AMD Athlon 64 4000+ & FX-55: A Thorough Investigation
by Anand Lal Shimpi on October 19, 2004 1:04 AM EST- Posted in
- CPUs
Workstation Applications
Visual Studio 6
Carried over from our previous CPU reviews, we continue to use Visual Studio 6 for a quick compile test. We are still using the Quake 3 source code as our test and measure compile time in seconds. The results are pretty much in line with what we've seen in the past.
Without compiler speed optimizations for NetBurst, the Pentium 4 architecture is not very well suited for highly branchy applications such as compiling and code optimization. The Athlon 64 is far better suited for this type of usage pattern and thus AMD wins the compiler test.
SPECviewperf 8
For our next set of professional application benchmarks we turn to SPECviewperf 8. SPECviewperf is a collection of application traces taken from some of the most popular professional applications, and compiled together in a single set of benchmarks used to estimate performance in the various applications the benchmark is used to model. With version 8, SPEC has significantly improved the quality of the benchmark, making it even more of a real world indicator of performance.
We have included SPEC's official description of each one of the 8 tests in the suite.
3dsmax Viewset (3dsmax-03)
"The 3dsmax-03 viewset was created from traces of the graphics workload generated by 3ds max 3.1. To insure a common comparison point, the OpenGL plug-in driver from Discreet was used during tracing.
The models for this viewset came from the SPECapc 3ds max 3.1 benchmark. Each model was measured with two different lighting models to reflect a range of potential 3ds max users. The high-complexity model uses five to seven positional lights as defined by the SPECapc benchmark and reflects how a high-end user would work with 3ds max. The medium-complexity lighting models uses two positional lights, a more common lighting environment.
The viewset is based on a trace of the running application and includes all the state changes found during normal 3ds max operation. Immediate-mode OpenGL calls are used to transfer data to the graphics subsystem."
The biggest surprise here is that there is a huge performance impact (13%) by moving down to a single channel memory subsystem with the Athlon 64. There are also a couple of other surprises, with the Pentium 4 560 doing surprisingly well, coming in at the heels of the Athlon 64 FX-55.
CATIA Viewset (catia-01)
"The catia-01 viewset was created from traces of the graphics workload generated by the CATIATM V5R12 application from Dassault Systemes.
Three models are measured using various modes in CATIA. Phil Harris of LionHeart Solutions, developer of CATBench2003, supplied SPEC/GPC with the models used to measure the CATIA application. The models are courtesy of CATBench2003 and CATIA Community.The car model contains more than two million points. SPECviewperf replicates the geometry represented by the smaller engine block and submarine models to increase complexity and decrease frame rates. After replication, these models contain 1.2 million vertices (engine block) and 1.8 million vertices (submarine).
State changes as made by the application are included throughout the rendering of the model, including matrix, material, light and line-stipple changes. All state changes are derived from a trace of the running application. The state changes put considerably more stress on graphics subsystems than the simple geometry dumps found in older SPECviewperf viewsets.
Mirroring the application, draw arrays are used for some tests and immediate mode used for others."
The single channel Athlon 64 3400+ does exceptionally poorly here, with the dual channel Athlon 64 parts holding a significant performance advantage. By now it's no surprise to see the FX-55, 4000+ and 3800+ at the top of the charts.
Interestingly enough, the CATIA benchmark appears to favor Intel's Prescott core over Northwood.
EnSight (ensight-01)
"The ensight-01 viewset replaces the Data Explorer (dx) viewset. It represents engineering and scientific visualization workloads created from traces of CEI's EnSight application.
CEI contributed the models and suggested workloads. Various modes of the EnSight application are tested using both display-list and immediate-mode paths through the OpenGL API. The model data is replicated by SPECviewperf 8.0 to generate 3.2 million vertices per frame.
State changes as made by the application are included throughout the rendering of the model, including matrix, material, light and line-stipple changes. All state changes are derived from a trace of the running application. The state changes put considerably more stress on graphics subsystems than the simple geometry dumps found in older viewsets.
