AnandTech's Integrated Video Chipset Roundup

Modus

Platinum Member
Oct 9, 1999
2,235
0
0
It may seem a bit obsessive to pick on a review of hardware that most people on this board consider a waste of time, but you must understand that many consumers have not the slightest interest in fragging a Quake3 monster. The vast majority of computers purchased today are used primarily for standard office/content creation tasks, and this is where integrated video chipsets show their value. If AnandTech wants to attract a broader audience and be the top source for those who make the purchasing decisions for such machines (computer resellers and consultants, network administrators, IT managers), they need to put just as much effort into their reviews of so-called "value" products as of "performance" products.

But AnandTech's recent Integrated Video Chipset Roundup leaves too many questions unasked.

1) Is it wise to use a batch of reference boards in such a comparison?

We've seen time and again how performance can vary a great deal from an engineeing sample to a final product. While AnandTech's reviews are usually most thorough, one consistent criticism has been their relative lack of attention to the true status of a product: Is it an engineering sample? How close is it to the final product? Are units shipping at this moment? How do those differ from the unit being reviewed? What room is their for improvement? Any hardware enthusiast looking for a solid judgement on a product will tell you that the one thing they can't stomach is uncertainty. There's nothing worse than reading a nice review and wondering if the benchmarks will be meaningless in two weeks due to new BIOS/driver/PCB revisions. This is especially true in the case of brand new chipsets -- manufactures work with chipset designers to implement dozens of little BIOS tweaks that collectively can have a significant impact on performance.

As we speak, KM133 motherboards are available from Microstar and Chaintech, and SiS 730S boards are shipping from PC Chips. Why not get your hands on these products and evaluate what consumers will actually be using, as opposed to engineering samples and reference boards that will never see the light of day?

2) Why does the KT133 perform 2 to 5% better than the KM133 on standard Office / Content Creation benchmarks when using an external graphics accelerator? Memory bandwidth issues should be non-existant here, leading one to believe that the deficiency is due to a pre-release, unoptimized board. But we can't say for sure. That's the problem.

3) Why is the Duron just barely edging out the Celeron using integrated video on Office / Conent Creation benchmarks?

Supposedly the ProSavage and SiS 300 graphics cores are not as fast in 2D as their Intel counterpart. But does 2D performance even matter today? Has anyone in the past five years ever been doing some computer work and thought to themselves, "Boy I wish that display speed was faster"? Performance in 2D screen drawing has been better than human vision ever since the advent of the PCI bus. I can't possibly imagine a situation where 2D speed is an issue these days.

Here's a possible explanation: the Intel chipset's more efficient memory controller enables it to better cope with the bandwidth burden of an integrated UMA archtiecture, resulting in less of a performance decrease for other memory-intensive processes running concurrently with the integrated video. Is this correct? I don't know. But it should have been explored, because we all know that the Duron should be wiping the floor with the Celeron, not just politely edging it out.

Perhaps current office/content creation benchmarks are too reliant on 2D performance. After all, the real delay in most business environments is hard drive access due to virtual memory when running multiple applications. Maybe we need some benchmarks that produce scores more realistic with everyday Windows usage. SysMark 2000, for instance, makes a big deal of the time it takes to apply a filter onto an enormous graphic in Photoshop (of the size that no one but professional DTP workers will encounter; typical web graphics are puny, filtered in milliseconds), to convert a digital media file with Windows Media Encoder 4.0, or to apply a special effect in Elastic Reality. Is that the kind of stuff your average SOHO worker is doing on a daily basis? Of course not. We need benchmarks that stress heavy business multitasking, which is where even today's systems can look slow. Just look at a typical business machine and you'll find about a dozen useless apps running in the system tray, a pointless virus scanner inspecting every single file the OS touches, and a backup tool constantly archiving data, while the user is doing relatively mundane things like accounting with Quicken or web browsing with IE or correspondence with Word. Disc access and memory performance are the keys here; 2D speed is as fast as it'll ever need to be.

4) Why is the Celeron consistently beating the Duron in gaming benchmarks with integrated video?

Even when the integrated graphics core remains the same (ie. KM133 vs PM133 or 730S vs 630S) the Celeron 800 scores higher than the much more powerful Duron 800 on 3D gaming benchmarks. What is to blame here? I have no idea, and the review didn't even raise the issue.

5) Why not examine the DVD performance issue with the KM/PM133?

