlon XP or the Pentium 4. They run very smoothly at high resolutions and refresh rates. In this case, it was up to game developers to tailor the 3D engines to work better with new technologies such as SSE and SSE2. Not even DirectX 8.1, which was launched only recently, makes full use of Windows XP capabilities.