Mirroring the application, both immediate-mode and display-list modes are measured."
No 925X based test system would complete this test, thus all we have are AMD chips to look at. Once again we see that the 3400+ is seriously crippled by its single channel memory interface.
Lightscape Viewset (light-07)
"The light-07 viewset was created from traces of the graphics workload generated by the Lightscape Visualization System from Discreet Logic. Lightscape combines proprietary radiosity algorithms with a physically based lighting interface.
The most significant feature of Lightscape is its ability to accurately simulate global illumination effects by precalculating the diffuse energy distribution in an environment and storing the lighting distribution as part of the 3D model. The resulting lighting "mesh" can then be rapidly displayed."
Maya Viewset (maya-01)
"The maya-01 viewset was created from traces of the graphics workload generated by the Maya V5 application from Alias.
The models used in the tests were contributed by artists at NVIDIA. Various modes in the Maya application are measured.
State changes as made by the application are included throughout the rendering of the model, including matrix, material, light and line-stipple changes. All state changes are derived from a trace of the running application. The state changes put considerably more stress on graphics subsystems than the simple geometry dumps found in older viewsets.
As in the Maya V5 application, array element is used to transfer data through the OpenGL API."
Pro/ENGINEER (proe-03)
"The proe-03 viewset was created from traces of the graphics workload generated by the Pro/ENGINEER 2001TM application from PTC.
Two models and three rendering modes are measured during the test. PTC contributed the models to SPEC for use in measurement of the Pro/ENGINEER application. The first of the models, the PTC World Car, represents a large-model workload composed of 3.9 to 5.9 million vertices. This model is measured in shaded, hidden-line removal, and wireframe modes. The wireframe workloads are measured both in normal and antialiased mode. The second model is a copier. It is a medium-sized model made up of 485,000 to 1.6 million vertices. Shaded and hidden-line-removal modes were measured for this model.
This viewset includes state changes as made by the application throughout the rendering of the model, including matrix, material, light and line-stipple changes. The PTC World Car shaded frames include more than 100MB of state and vertex information per frame. All state changes are derived from a trace of the running application. The state changes put considerably more stress on graphics subsystems than the simple geometry dumps found in older viewsets.
Mirroring the application, draw arrays are used for the shaded tests and immediate mode is used for the wireframe. The gradient background used by the Pro/E application is also included to better model the application workload."
When the Athlon was first released, it was a very solid performer in Pro/ENGINEER. But of course, back then, very few companies would think about Athlon workstations. With the Athlon 64 times have obviously changed, but the performance advantage does not seem to have changed at all, AMD continues to lead the way in the proe-03 viewset.
SolidWorks Viewset (sw-01)
"The sw-01 viewset was created from traces of the graphics workload generated by the Solidworks 2004 application from Dassault Systemes.
The model and workloads used were contributed by Solidworks as part of the SPECapc for SolidWorks 2004 benchmark.
State changes as made by the application are included throughout the rendering of the model, including matrix, material, light and line-stipple changes. All state changes are derived from a trace of the running application. The state changes put considerably more stress on graphics subsystems than the simple geometry dumps found in older viewsets.
Mirroring the application, draw arrays are used for some tests and immediate mode used for others."
Unigraphics (ugs-04)
"The ugs-04 viewset was created from traces of the graphics workload generated by Unigraphics V17.
The engine model used was taken from the SPECapc for Unigraphics V17 application benchmark. Three rendering modes are measured -- shaded, shaded with transparency, and wireframe. The wireframe workloads are measured both in normal and anti-alised mode. All tests are repeated twice, rotating once in the center of the screen and then moving about the frame to measure clipping performance.
The viewset is based on a trace of the running application and includes all the state changes found during normal Unigraphics operation. As with the application, OpenGL display lists are used to transfer data to the graphics subsystem. Thousands of display lists of varying sizes go into generating each frame of the model.