Obviously something is wrong with the benchmark or the drivers or both. It's a bit misleading because CPU utlization while watching a DVD shouldn't be a big deal anyway -- you're not trying to play Quake at the same time. But there should have been some research into the problem, because the Savage4 has historically featured excellent DVD performance, and its Hardware Motion Compensation and iDCT features should at least allow it to beat the ALi Alladin TNT2 which lacks them.

Modus
 

Modus

Platinum Member
Oct 9, 1999
2,235
0
0
And you could add

5) What size frame buffer was used for each chipset? Even in a UMA environment, there can be a significant performance difference between 4M and 8M shared buffer size.

6) How does the integrated ProSavage core compare to an external Savage4 Pro card?

This could be easily tested by increasing the shared video memory to 32M, adding another 32M SDRAM chip, benchmarking it, and then yanking out the extra 32M of RAM and popping in a 32M Savage 4 Pro card. Then we would really see the effect of UMA's decreased memory bandwidth.

Modus
 

Modus

Platinum Member
Oct 9, 1999
2,235
0
0
Oh and in connection with 6) you could do the same thing with the SiS chipsets as there are separate Sis 300 AGP cards available.

Modus
 

ET

Senior member
Oct 12, 1999
521
33
91
Interesting questions. I was thinking that the articles forums is the place for them, but I guess that more audience can't hurt, since it seems that just commenting in the articles forum didn't bring about any action (regarding the DVD issue, for example).

BTW, the Savage4 core only has motion compensation, and not iDCT. Still, it definitely shouldn't perform any worse than the TNT2 core, and should be around the performance of the i815.

As for the SiS300, it would have been nice to have a review of a card based on it. Always wanted to know what it's worth. Anyway, I definitely think that there's need to discover why the SiS630 gives half the performance of the SIS730 in UT.
 
Jan 16, 2000
94
0
0
Modus, please allow me to address your questions.

1. The "reference boards" used in this comparison were used because these boards provide an accurate measure of the chipset performance over a wide variety of solutions. These are not pre-release boards, but claim to offer the same performance and stability that "retail" products. The reference boards used in the benchmarks will be the same boards that manufacturers turn to when making their retail products. To answer your question, all the "reference boards" used are precursors to the final product and thus should be an accurate gaguge of the chipset's performance.

2. The most likely cause of the slower performance of the KM133's ProSavage integrated video controller when in 2D. As we were made certain to point out in our KM133 review that is linked in the integrated chipset roundup, the "Savage 2000 2D is still the slowest [2D] in the industry." For more information about the 2D bottleneck faced with the ProSavage video controller (which uses the Savage 2000's 2D controller), please see the KM133 review.

3. As we discussed in both the
KM133 review as well as the SiS 730S review, the Celeron is able to beat out the Duron when using integrated video solutions due to the slow 2D performance that the Socket-A solutions provide. It is true that the benchmarks used do stress a card's 2D performance, however it does so in an accurate way. Just watching benchmarks such as SYSmark 2000 and Content Creation Winstone 2001 run shows that these benchmarks do accurately simulate actually computer use. You may not think of it as a common function, but moving pictures, graphs, and objects does require 2D performance. Since the business and content creation benchmarks are aiming to finish the benchmarks in as little time as possible, this often means that items on the screen are moving around very quickly, even more so then would be present during normal computer use.
Add the 2D performance with the overall lack of maturity in the integrated Socket-A platform (even some things had to be changed to get the ProSavage controller on the KM133 chipset) and you can understand why these platforms are they are limiting the performance of the Duron and putting a hamper on the performance advantages the Duron has.

4. Please refer to number 3.

5. An explanation of the poor DVD performance of the KM133 and PM133 chipsets is mentioned briefly in the roundup ("This is most likely due to the fact that the ProSavage chip seems to be using the Savage4's hardware motion compensation that is most likely not supported in our DVD tests&quot as well as in the KM133 review's DVD performance section. Since there is no benchmark that uses older or proprietary DVD players, it is rather hard to see exactly how the chipset will perform using one of these other options.
As far as your claim that CPU utilization should not matter while watching a DVD, I believe this is a bit of a stretch. First off, the higher the CPU utilization, the harder the CPU is working to keep up with the DVD stream, meaning that your system is much more likely to experience choppy video playback or lapses in audio. Secondly, when a CPU is dedicating up to 79% of its power to DVD playback it has very little room to do other tasks. Even while in Windows the functions of the operating system require CPU power just to keep the system running. Tag onto this the CPU power needed to run virus checkers, messaging services, ect and you once again run into the problem of choppy DVD playback. Perhaps the biggest concern, however, should come with the fact that the software based sound that many of these platforms will end up using also require CPU power, and when 79% of the CPU's power is going to just video playback, you have yourself a problem.

