Media Encoding Performance
A traditional weak spot of the current AM2 processor offerings has been in the media encoding arena when compared to the Intel Core 2 Duo. Considering the main optimizations for the Core 2 Duo centered on overall media encoding performance, we were very interested in seeing how this platform compared to current AMD lineups. We are utilizing an updated video encoding test suite for this article that includes AnyDVD, Nero Recode 2, Windows Media Encoder 9, and QuickTime Pro 7.1.
Our first series of tests is quite easy - we take our original Office Space DVD and use AnyDVD and Nero Recode 2 to copy the full DVD to the hard drive without compression, thus providing an almost exact duplicate of the DVD. We then fired up Nero Recode 2, selected our Office Space copy on the hard drive, and performed a shrink operation to allow the entire movie along with extras to fit on a single 4.5GB DVD disc. We then go back and use Nero Recode to shrink the full DVD copy and burn it onto our DVD disc. We left all options on their defaults except we checked off the advanced analysis option. The scores reported include the full encoding process and are represented in minutes/seconds, with lower numbers indicating better performance.
The results are very interesting as we did not expect our AM2 systems to perform this well and to do so is very surprising when reviewing other media encoding results. In fact, we ran the test several times and verified our settings before accepting the test results. Perhaps the last few months of Core 2 Duo results constantly exceeding AM2's best by a wide margin have tainted our cognitive abilities.
Of course, we are using a mid-range AM2 processor against the budget C2D part (the AMD price cuts have helped matters there, as the price difference is currently only about $35) but we fully believe the majority of the performance difference lies in the chipset selection on the AM2 systems. It is only in the shrink and burn tests that we see the Intel platform flexing its muscles. The difference in the shrink portion of this test was minimal and mirrored our individual shrink test results. We noticed our Intel platform was able to feed data to our DVD drive in a consistent manner with each of the AM2 platforms suffering from stutters several times resulting in our drive slowing down during the burn process. It was difficult to figure out if this was a driver, disk controller, or CPU issue. We are still testing on other chipset and CPU combinations.
Our next test has us converting our day at the beach AVI file into a high definition WMV file suitable for our Aunt Gertrude to view on her new high definition LCD TV. We ensured our quality settings were set for High Definition output. The remaining options are left at default values and then we let this program do its magic. We are reporting the numbers in minutes/seconds to complete the conversion.
These results left our heads shaking and we busily scurried off to the far corners of the lab looking for answers. After several test repeats, a few new images, and a couple of drinks, we stand by the results. Our answer, which we came up with after we switched out the MSI G965 board for a Gigabyte G965 board and had the same results, is that our E6300 is just incapable of keeping up with our AM2 5200+ X2 in this test. Why? It's all about the cache and CPU speed. Our E6600 trounces the 5200+ X2 at stock speeds and scores near it when running at E6300 speeds thanks to the increased cache. Meanwhile, our E6400 is just slightly off the 5200+ X2 pace at a modest 300MHz increase in speed over the E6300.
Next on the list is our QuickTime Pro 7.1 test that will convert our newly downloaded .Mov file into a plasma screen pleasing H.264 format. We ensured our quality settings were set to their highest levels and then let the horses loose. The values reported are again in minutes/seconds for the conversion time, with lower numbers being better.
Once again, the AM2 platforms have a slight advantage with the 6150 scoring outside of the norm in this test - so far outside that we thought something was not kosher in the land of NVIDIA. Fortunately for us, the answer was easy once we reviewed the playback files. It seems as though the quality of our encoding results on the NVIDIA platform left much to be desired along with numerous frames missing during playback or other frames just tangled together at times. We ran this same test under XP without an issue so our first blip with Vista has now occurred. Our video images are presented below.
In these particular screen shots both the G965 and 690G image quality results are basically even, but the 6150 is showing a case of the jaggies at full screen resolution. The entire playback file suffered from this issue.
A traditional weak spot of the current AM2 processor offerings has been in the media encoding arena when compared to the Intel Core 2 Duo. Considering the main optimizations for the Core 2 Duo centered on overall media encoding performance, we were very interested in seeing how this platform compared to current AMD lineups. We are utilizing an updated video encoding test suite for this article that includes AnyDVD, Nero Recode 2, Windows Media Encoder 9, and QuickTime Pro 7.1.