To increase model size and complexity, SPECviewperf 8.0 replicates the model two times more than the previous ugs-03 test."
Facebook
Twitter
RSS
89 Comments
View All Comments
mlittl3 - Thursday, October 21, 2004 - link
Oh and one more thing and then I will shut up. With regards to trusting your research, how about I trust Anandtech's, Tomshardware's, Xbitlabs', Acehardware's, [H]ardocp's, etc. research over yours. None of these sites have made an official statement that you are mad to choose AMD over Intel. I will trust their research before I trust yours (unless you run a hardware review site that proves that AMD processors are less stable and slower than Intel processors).mlittl3 - Thursday, October 21, 2004 - link
t, Valeria was refering to the AMD 8000 series chipsets lacking the features you listed. If you want an AMD chipset, you will have to have third-party addon chips to support SATA, firewire, etc.Valerie, you said you liked it when people answered your comments so here goes.
"More competitors to the battlefield or some kind of regulation is needed."
I agree with you here. We need companies like AMD because they help continue progress and make equal if not better processors than Intel, forcing Intel in turn to make equal or better processors than AMD and so on.
"...that AMD madness will end one time..."
This contradicts your statement about competition. You want regulation and competition but you want AMD gone. Why do you hate other companies besides Intel?
"Name me one company which prefers AMD and doesnt produce intel, name me one industrial computer who support AMDs, one automotive rack test system provider, hospital equipment, avionic systems, and so on"
According to your request for more competition, why would you want a company that only supports AMD. If all companies support all processors, we have the ability to make choices based on the strong points of each technology for a given application that we desire.
"its about quality and support what you will never get from AMD/taiwan"
I don't understand why you typed Taiwan (refering to Via I guess), but anyway, Intel has delayed/cancelled products over 10 times in the last year. They cancelled Tejas and 4GHz Prescott. They delayed Dothan, Whitefield (dual-cores I believe) and Itanium progress has been delayed like crazy. AMD has had a lot less setbacks. I don't know what the definition of quality is in Europe, but here in America, such cancellations and delays are unacceptable.
"Get Intel, and dont fall to temporaly madness"
No one will respect your "research" with statements like these. Again you contradict your desire for competition. No one is mad to choose AMD, Via, Transmeta, Apple, IBM, Sun, etc. Everyone has a specific application and no one processor does well at everything. Do you know anything about the following processors?
Via C3
Transmeta Crusoe
IBM Power series (4,5)
Sun Ultrasparc (III, III+, IV, IV+)
Apple G4 (Motorola) G5 (IBM)
Do you think people who choose these are "mad" or do you only hate AMD?
"Respected companies doesnt change so fast"
My respect for your research is getting worse. Intel has changed their mind about issues left and right in short periods of time (sorry but I have to say it, they are "flip-flopping"). At one point, they insulted and ridiculed model numbers for processors. They also said that only megahertz(hurtz) matters. They also said that it is not the time for 64-bit. They have changed their position on all these points. Intel changing their minds so much is not a good quality a company should have that is supposedly "leading" the industry.
"everybody makes mistakes, but with intel you have allways choice"
If Intel and AMD both make mistakes, why are you so much in bed with Intel. If you want all the "mad" people to get over AMD and only go with Intel, how is this choice? How do you have choice without competition. If Intel only existed, then we would have no choice. You contradict yourself again.
"now i am happy that AMD is here, same as i am happy that there is cheap cars to buy, competing is good for all."
You are all over the place. First you say that people should get over their temporary madness now you are glad AMD is here. What are you trying to say? If competing is good for all, then we should have many processor companies driving technological progress. One day, AMD might be bigger than Intel. Do you think Intel will be the largest company for processors over the next 100 years? No one can predict that. Are you going to tell people they are "mad" when they choose Intel if AMD one day gets larger than Intel. With regards to your cheap comments, Intel and AMD both make expensive and cheap processors. Their price points are almost identical. Do some freaking research why don't you?
"Stop that anti intel BS which is based on fool synthetic benchmarks and theoretical "if than" visions"
You obviously have not done any research. Intel wins the majority of the synthetic benchmarks. That is why many companies pick them because there are so many apps out there, synthetic benchmarks represent an overall processor goodness. However, hardware enthusiats know that synthetic benchmarks don't mean crap, and they hunt down a review site that did tests with the application they are interested in. People who play games buy AMD, people who do video/audio encoding and rendering pick Intel, people who do scientific apps pick AMD, etc. (I know this is always changing people so don't flame me). Try doing your research again.
"I did lot of research, trust me."
How about I do not trust you based on your above statements which contradict themselves constantly. I believe you are an Intel Fangirl just like an AMD fangirl/boy with very little info.
"We need better chipsets, better support from OS and for developers to be close to intel."
Again, you are contradicting your statements about competition. How are other companies supposed to compete if they only work with Intel? This is the problem we have with the horrible Windows operating system. Developers only work with Microsoft and we are plagued with problems because of that monopoly.
Sorry, Valerie, but your statements are flawed and they contradict themselves. You are an Intel Fangirl and nothing else. Both AMD and Intel Fangirls/boys never present a valid argument. Technology is technology. It benefits mankind regardless of whether or not AMD or Intel is in front of the name. People will pick a product that suits their needs. Both AMD and Intel provide for different needs from different people.
blckgrffn - Thursday, October 21, 2004 - link
Val!Enough!
It is really obvious that you have no clue what you are talking about! You made your "points" and now it is time to stop. The comments you made here will be remembered and if you try to make any discussion in the future, not matter how valid, you will be dismissed as the intel fan"girl" that you seem to be. Whatever, I am glad that someone will stick with intel through their development slump...
PrinceGaz - Thursday, October 21, 2004 - link
#72 Val: "[AMD] Is cheaper, but not much. Whats the difference some stupid 100euro for CPU which lasts for one year? I dont think that they are worse because they are cheaper, but they are cheaper because they are worse."Its news to me that AMD processors only last one year. My Athlon XP 1700+ purchased in Dec 2001 and used almost continually for the three or so years since then is still working fine. I must have been very lucky. I've been doubly lucky as my other machine, with an AMD K6-III/400 purchased early in 1999 is also still running without any problems.
Do you have a link to a reliable source which supports your statement that AMD processors only last one year? Or is this just some more BS you invented?
t - Thursday, October 21, 2004 - link
74:why exactly do "we need...better support from OS and fro developers to be close to intel"?
given that there are a number of x86-64 linux distrobutions available now... and linux is a fast growing segment of the server market, it would seem that the 'better support' is in place. Remember that it is intel who is also releasing x86-64 based xeons... as a direct response to AMD's strategy.
http://www.realworldtech.com/page.cfm?ArticleID=RW...
also, sun are porting solaris (like it or lump it, its still a major player in the server market) to x86-64
and back to chipsets... one of the majore chipset suppliers for xeon systems is serverworks, who are also supporting opteron
http://www.eweek.com/article2/0,1759,1604929,00.as...
so...it would seem that both OS support AND chipset support is in place, or rapidly falling so.
intel is a miniscule percentage of the server market for non 32bit x86 systems.... see the realwordtech article linked above.... most 'serious' server developers don't work with intel, because intel systems don't make up a high percentage of the market...instead u have sun/hp/ibm being the majority.
heh, 'special boards'... im sure tyan would happily sell u an amd OR intel based board for a hefty price above any mainstream product. its not like intels server boards are dirt cheap either...
what do you mean 'affordable mainboard with enough features'... the only 'feature' lacking from amd system chipsets is ddr2. They have sata, pci-e, raid, etc, etc.
i dont wanna be all condescending...but if you did 'a lot of research' and want ppl to 'look away from [their] overclocked gaming machines', perhaps you should practise what you preach a little... or do some more extensive research.
the world is not x86, and in the non x86 market, it is definately not intel.
out of curiousity... what are your thoughts on the non intel power5 processors? :)
t.
val - Thursday, October 21, 2004 - link
71:>As to verification... seeing as sun/ibm/hp
> offer k8 based servers, they HAVE to be
> verified, you know tier1 oem..that kinda
> thing.
where AMD chipset is used and completely different design than you can ever buy for normal PC. so the efect is same. You are right in that k8 is changing things (for better for AMD) but it is still far from perfect. We need better chipsets, better support from OS and for developers to be close to intel.
>also...amd do make their own chipsets, always >have....just that sis/via/nvidia chipsets are >more targetted at the mainstream.
yes they are much worse and with AMD chipset you cannot buy affordable mainboard with enough features. Not counting those special mainboards.
So it is nothing else than what i said: no good chipset available.
Valerie
Gnoad - Thursday, October 21, 2004 - link
god, someone just end it already. Its obvious arguing our points is gonna accomplish anything here, so lets just let it be now.val - Thursday, October 21, 2004 - link
68: i dont have only celerons. And i am not him, but her.67: at least somebody answers my questions, unfortunately only some and catching only wrong data out of them.
1.Who cares that 1.7 AMD outperforms it? 1.7 AMD is rated 2000+ and if you tell me that you need 2000+ to play movies, get mails and write office stuff, than heh. i have celeron 2 Ghz and even video encoding is fast on it (23% under athlon 64 3200+) who cares? Far cry i have 60 FPS, Doom3 30 and race driver 2 67 fps. Please do not make me tired.
2. Mostly outsourced as i heard, i am not sure right now here.
3. They don't! They should rather take few bucks they are giving to marketing company which is one year raising PR numbers with same CPU frequency and give them to microsoft to include support in VC++ and speed up the XP64bit. There is no good support for hardware developers at all. No discussion here, look to industrial business.
4. Is cheaper, but not much. Whats the difference some stupid 100euro for CPU which lasts for one year? I dont think that they are worse because they are cheaper, but they are cheaper because they are worse. I am 19+ and dont call me kid.
5. PR filtered.
I did lot of research, trust me. seems like you dont. Please look away from your overclocking gaming machines.
>go buy a Dell, you silly ignorant little fool.
i hope you feel big and smart, not like those idiots making ten times more moneys then you.
t - Thursday, October 21, 2004 - link
Val: What i am trying to tell you, that AMD is only manufacturing what somebody else designs, they dont care and dont support products and developers, dont care for chipsets quality, certificates, nothing. This leads to that overall quality of AMD platform is far from what you can get with similar price based on Intel.umm....who designs them then? oh thats right...the k7/k8 were largely designed by the alpha engineers that AMD hired after hp killed off the alpha. Intel also have a large body of ex-alpha guys, mainly working on the itanium iirc.
oh...AMD also have a fab tech agreement with IBM.
unless u mean that AMD only makes intel compatible chips? In that case, yes, yes they did, until the K8 when AMD extended the x86 ISA. Making compatible, ie. x86 processors is also what transmeta/via do. And anyway, intel and AMD have a cross patent agreement, meaning that they CAN copy each other if they so wish.
As to verification... seeing as sun/ibm/hp offer k8 based servers, they HAVE to be verified, you know tier1 oem..that kinda thing.
also...amd do make their own chipsets, always have....just that sis/via/nvidia chipsets are more targetted at the mainstream.
sorry for the huge post.
Gnoad - Wednesday, October 20, 2004 - link
Thats a good pic. I wonder if Intel actually tried testing to see how prescott performed at 4ghz before they released it. I know I would have, but it seems they just kinda let it loose then realized, "Oh crap, we screwed up."