6. The frame buffer size used in each solution was set to 8MB. Increasing the size of the frame buffer should result in no performance increase, simply because at 640x480x16 and 800x600x16 there is no way that a scene will require 8MB of frame buffer with the games tested as well as almost every other game out there.

7. There is no question that the integrated ProSavage solution performs slower than the Savage4 in 3D gaming. It does not seem relevant how much faster the ProSavage performs when compared to an external Savage4 video card. There is no question that the external Savage4 card would perform better, as would a GeForce2 MX, an ATI Radeon SDR, and a 3DFX Voodoo4 4500. The goal of the review was not to compare the integrated video solutions with external AGP solutions but rather to show how the solutions compare to one another. In addition, VIA does not claim that the ProSavage integrated graphics controller is the same as the Savage4, rather they say it is a combination of the Savage4 and the Savage2000.

I hope this answers your questions. Thanks for the feedback.
 

Modus

Platinum Member
Oct 9, 1999
2,235
0
0
Matthew Witheiler,

<< To answer your question, all the &quot;reference boards&quot; used are precursors to the final product and thus should be an accurate gaguge of the chipset's performance. >>

That's a non sequitor: if they are precursors to a final product, then by definition they cannot represent exactly its performance. Now, with video chipsets this might not be as much of an issue, but we all know how sensitive third-party motherboard chipsets are to the BIOS and driver tweaks implemented by most major manufacturers.

<< These are not pre-release boards, but claim to offer the same performance and stability that &quot;retail&quot; products. >>

Which is rarely the case. Surely you've seen how engineering samples and reference designs are often beaten by the final shipping product due to BIOS/driver/PCB revisions. The SiS solutions, in particular, seem to have much room for driver improvements. Wouldn't it have been prudent to at least mention that in the review? Something like &quot;please bear in mind that all the motherboards used in this roundup are reference parts from the chipset designers themselves, and may not perform as well as the final, optimized boards commonly available.&quot;

<< 2. The most likely cause of the slower performance of the KM133's ProSavage integrated video controller when in 2D >>

You didn't read the question. It was, &quot;Why does the KT133 perform 2 to 5% better than the KM133 on standard Office / Content Creation benchmarks when using an external graphics accelerator?&quot; With an external AGP card, the KM133 should perform identically to the KT133. . . Do you still claim the KM133 reference board is as good as the final product?

<< As we discussed in both the KM133 review as well as the SiS 730S review, the Celeron is able to beat out the Duron when using integrated video solutions due to the slow 2D performance that the Socket-A solutions provide. . . Add the 2D performance with the overall lack of maturity in the integrated Socket-A platform (even some things had to be changed to get the ProSavage controller on the KM133 chipset) >>

First of all, the Celeron is NOT able to beat the Duron when using integrated video solutions in office and content creation benchmarks. Just look at page 12 of your own review. The Sis 730S always beats the Socket 370 chipsets, and the KM133 only looses out to the Socket 370 chipsets on one of three tests.

Second, why should the 2D performance of the KM133 be at all slower than the 2D performance of the PM133? Neither could be called mature. You claim &quot;some things had to be changed to get the ProSavage controller onto the KT133 chipset.&quot; Says who? What things? You sure didn't mention that in the review. It seems like you're theorizing explanations for questionable results instead of figuring out how to remedy the those results.

<< Just watching benchmarks such as SYSmark 2000 and Content Creation Winstone 2001 run shows that these benchmarks do accurately simulate actually computer use. . . Since the business and content creation benchmarks are aiming to finish the benchmarks in as little time as possible, this often means that items on the screen are moving around very quickly, even more so then would be present during normal computer use. >>

That's another non sequitor: if the SYSMark 2000 and CC Winstone 2001 are &quot;real world benchmarks that accurately simulate actual computer use,&quot; then by definition they should not be stressing 2D performance so much more so than would happen under normal human use. Right?

<< 4. Please refer to number 3. >>

Number 3 doesn't have anything to do with it. We're talking about 3D gaming performance now. Why does a C/800/PM133 beat a D/800/KM133 by a full 10% on your Quake3 tests when VIA Hardware's KM133 preview showed a C/667/PM133 loosing to a D/600/KM133 by the duron by the same margin? Could it be that their review used boards from DFI while yours were just reference designs? I have no idea.