Our first series of tests is quite easy - we take our original Office Space DVD and use AnyDVD and Nero Recode 2 to copy the full DVD to the hard drive without compression, thus providing an almost exact duplicate of the DVD. We then fired up Nero Recode 2, selected our Office Space copy on the hard drive, and performed a shrink operation to allow the entire movie along with extras to fit on a single 4.5GB DVD disc. We then go back and use Nero Recode to shrink the full DVD copy and burn it onto our DVD disc. We left all options on their defaults except we checked off the advanced analysis option. The scores reported include the full encoding process and are represented in minutes/seconds, with lower numbers indicating better performance.
The results are very interesting as we did not expect our AM2 systems to perform this well and to do so is very surprising when reviewing other media encoding results. In fact, we ran the test several times and verified our settings before accepting the test results. Perhaps the last few months of Core 2 Duo results constantly exceeding AM2's best by a wide margin have tainted our cognitive abilities.
Of course, we are using a mid-range AM2 processor against the budget C2D part (the AMD price cuts have helped matters there, as the price difference is currently only about $35) but we fully believe the majority of the performance difference lies in the chipset selection on the AM2 systems. It is only in the shrink and burn tests that we see the Intel platform flexing its muscles. The difference in the shrink portion of this test was minimal and mirrored our individual shrink test results. We noticed our Intel platform was able to feed data to our DVD drive in a consistent manner with each of the AM2 platforms suffering from stutters several times resulting in our drive slowing down during the burn process. It was difficult to figure out if this was a driver, disk controller, or CPU issue. We are still testing on other chipset and CPU combinations.
Our next test has us converting our day at the beach AVI file into a high definition WMV file suitable for our Aunt Gertrude to view on her new high definition LCD TV. We ensured our quality settings were set for High Definition output. The remaining options are left at default values and then we let this program do its magic. We are reporting the numbers in minutes/seconds to complete the conversion.
These results left our heads shaking and we busily scurried off to the far corners of the lab looking for answers. After several test repeats, a few new images, and a couple of drinks, we stand by the results. Our answer, which we came up with after we switched out the MSI G965 board for a Gigabyte G965 board and had the same results, is that our E6300 is just incapable of keeping up with our AM2 5200+ X2 in this test. Why? It's all about the cache and CPU speed. Our E6600 trounces the 5200+ X2 at stock speeds and scores near it when running at E6300 speeds thanks to the increased cache. Meanwhile, our E6400 is just slightly off the 5200+ X2 pace at a modest 300MHz increase in speed over the E6300.
Next on the list is our QuickTime Pro 7.1 test that will convert our newly downloaded .Mov file into a plasma screen pleasing H.264 format. We ensured our quality settings were set to their highest levels and then let the horses loose. The values reported are again in minutes/seconds for the conversion time, with lower numbers being better.
Once again, the AM2 platforms have a slight advantage with the 6150 scoring outside of the norm in this test - so far outside that we thought something was not kosher in the land of NVIDIA. Fortunately for us, the answer was easy once we reviewed the playback files. It seems as though the quality of our encoding results on the NVIDIA platform left much to be desired along with numerous frames missing during playback or other frames just tangled together at times. We ran this same test under XP without an issue so our first blip with Vista has now occurred. Our video images are presented below.
 690G - Click to enlarge |
 G965 - Click to enlarge |
 6150 - Click to enlarge |
In these particular screen shots both the G965 and 690G image quality results are basically even, but the 6150 is showing a case of the jaggies at full screen resolution. The entire playback file suffered from this issue.
Facebook
Twitter
RSS
70 Comments
View All Comments
TA152H - Tuesday, March 6, 2007 - link
I just read the first page of this review, and wonder if a 15 year old kid wrote this.Let's grow up a little and stop thinking the world revolves around people that play video games all day and whine about big companies that actually think people use computers for real work!
There's a perfectly good reason that Intel, et al, release IGPs that don't have great 3D processing power. I mean think before you write something so inane. The world doesn't all play games, and not everyone is an idiot and wants to play a mindless 3D shoot 'em up game and thinks extra resolution makes it so much more fun. Pac-Man, Space Invaders and Defender were incredibly crude by today's standards, but people loved them. It's the idea, not the resolution.
Also keep in mind it's not like there is no downside to creating IGPs with all this functionality. You burn up electricity with those extra transistors, even when the vast majority of people, the vast majority of the time, aren't using them. You also generate more heat, and you make them more expensive. Why should someone that doesn't play 3D games pay for this when they don't use it? Why should be use more electricity so some simpletons that want to zap aliens can have that ability in an IGP? Isn't it enough they make cards specifically for that? Shouldn't only the people that want this capability have to pay for it? Why should other machines be crippled with this heat generating, electricity using capability when they never use it?