The results are obviously questionable, and that should have been addressed in the review. It makes AnandTech look so much better when they actually investigate certain results rather than take them at face value. Notice how Tom wrote three or four different updates to his P4 review? Your P4 review was still better, but it makes him look very dependable when he investigates anomolous results.

<< An explanation of the poor DVD performance of the KM133 and PM133 chipsets is mentioned briefly in the roundup (&quot;This is most likely due to the fact that the ProSavage chip seems to be using the Savage4's hardware motion compensation that is most likely not supported in our DVD tests >>

Then you should have figured out to fix the problem. The way you've done it makes the KM133 look unecessarily bad for DVD playback. Even forcing the decoder to software-only mode would have done the trick. Obviously, more research was needed. You're putting the blame on the chipset when the software is obviously at fault.

<< As far as your claim that CPU utilization should not matter while watching a DVD, I believe this is a bit of a stretch. >>

I never said it shouldn't matter, I said it wasn't a big deal. And it isn't. If you can watch a DVD with sound enabled at less than 100% CPU utlization, you should be OK. An MPEG2 movie essentially demands your full attention -- you won't be defragging your hard drive or rendering Toy Story in the background. And a straight AC'97 sound stream isn't too demanding on today's CPU's (though that would make a perfect item for a quick review/investigation).

<< The frame buffer size used in each solution was set to 8MB. Increasing the size of the frame buffer should result in no performance increase >>

Another thing that needed mentioning in the review itself. As for the shared memory size, there would be a performance increase going from 4M to 8M because the frame buffer only occupies 4M and the rest is reserved for texture space.

<< There is no question that the integrated ProSavage solution performs slower than the Savage4 in 3D gaming. It does not seem relevant how much faster the ProSavage performs when compared to an external Savage4 video card >>

It's relevant because it goes to the question of how much of a memory bandwidth hit is forced by a UMA integrated video solution.

Modus
 

BigD01

Junior Member
Aug 11, 2000
10
0
0
What I really want to know is if any of these solutions let you use both an AGP card and the integrated video chip. Are all the integrated video solutions running on the AGP bus or do they use the PCI bus? I know the 815 can't but I'm wondering about the other vendors solutions. I believe this would be an extremely useful feature. My $0.02
Don
 

formulav8

Diamond Member
Sep 18, 2000
7,004
522
126
I agree with Modus. That was a pretty crappy review. You guys usually explain certain reasons and so forth on why this or that is performing the way it is. Could it be you guys are turning bias on us? j/k
 

TravisBickle

Platinum Member
Dec 3, 2000
2,037
0
0
I have an integrated video myself and I can't say I found the review very encouraging anyway. for 3d they are horrible. I do think it's for bargain basement and business.
maybe with ddr or the bandwidth-reduced new chips integrated video will be a lot better, but it will always lag, obviously. I would add, though, that DVD performance is a standard requirement of a PC today, and the issue of cpu utilisation is not to be swept under the carpet.
 

ET

Senior member
Oct 12, 1999
521
33
91


<< &quot;This is most likely due to the fact that the ProSavage chip seems to be using the Savage4's hardware motion compensation that is most likely not supported in our DVD tests&quot; >>



That was an okay response when the review was posted, three weeks ago. Even then I didn't like it, and said the IMO you should have checked this with VIA and Intervideo. But three weeks have passed since, which IMO should have been more than enough time to get a reply on this. I can only assume that you don't really care about getting to the bottom of this. You could have tried other DVD players, playing with the options, or whatever, and maybe you could have found a good tip for DVD performance on this chip. But you didn't even try. That's what's so disappointing.
 

theplanb

Golden Member
Jan 12, 2001
1,239
0
0


<< , KM133 motherboards are available from Microstar and Chaintech, and SiS 730S boards are shipping from PC Chips. Why not get your hands on these products and evaluate what consumers will actually be using, >>


I'd like to know how they are performing..too
I think the PC-CHip's board doesn't come with the external AGP port though.
 

Modus

Platinum Member
Oct 9, 1999
2,235
0
0
Sorry for abdonding this thread, I've been pretty busy lately. Well, I got my hands on a Microstar K7TM Pro (6340). It's been available for the past week. It's rock solid so far, though it doesn't have the forest of massive capacitors that the K7T Pro-2a sports. I'll try to benchmark it against a new Matsonic KT133 if I have time. Both systems are Duron 800's with 128M PC133 CAS3 and Maxtor DiamondMax Plus 45 20G 7200rpm drives. The KT133 uses a Chaintech GeForce 2 MX. I feel this is a very typical mid range confiuration.