So, if you're still incredulous about why companies still make relatively slow IGPs, it's because they fill the need for the vast majority of the people that buy them. And if they do not, there are options available that do. That's a perfect solution, you get it if you need it, and don't if you don't. Why whine about it so much?
Gary Key - Tuesday, March 6, 2007 - link
I wish I was 15 again. :)
I fully understand the world does not revolve gaming and that most people use computers for work. However, these same people might have kids who want to play games or they decide to try it also and without knowledge of what it takes to run a game properly they pick up the phone and call Dell or HP or go on-line. Unfortunately, the majority of the machines they will see and probably can afford are going to be IGP solutions. The reps or the machine specs on-line will lead you to believe the $650 unit you just picked out will cover all of your needs and that is wrong. One of Dell's top selling units is the 521 series and is advertised as "Built for Smooth, Advanced Performance, Edit your videos and photos, play games and more." Guess what, it just barely does most that and we hear people complain all the time that their PC is not fast enough for this application or this game, or a myriad of other uses.
Probably the same reason we whined about Yugos and Pintos and the reason you no longer see them in production. Sure, they filled a basic need but once you got in a Civic or a Corolla you fully realized that something faster, better, and of higher quality was available for nearly the same price. The technology is there to greatly improve IGP performance at very little cost. We do not expect these solutions to ever provide the same level of performance as dedicated GPU cards but we do expect more out of them. Intel promised the world they would have a class leading DX9 IGP solution in the G965, what we ended up with was a solution that barely worked better than the 945G. If you are going to advertise functionality that leads one to believe you can play the latest games on it and mass market it, then expect us to report our findings and to call them out on it if it does not meet the marketing specifications. The same holds true for NVIDIA, AMD, SIS, or VIA in this market sector. We have the same issue with the low end video cards also, in some cases their performance is no better than the IGP solutions which means you just spent a $100 for the same performance yet those words all over the product box would lead you to believe otherwise.
Thank you for your comments! :)
TA152H - Wednesday, March 7, 2007 - link
Gary,My big problem with what you said in the article is that you are blaming Intel, and to a lesser extent, for creating a great chipset for the vast majority of people that use it. The negative remarks about how they are making it difficult for the those poor developers is completely misplaced, they make an excellent product that does what it should. In your response, you are now changing your position, I believe. It seems that Dell and HP are misrepresenting products they are selling, and that is the root problem. I am not privvy to that, but if that is your contention, I have no problem with it. It could be correct, but pointing fingers at Intel and other IGP makers is misplaced. They make a product for a huge market and address it well, and if some maker misrepresents it, that's their issue and not the maker. That would be like complaining that a Ford Mustang isn't as nimble as a Lotus Elise, because sometimes dealers say it's the best handling car. Is it Ford's fault that it is being sold incorrectly?
Your remarks about the Pinto and Yugo are totally misplaced. The Pinto died because it had a nasty tendency to explode when hit in the rear, and the Yugo was unreliable to the extreme. Neither was particularly well suited for its target audience, but the class of cars itself was. You kind of make my point for me. A lot of people still like subcompact cars for exactly the reason I stated - they don't need a big monster and they don't want to pay for the gas or the car in the first place. They get what they need that does the job adequately.
It's kind of childish, and megalomanical to think the whole world should have to pay more for their computers, and burn more power doing it, so a small subset of people can benefit because game developers can develop to a higher standard. So the business world, and everyone that doesn't play 3D games should subsidize gameplayers by having to buy something useless that uses extra power, probably makes more noise and costs more. All so gameplayers can get developers to design to something better. Sorry, I don't agree, and apparently neither does any of the big companies that actually make these decisions.
With regards to operating systems, a lot of companies still use Windows 2000. The business world is a lot more pragmatic and don't see a need to move to something because Microsoft came out with it. Outside of new machines coming with XP instead of 2000, was there really a compelling reason to move to it? Not for most people.
This last remark is directed at whoever it was that said that the Intel chipset took more power and was slower. Well, think a bit, it's a fully functional chipset and should, whereas the memory controller is not in those for the AMD processors. Also, Intel typically makes their chipsets on old lithography, not the newest stuff, although I don't know anything specific with regards to these chipsets.
Gary Key - Wednesday, March 7, 2007 - link
Our issue continues to be the technical and marketing spins of the various companies that sell these solutions as if they will do everything a consumer expects out of a PC. I am not addressing the office crowd and was clear about that in the article. I am addressing the home user who buys a machine and has certain expectations out of it based upon the sales literature or advertisements. I deal with the OEMs on an almost daily basis and while they have certain cost targets to achieve, one of their primary issues is that they have to constantly up sell a unit if the consumer says the word "gaming". Generally, this up sell does not work and they lose the sell as the consumer thinks they are getting the bait and switch routine. One of the top listed support call questions or concerns is, "My computer will not run this game or the game does not run properly." Like it or not, people do use the Home PC to play games or enjoy multimedia applications. The last time I looked at the numbers the PC game market is still very viable and alive. I also realize that not everyone plays games but that does not mean the requirements for decent 3D performance is not required, especially with Vista being launched now. This changes everything for the consumer market as it does not matter if Windows 2000 or XP is still best option for most people as the consumer buying these machines will receive Vista.
I am not advocating that the IGP solutions run the latest games at 100FPS. I am advocating that the base level of performance should be better, at least to the point where 1280x1024 at 30FPS with decent settings is viable, not because I want it but a significant amount of consumers would like an inexpensive PC solution that the kids can also use in playing games or maybe they are buying a second system for the family and do not want to spend as much for a discreet video card as they do for the system. Settling for 640x480 graphics is not the answer, improving the performance of the solution is the answer. We constantly expect our machines (yes, even the office crowds) to perform faster, whether we admit it or not. I have yet to come across a coworker that was satisfied with their current hardware once they used a next generation system. So, I do not think it is wrong to expect an improvement in the advance of IGP solutions targeted to the home crowd as an all-in-one solution.
As for the marketing spins, this is one of my main issues with the G965, and why I have a real issue with it when I address the gaming issues. This is straight from their webiste -
"Gamers and Media enthusiasts have been long demanding better technology for unparallel ultra-realistic visual experience and the demand will be even greater in the future. Intel has answered the current demand to support future technologies by unveiling its hybrid graphics architecture with the introduction of Intel® Graphics Media Accelerator 3000 (Intel® GMA 3000) family engine. This state-of-the-art hybrid architecture evolved as a balance between fixed function and programmability. The graphics engine of the Intel® Graphics Media Accelerator 3000 family consists of a scalable array of symmetric processing components known as execution units (EUs) which can be programmed to dynamically process graphics or media data alike. Programmability of the EUs adds flexibility and enables support for future graphics and media usages by upgrading the driver. Execution units support dynamic load balancing, multi-threading, and multi-functional data processing, resulting in increased performance to enable a more compelling gaming and visual experience for main stream users.
Intel’s next generation architecture shows Intel’s continued innovation by delivering greater flexibility and performance to meet the needs for the current and future consumer and corporate applications.
The new Intel Graphics Media Accelerator architecture with its greater performance and flexibility will be the basis for many new consumer and corporate applications. For consumers, the Intel® Graphics Media Accelerator X3000 engine supported on Intel® G965 Express chipset has been optimized for enhanced 3D to allow for greater game compatibility and realism with new video and display features to deliver a theater-like entertainment experience through Intel® Clear Video Technology. For corporate users, Intel® Graphics Media Accelerator 3000 engine incorporated in Intel® Q965/Q963 Express chipsets continues to offer stability, ease of use, and lower power consumption, in a cost effective solution. The Intel® Graphics Media Accelerator 3000 family engine is eligible for the Microsoft Windows Vista*Premium logo.
Intel’s next generation Graphics Media Accelerator that provides new levels of graphics and video responsiveness in a cost effective solution. The Intel graphics engine is integrated into the Graphics Memory Controller Hub (GMCH) and enables a low power, reduced form factor solution compared to power-hungry discrete graphics cards.
With Intel’s next-generation Graphics Media Accelerator, Intel continues to drive innovative platform enhancements that increase the overall performance for the end user. This next-generation Intel graphics architecture is designed to be extensible and offers extraordinary features. This section discusses the benefits of the Intel® Graphics Media Accelerator 3000 architecture features and innovations: programmability, dynamic load balancing, multi-threading, multi-function, 32 bit full precision compute, and dynamic and static flow control which enable more stunning graphics in games and video applications."
Notice they lead off their marketing spin with the word "Gamer" and continue with a demanding better technology statement. I think that pretty much indicates the market they were addressing with this chipset when the X3000 core was being developed. Guess what, the words that Intel used to launch this chipset are the same ones I am advocating in my opinion.
Thanks for your comments, they are certainly welcome and we respect your opinion. :)
TA152H - Thursday, March 8, 2007 - link
Gary,You mention gamer, and for the games I play this chipset works fine for almost all of them. Did you notice the word "mainstream"? It's not for enthusiast gamers, that want the most modern games, and the highest settings. But, still, I agree they exaggerate. I think the crux of our disagreement is, you seem to think that an IGP should run very demanding, very modern games at fairly high settings well. I think that's entirely unrealistic. I think that's what add-in cards are for. These are not for the enthusiast, and I don't think it's feasible to expect so much of them. Anything that could would be very power consuming, expensive, and big. They are already putting fans on some chipsets! This to me is entirely unacceptable, I like my PC very quiet. If you like performance and fans and a hot room, that's what cards are for.
I guess we mainly disagree on the target market for these items. You expect them to be serious gaming platforms, I don't. There are so many tiers of add-on cards, I don't see that being proper segmentation, and people don't want to pay for it. Put another way, if you make these more powerful than entry level add-in cards, what do people buy that want basic 3D performance without needing anything too powerful? They have to spend more money, and get a hotter power hungry machine because there is nothing that works well for them? Whereas, the way it is now, if they want something better, they go with a 965 and get a add-in card. Not only that, I think these things fit a lot of people pretty well. There is almost nothing I run that requires more than a Radeon 9000, and I'm not nearly as unique as I might like to think I am. There a big market for this level of performance.
Now, do computer companies exaggerate their claims. Well, as I said we agree on that part. They always have, always will. It's capitalism, so I guess we have to accept it. Capitalism isn't perfect, but overall it works.
I agree Vista will necessarily raise the bar a bit for 3D performance, and in fact Intel did improve their IGP to address that. We only disagree on the amount, not that it was done at all. But, even with Vista Microsoft hedged their bets and did not make the requirements have any teeth. You don't need "Aero" after all, and in some version you don't even get it. Don't they deserve some blame too? How difficult would it have been for them to make it a requirement? Think about it, if they don't sell Vista, they are still going to sell XP, and the upgrade market has never been that big compared to the amount they sell with new PCs. So, why no egg on their face?
I disagree completely with 1280x1024 being necessary, or that parents want their kids playing brainless games all the time. Playing at 640 x 480 is fine, or 800 x 600, and just saying it isn't without any backing isn't very useful. Do you remember Pac-Man, or the Atari 2600? Or Galaga? Or Space Invaders? They were much more popular than any single game now, and they were a lot of fun (Defender was my favorite). They were incredibly crude by today's standard, but people still loved them. Some poor little kid that has to play at 800 x 600 is just not going to get my sympathy, if the game is a fun game, it just won't matter. If you think the level of detail of an explosion is what makes the game fun, then I think you've lost touch with mainstream human behavior and a bit obsessive about this level of detail. It's not what matters, even if you have a bunch of obsessive people think it is. If it were, games played at 256x192 or whatever, with 4 colors, that were two dimensional, that had big time lag when a lot was going on, would never have been popular. For those that are extreme though, there are cards.
I would agree with you if IGPs were making NO progress. But they are, and they will continue to. You know what would be interesting for you guys to do (at least to me)? A comparison between IGPs and older video cards, and do a trend where you compare them for the last several years to see how far IGPs have been behind cards, and see what the trend is with that. It would be imprecise, and in some cases difficult, but I think the results would be illuminating for just about all of us. I'm kind of curious if they are further behind now than they were say in the 810 days, or if they're trailing by the same amount of time. My guess, and it's just a guess, is it's about the same lag. Wouldn't you be interested in that too?
JarredWalton - Wednesday, March 7, 2007 - link
If you want the best example of misrepresenting the capabilities of a product, look at the 3DMark06 results and then look at the gaming results. The Intel GPU that you're so fond of happens to score the best in 3DMark (nearly twice as fast!), and yet it almost entirely fails to run two popular games/game engines. That's Intel's fault, and it's Intel's drivers that appear to be mostly to blame. How many places will post the 3DMark score as representing a product, because it's "easy to compare", and completely neglect to show that actual performance is very poor?For all we know, the G965 has the power to actually outperform the other IGPs and it's only the drivers holding it back. Unlikely, but possible. And while you appear to be more than happy with the IGP offerings, we would like more, and we would like more quality - in drivers, performance, features, etc.
You complain about us putting out our opinion on what an IGP should be, but basically opinion is all you give in response. Plenty of people will wholeheartedly agree with our stance; others will agree with yours. That graphics are becoming more necessary in modern computers is basically a given - though certainly a lot of people and businesses still run Windows 98 and 2000, and they'll continue to run XP likely until after Vista's successor ships. Me, I'll take XP over 2000 any day - if for nothing else than the task manager than now has a networking tab. Yup, call me crazy, but I appreciate most of the new features that went into XP, and I'm sure I'll eventually appreciate Vista as well. (SuperFetch alone is going a LONG way towards winning me over!)
G965 is hardly better than 945G, which was barely better than 915G. X1250 likewise is only an incremental step up from X1100. Is there a compelling reason to upgrade operating systems? Sometimes. Is there a compelling reason to not upgrade? Sometimes. What about IGPs - what's the compelling reason to upgrade from 865G if all you need is 2D acceleration?
I guess we should all just forget DX8, 9, 10... probably ditch DX7 as well, forget 3D and move back into the glory days of pure 2D graphics acceleration. For all the people that would like that move, there are lots of others (many more, I'd wager) that would feel it is castrating innovation. Long term, we need new software and hardware and advances in both. How many people switched to OSX because it was an overall better OS for their needs? Only a few percent, but if Apple continued to evolve their OS and nothing happened with Windows, eventually Windows would disappear.
Q965 and 690V are more stripped down versions of the IGPs in question, and those are targeted at the business market that you seem to prefer. We're not looking at those; we're looking at the "consumer" line of IGPs, and we're finding them severely lacking. If there were no real difference in performance between the business and consumer offerings, then there would be no point in having different products. There is a difference, however, indicating that Intel and AMD are both aware of the need for better performance in the home market. Right now, that better performance is largely missing. What it comes down to is that we feel Intel, AMD, and others could do a lot better. For you, that might not be important, but right now a lot of IGPs are only good for a bare level of functionality. Fine for business, but not for a lot of home users; they might as well forget the home market and just focus on business IGPs if the offerings are going to continue to unimpress.
Finally, your comment about power is once again incorrect. Intel Core 2 Duo with GPU x uses less power than AMD Athlon X2 with GPU x, in the majority of situations. Intel doesn't have an IMC in that scenario either. Both X1250 and G965 are "fully functional chipsets". The AMD platform ends up using quite a bit less power. Is it because Intel uses old lithography? Probably, but again, that just reillustrates the point that Intel is not doing anything to move the IGP market forward. If we could run X2 processors with a G965 chipset, we would find as usual that Intel uses less power for the CPU. The net result is that the better platform for low power computing currently looks to favor AMD, at least in the IGP market.
TA152H - Wednesday, March 7, 2007 - link
Jared,You're changing the argument. I didn't say Intel didn't exaggerate claims, I'm saying the world doesn't need 3D graphics with superior performance for most people, like the supposition in the article said. I don't want it, and if you do, go buy it and stop insisting that everyone gets it too. You have options.
You may want more performance, etc..., and there are cards for you. Go buy them instead of complaining about an item that wasn't made for the enthusiast. It's like complaining a Chevrolet Cavalier isn't fast enough. So go buy a Corvette, and leave the gas mileage people the Cavalier. Stop insisting that we need it.
I don't complain about you giving an opinion, I complain about your opinion. You think that everyone should subsidize game players. I don't. I never said XP wasn't for anyone, just most people don't care one way or another. I prefer 2000, it's faster and I can run it on machines that are less powerful. I still think OS/2 is better than either, but that's another topic :P. Vista might be good eventually, but I don't feel compelled to buy it because it doesn't do anything I need. When it runs something I want to run, and Win 2K doesn't, I'll probably buy it. Until then, I saved a lot of money skipping XP.
"Hardly" is too ambiguous a term, and what you consider hardly could be considered considerable. What is obvious is they run fine for most of the people that need them, and they are not for enthusiasts that need more than they are targetted for. Intel is in the business of making money, and their graphics sell better than anyone else's. So, they're obviously doing something right, even though you can shoot aliens with them as well as you'd like. But then, that's what Nvidia and AMD/ATI are for. That's what the video slots are for.
Innovation for the sake of innovation is always a mistake, and if you don't believe that look at the P7 and you'll see a perfect example. And exaggerating something to make a point shows just how weak your point really is. Intel will continue to boost performance when it makes sense to, meaning it doesn't add much in terms of power requirements and cost, and adds something useful for the target of the product.
Your supposition on Apple is without support. Unless you think you know more than analysts that cover Apple and think they gained a lot of mindshare because of the Ipod. Also, moving to Intel helps, since it is better than the POWER platform was for them.
With regards to home users, you sort of have a point, because they do sell cheaper versions. But, you seem to think the world is filled with kids that play games, and simple ones that don't want to play games that require a lot of thinking. A lot of people, myself included, like playing strategy games, and some are fine playing games without all the bells and whistles running. I don't mind playing at 640 x 480 even for arcade games, because it doesn't change the game ideas and that's what is engaging. The funny part is you guys seem to think the world is filled with people that play extreme games because that is a lot of what reads your web site, but it isn't. Most people are happy to surf the internet, email, play card games online, and do other things that don't stress 3D capabilities much, and would prefer to save the money and electricity, and noise, and heat, and not get something they don't need or want. If they need or want more, there is more out there. Put in a different way, you say they should be more powerful because they have weaker siblings, has it occurred to you there are several lines of cards above them, and by contrast they don't have to be so powerful because of this. They are home machines, not enthusiast machines. That's what add-in cards are for.
You are misrepresenting my comments when you call them incorrect. You said that the highest performing IGP used the least power. I mentioned the chipset, not both pieces, and you would expect, everything else being equal, the Intel would use more, because it has a memory controller as part of the chipset.
You're dead wrong about Intel not moving the IGP market forward, and you even contradict yourself earlier in your remarks. It's moving it along slowly by your own admission. I'm not sure I agree on the speed, but we both agree that it is moving along. The fact is, Intel chipsets with graphics sell really well, so they are meeting the market demands. Maybe not yours, but, again, that's what add-in cards are for.
JarredWalton - Wednesday, March 7, 2007 - link
I have to say that it appears to be the same thing. Gary has further addressed the situation below. You are certainly entitled to your opinion on things, but likewise we are entitled to ours. For non-gaming, non-3D work (and to a lesser extent, non-video tasks) just about any IGP released since 2000 will perform acceptably. If that's all you need for a PC, you're set. Clearly, Intel *doesn't* feel that's all that's necessary, as they have integrated a lot more into their IGP and their marketing touts those new features. However, while we have more, it is in many instances inadequate.
If someone is happy playing older games at 640x480, that will almost certainly lead to them wanting to play newer games. If they only want to play strategy games, that will inevitably lead to looking at newer strategy games (Civ IV for example) that require 3D graphics. I think we can make a pretty good case that for anyone that does any gaming beyond solitaire and web-based games, the move to 3D quickly becomes necessary unless you simply want to relieve the glory days of 2000 and earlier.
There is a market for add-in cards, but that market exists in part because the IGP offerings are so anemic in terms of performance, and there's plenty of obfuscation there as well. Many OEM budget PCs don't even give the buyer the option of installing a truly capable GPU (at least GeForce 7600 or Radeon X1650) in the online configuration tools. The best IGP on the market currently rates about the performance of an X300/9600. Given that we have discrete GPUs ranging from entry level $50 cards up through uber-powerful $600 monsters, we'd like to see a few more options for IGP that get above the "not even entry-level performance". Both AMD and Intel have two IGP chipset levels, but they are simply "too slow" and "even slower - but it's for business so that's okay" (at least when it comes to 3D). If they're going to the trouble of making a faster model, why not actually make it useful? And why all the bogus marketing if it doesn't matter?
I too have seen far too many support questions wondering why a "recent" PC can't run a game some child got for Christmas or whatever. The companies selling and producing these products aren't out to educate the people about what's really required for certain computing tasks, so that's what we do. Most people apparently get that, and we're not the only ones complaining about the state of IGP. I've read about it in quite a few magazines, but I'm sure you disagree with their opinion as well.
For better or worse, IGP still goes into 40% of PCs sold. No wonder people get frustrated and purchase gaming consoles... not because a PC can't handle games with the correct configuration, but because the base models that so many buy are crippled at the starting gate. Don't even get me started about OEM systems that don't include GPU expansion slots (which are thankfully starting to become less common).
I understand your argument, and I respectfully disagree. Intel will sell millions of their IGP whether they suck rocks or they actually perform okay. They will sell millions of CPUs whether they are worse than the competition (P4 vs. Athlon 64) or better than the competition (Core 2 vs. Athlon X2). They are taking the easy way out on the IGP and millions of people unknowingly get "Powerful multimedia and gaming capabilities" that simply aren't. Intel is capable of doing better on IGPs - lower power and better performance, all in the same package. They don't do that for a variety of reasons, but what's good for a big corporation is rarely ideal for consumers. I for one care more about what will benefit consumers than what will make Intel, IBM, AMD, Dell, etc. happy. You don't work for one of them, by chance? You're awfully upset about a few minor paragraphs.
TA152H - Thursday, March 8, 2007 - link
Jared,I think we disagree on the purpose of this chipset. I have never liked shared memory video, ever since the PC jr. I thought it was a bad idea, and I still do. They are a low end segment and people should not expect performance from them. My expectations for them are not that high, and I think they're perfect for their market segment. I don't think IGPs should be particularly powerful or play the most modern games well, that's what add-in cards are for. It probably comes down to market segmentation.
You're doing the same thing Gary is, you are blaming Intel, et al, for what OEMs are doing with their computers. Intel has improved the video capability, and no doubt will continue to, and it will inevitably lag behind video cards. But then, why shouldn't it? They make the product, they aren't like IBM used to be and sold the end product to consumers too, so I'd say the makers have to fix that part and segment it well. Then again, who is really to blame? You can't blame makers, you have to put the blame squarely where it belongs, the consumer. If the consumer demands better stuff, they will get it. I totally disagree with you on items sucking "rocks". For the first time I can remember, the technology actually has a huge impact on what sells. Look at the market share losses Intel suffered because of the miserable Prescott. Look at the the low prices AMD is forced to sell their pathetic K8s for now that the P8 is out. I don't ever remember this happening before, if it had would x86 have won out (68K was much better)? It didn't happen completely, but it DID happen to a degree that had a material impact on both companies, and certainly helped Intel create a more prosaic and effective product in the Core 2. This is a company that is known for bizarre microprocessors too, from the 432 to the 860 to the Itanium (really more HP, but still), so a normal microprocessor that just works well probably killed them to make :P.
We have a fundamental disagreement on the segment, I guess that's it. I think they should leave off where the add-in cards start. Cards having their own memory and processor should be better than a integrated chipset with shared memory and shared die space with other functions. Call me a purist, but this makes more sense than this being faster than a card made specifically for that. Also consider that the reason why these machines sell so well is because of the cost. Intel, et al, has to balance the cost with the features, including other aspects like buffers, sound, etc... and make it all in the best package for their intended audience. Since you can't upgrade the chipset on the motherboard, and you can upgrade the sound and video, I think they do a pretty good job.
You probably have never played Civ 4, or other strategy games like it, if you make that remark. They do not need even anything remotely powerful to play well. I run it fine on a Radeon 9000, although the processor speed is extremely important for it. Civ III was a beast too in terms of CPU performance.
I'm sure Intel has their reasons for making the IGP the way they have, and you certainly do not know better than they what they are capable of. Could they make it lower power and more powerful? Well, it's possible, but then it would cost more. Would it target the market better? Probably not, otherwise they'd probably do it. But, of course, they are working on the next chipset that will, so it's not all bad.
The root problem with stuff like this is the consumer, and the consumer has chosen these chipsets, either actively or passively. If they rejected it like they rejected, for example, RD-RAM, or to a lesser extent the Prescott, then Intel wouldn't keep making them and would come out with a product that fit the market better. But these products do fit the market. You talk about how many people that complain about this or that, but do you ever think about how many people that buy these units are not complaining and are perfectly happy and saved some money because it was all they needed? Anandtech readers are a small, very inaccurate subset of the computing world at large, and you can't base a whole lot off of what readers here say. Computers 20 years ago were mainly for hobbyists, but now the general public uses them and the reality is, there is not an industry in the world where there isn't some mismatching. You think no one complains about cars? Dishwashers? Speakers? Call your local Sears to find out if they ever get complaints :P. It's the nature of the beast, and it won't go away if Intel makes ultra-powerful IGPs (which again, I find oxymoronic). There's always a tradeoff.
Final Hamlet - Tuesday, March 6, 2007 - link
I don't agree to _how_ this criticism was stated, but I _do_ agree to its content.I am PC worker, too.
I am not getting paid for playing directX 10 games.
I have a PC for using Word, Excel, Access, maybe a browser and a PDF reader. That is basically it. Now you tell me: Do I need a Core 2 Duo 6800 and a Geforce 8800GTX for that?
Nope. I need no graphics power at all (1280x1024x32 2D desktop - that's all) and my no. 1 priority for a processor is one that is _silent_ and nothing else.
I am quite sure I am not alone out there...