Modus
 

OC-Luvin

Junior Member
Mar 15, 2000
7
0
0
While reading this article I couln't help but notice that for the integrated video bench marks, anandtech used a Celeron 800 which currently sells for about $204. As everyone knows the processor uses the 66.67 Mhz bus. Now, to my knowledge only the SiS and Via chipsets are able to run the memory asynchronously up to 100 Mhz in this case (I don't know if they were run at 100 Mhz for these test). My questions are:

1) Why didn't Anand use a 133.33 Mhz chip like the 866.67 Mhz PIII which currently sells for about $210 ($6 more than the 800 Mhz Celeron)? This would be double the memory speed over the celeron.

2) What's the use of PC133 CAS2 memory in a &quot;budget&quot; celeron system&quot;? PC100 CAS3 is more than enough for a celeron and it's cheaper too. One can even set it to CAS2 as the memory speed will only be 66.67 Mhz.

In a nutshell retest using the 133.33 Mhz. memory speed with CAS2 memory and see how that stacks up in 2D.
 

OC-Luvin

Junior Member
Mar 15, 2000
7
0
0
I stand corrected.

Now did the SiS and Via boards use the 133 Mhz memory speed? I know that the 815e didn't. Also it would have been nice if the 810e were tested along with these other chipsets all at 133 Mhz.
 

HockeyPlayer

Junior Member
Jan 12, 2001
1
0
0
Actually one of thing I found missing was the 2D image quality test. At work they upgraded me from a Pentium 3/450/BX chipset with an on-the-motherboard ATI video chipset to an Pentium 3/733/i810 chipset. I went back to my old computer because the video quality at 1280x1024 was terrible. I have a 21&quot; monitor so the monitor was not the issue. In addition the sound quality was so bad I could not even listen to MP3s. It sounded like the vocals were recorded in a bathroom.
 

Modus

Platinum Member
Oct 9, 1999
2,235
0
0
OC-Luvin,

<< Why didn't Anand use a 133.33 Mhz chip like the 866.67 Mhz PIII which currently sells for about $210 ($6 more than the 800 Mhz Celeron)? >>

The reviewer probably intended to evaluate the chipset under a typical &quot;low end&quot; configuration. And it's doubtful the C/800 will remain more expensive than the P3/800 for very long.

HockeyPlayer,

<< Actually one of thing I found missing was the 2D image quality test. >>

Yes, some integrated video implementations suffer from extremely poor 2D image quality. However, in playing around with the KM133 Microstar K7TM Pro (6340) I have foud the image quality to be quite good. Running 1024x768x32x85x19&quot; seems at least as good as a typical nVidia card and almost as good as an ATi card.

ABOUT THE DVD ISSUE:

I can confirm that at 1024x768x32, the KM133 does seem a bit slower than other chipsets for 2D graphics. It's nothing that will affect your everyday work in the slightest, but it is noticeable when scrolling a complex browser window, for instance. And when I play a DivX movie using Windows Media Player -- The Killer Bean 2: The Party, a riot BTW -- everything is fine until I attempt full screen mode, where there is clearly a decrease in frame rate, and an absense of the &quot;smoothing&quot; that usually gets done to mask pixelation on smaller MPEGx streams.

However, as soon as I drop the resolution down to 1024x768x16, it looks flawless. I don't know if this is a driver issue with hardware acceleration not kicking in at certain color depths, or if the ProSavage 2D core is just slow at high resolutions with 32bit color depths. But at 16bpp, it's perfectly smooth even at full screen (and DivX takes a lot more decoding power than DVD/MPEG2).

Modus
 

theplanb

Golden Member
Jan 12, 2001
1,239
0
0
Have you tried the PC-CHIP's Sis730s board? If so, how do you think they fair with other boards?
 

Modus

Platinum Member
Oct 9, 1999
2,235
0
0
I will probably be building quite a few systems based on it in the near future. I'll let you guys know how it turns out. It's certainly cheap: I can sell them for $83 US, and that includes video, sound, 56k HSP softmodem, and SiS 10/100 PCI Ethernet. How can you beat that?

Modus
 

RalphTheCow

Senior member
Sep 14, 2000
942
366
136
I experienced the same problem as Hockeyplayer. At work they upgraded me from an integrated Matrox to an 810 with the integrated Intel video, and it kills my eyes. The monitor and refresh and resolution settings are the same. What can the difference be?
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |