AMD vs Intel Question

[Jester700]

So much nonsense regarding dual-core processors. I can't even keep up anymore. People really need to wake up and realize that until companies start programming more software to take advantage of SMP that having that second core is really just a marketing ploy. For the general user it may take some of the mundane OS tasks and leave them on one core and free up more of the clock from the other core, but that's about it. Not very many users really multi-task like such an SMP platform is meant to do. We're talking having a few design applications open and working between them as one or the other is rendering something out.

OK, but EVERY video encoding program on Anandtech's charts gained substantial speed going from a 1.8GHz single core to a 1.8GHz dual core Opteron. The smallest change was a 50% increase in frame rate for WME, and one of the benches gained almost 100%. These may have been selected to show off dual core CPUs, but if someone wants to keep a CPU for a couple years, dual core seems a valid choice at this point.

And to be really honest, for the Windows OS nothing, nothing is going to touch two Opteron 280s on a good platform. If we want to get really ******* serious about putting down some silly-short encoding benches then that's where to start. Anything else is laughably consumer. I even feel inferior to such a machine with both my Intel SMP rigs.

Well, yeah. Silly money buys silly performance. Always has, always will. The question is, where on the diminishing returns curve do you want to pitch your tent? Give it 2 years, and the then-current $300 quad core CPU will be matching that dream setup, and the consumer will be laughing indeed.

[lordsmurf]

I don't consider WME to be a test that means anything or is worth anything whatsoever. Try some MPEG-2 encoder tests, or at very least something more common and mainstream.

This all gets back to my out-of-touch with reality comments from earlier.

[deadrats]

I don't consider WME to be a test that means anything or is worth anything whatsoever. Try some MPEG-2 encoder tests, or at very least something more common and mainstream.

This all gets back to my out-of-touch with reality comments from earlier.

i would say that windows media encoder is one of the most relevant apps to test with for one big reason: it's the only app currently available for creating VC-1 compliant files. as you know (or should know) VC-1 along with mpeg-2 and H.264 is one of the 3 formats approved for next generation high definition files. it's also way more intensive, both in encoding and decoding, than mpeg-2.

my current setup can encode mpeg-2 at full D1 resolutions at 2-3 times real time depending on the app used and the bitrate, while i'm lucky if i hit 10 fps while encoding to wmv or avc.

[Jester700]

I don't consider WME to be a test that means anything or is worth anything whatsoever. Try some MPEG-2 encoder tests, or at very least something more common and mainstream.

This all gets back to my out-of-touch with reality comments from earlier.

I'd be glad to look at any tests you deem more relevant. But these are the only ones I've seen (I haven't really started looking for upgrade parts yet). Assuming no such MPEG2 tests are published, I can either assume the several tests done (more than just WME), which all came to similar conclusions, do not correlate with my own different (yet still video encoding) software and go on...(what again?), or I can say these tests point in a general direction that I should take into consideration. I don't mean to imply these are the be-all, end-all; I know they're not. I'm just trying to see why there are those still saying "Intel=video encoding" when the only tests I've seen recently don't agree. If there is better evidence that says differently, I'm all ears. er, eyes.

[Doramius]

But you just said you don't believe anything Intel puts on their site. What else won't you believe when there is physical and publicly available proof?

That's where bias can obviously be seen and regardless of a benchmark test you're always gonna argue something to your opinion deeming it as truth.

Again, IN REALITY, you can't say one is better than the other unless you state a specific task. Even at that, you have to compare, as close as possible, VERY similar chips. Also, as I said before, it's getting to the point where a major split will come into play and you won't even be able to compare the 2, because they won't even be doing the same thing. It's Like the APPLE/Macintosch vs PC/WINDOWS war. Honestly, they may run similarly, but they are completely 2 different things. I think Macs are top of the line Elite in some ares; And PCs are the best you can come across in other ares. Now if you state something like which one would you get because of software compatibility, I'd tell you flat out PC would be better, though I know I'd get Mac users to argue with me to no end.

[lordsmurf]

Benchmark tests also assume the person doing the testing knows what they're doing. I have to say many tests seem awful fishy. How many people tend to know all of those programs and can operate them correctly? Not many, and if they do, I can pretty much assure you they'll be doing more than playing the review game at a magazine/site.

[hech54]

Any opinions about which processor is better

Damn....why don't you post a Ford vs. Chevy question next....you'll get the same arguements.
MINE is better....no MINE is better....YOURS sucks!!!!

[lordsmurf]

Tastes great, less filling.

[hech54]

[Doramius]

Everyone knows Dodge would kick Ford and Chevy's A**es.

Again, it's not even comparing apples to oranges. It's comparing real fruit to glass fruit. It you get it for looks, they both will work. real fruit spoils, Glass doesn't. You can eat Real fruit, but it would ruin the display. Real fruit isn't as fragile as glass in case someone picks it up. Real fruit might not have the ideal shape people enjoy seeing. ETC. ETC.

Though they are closely related to each other, they just aren't identical enough to compare unless you want a specific task to be performed. Then you match out with performs better and make your decision. You just have to find one that matches what you use most.

[INFRATOM]

You said you already have P4, my suggestion is wait another year or so there is not much gain if you have a good P4 machine and also these double CPU's are new. Wait until there are enough 64 bit programs out there and windows 64 bit has enough drivers. Meanwhile if you can optimize and maybe overclock, cleanout your computer a bit, it'll feel like you got a new computer.

[Jester700]

But you just said you don't believe anything Intel puts on their site. What else won't you believe when there is physical and publicly available proof?

That's where bias can obviously be seen and regardless of a benchmark test you're always gonna argue something to your opinion deeming it as truth.

Again, IN REALITY, you can't say one is better than the other unless you state a specific task. Even at that, you have to compare, as close as possible, VERY similar chips.

Is it just me who thinks there's a difference between what's posted on an independent geek site vs. one of the two competitors? That's not bias, that's healthy skepticism - I wouldn't trust anything posted on AMD's site either. I figured that'd go without saying, but maybe not.

Anand may not have done a perfect battery of tests, but he's been trustable - in fact, his current "bang for buck" dual core pick is actually an intel chip - the cheapo 805.

[Doramius]

Hey don't think I'm going all Intel. AMD also has publicly available information showing their processors are easier to overclock and run on low power. The Geek sites do tend to be biased. I haven't found any that aren't. Though many will find something decent in the competition's processor, it's like someone who's Pro-Ford saying, "The OnStar system is the greatest." knowing full well it's a feature that GM is strongly incorporating into many of their vehicles.

With all the different Processors out there there needs to be more than a general benchmark.

Though this would be extremely difficult to do, It'd be great if there was a performance test chart at the websites of the software. Let's say you go to Blizzard.com and you pick a game and hit the performance test. You get a bargraph of your systems performance with the hardware you have installed in your computer. It records the information (anonymously) and matches it against all others who have taken the test. Then gives you mean median & max performance of each processor class and other hardware. While some processors will peak out high, it may have low areas.

SOmeone who wants to play the game and is looking to build a special machine for it can look at that and see what they need to upgrade, or just outright build the best machine from the data shown. THey'll choose and AMD 3000+ and a Asus MoBO with an ATI 9600 with only 512MB RAM, because those are the ones that may not be the highest, but seem to perform overall the best together.

Any site that did this would almost certainly benefit and boost the sales of much more software and certain processors.

[pchan]

Both AMD and Intel are charging too much for their chips. 80% (Intel) and 20% (AMD) market share is not ideal. A 50-50% market share will be great. Consumers will enjoy value for money products when they both slug it for market share. I gave up searching for the best system. Whatever best system now will be the 2nd best in 3 months. So, I cut the chase. Look for the system that meet your current and near term needs.

[Dayna]

I finally made a decision and over the weekend I got an AMD Athlon 64 X2 Dual Core Processor 4200+ 2.2GHz, 2000MHz FSB, 512 KB L2 cache per core with TV Tuner, Windows MCE 2005, 2GB PC 3200 DDR Memory, 250GB Hard Drive, and I also added in an ATI Radeon X1600 Pro 512MB Crossfire Edition Video Card and man I am so happy with it. I spent the weekend installing my softwares and photos in place, etc. Then I started the S10 installation. Never had any errors installing on here. I then installed the patch that is supposed to have came out for helping with the AMD PCs, and I restarted the PC and opened S10 and everything was perfect....none of my transitions were studdering in the preview window, no loud popping sounds, no problems whatsoever!! I rendered up a small sample and it rendered up just fine. Even when I did the Data Rate Test, this new PC was in the 64 Thousands whereas my P4 was only in the 50 thousands. I really like the ATI TV Tuner as I am able to capture through it and not have to use my AV/DV card. So I have a spot to put a Soundblaster Audigy when I get it. The Extra GB of memory I installed DID make a difference as it only came with 1GB and I noticed the difference as soon as I turned the PC on after installing the 2nd GB. I was going to get a Gateway Dual 3800+, but I got this one for $3 less that the other one...only because they came out with another model ofter this one and he gave me 15% off of it and I found it in a local Best Buy Store. It was the only store that had it. I wasn't going to turn it down as it was about the same price as the other one. I was really surprised at the difference I have seen so far in the Pinnacle Studio 10 though. Gotta go change my profile!

[jimmalenko]

https://forum.videohelp.com/viewtopic.php?t=298718

Deja vu ?

[Dayna]

Thats where I was asking about them the other day before I found out about the deal I could get...so I chose the 4200+...make sense?

[jimmalenko]

Errrr .... my bad.


Try this link:

https://forum.videohelp.com/viewtopic.php?t=299132

[Dayna]

Sorry, I read most of those posts, but still are a little confused on what you meant. There will always be opinions on the AMD vs Intel. I just got this one Saturday and so far its been great...I am really enjoying learning the MCE as this is the first time I have had it. Since I put the extra GB of memory in it I also noticed a difference. This is my first AMD as I have always used Intel, but I really like it so far. Got my 19" LCD in front of me and lots of new goodies on here to play with and I am happy.

[jimmalenko]

Nevermind. Was just intrigued by how the AMD Vs Intel war can rage on in two concurrent threads is all.

[Dayna]

LOL, I see, thats kinda what I thought you meant. Ok, thanks!

[SingSing]

This is not like Beta vs VHS. This is more like Pepsi vs Coke, most people already make up their minds, and are only going to choose the flavor from one of these CPU makers.

[SCDVD]

One if the things that goes on is that people do it backwards. They buy something and then look to find "proof" that their purchase was a good one. In practical application, the differences are slight IF it is the same class of MB and processor/memory. Since there are so many benchmark tests and even more ways to jury-rig these tests, it's easy to find your favorite test to "prove" that your purchase was a good one.

[Doramius]

This is not like Beta vs VHS. This is more like Pepsi vs Coke, most people already make up their minds, and are only going to choose the flavor from one of these CPU makers.

This is why I drink Sprite....No 7*UP....No Sprite....uh, Sierra Mist? Ok then Mountain Dew,....No Mello Yello,......is Vault anygood? :P

Again, wait for carbon processors to land in the retail market. That's where you're going to see a whole new war.

[jagabo]

Intel just made huge price cuts on the 900 series dual core processors. If you're about to buy one you should wait a few days/weeks for the price cuts to filter through the pipeline.

http://www.intel.com/intel/finance/pricelist/April_23_06_1ku_Price_web.pdf?iid=InvRel+...Pricelist_pdf&

950: $637 --> $316 (-50%)
940: $423 --> $241 (-43%)
930: $316 --> $209 (-34%)
920: $241 --> $209 (-13%)

Presumably, part of this is to clear out inventory for the introduction of the Core Duo desktop processors (in June?).

[bevills1]

There's little if any difference between Intel and AMD performance, and actual tests show Intel a little better on some apps and AMD a little better on other apps. I've built just a single Intel system and 4 or 5 AMD systems over the years and never had problems with any of them.

[Doramius]

What if we talk about laptops? WHo's got the better processor there? 8)

[rallynavvie]

What if we talk about laptops? WHo's got the better processor there? 8)

Werd, I love my Core Duo lapper. It allows me to carry over my SMP fetish into portables 8)

[Cobra]

Rallynavvie has a benchmark set up here at VideoHelp - why don't we all submit non-overclocked, totally-standard-settings results to that and see how it goes?

There's no point arguing about AMD vs Intel. It's the way they are designed, mainly the way they pipeline tasks, that defines how they will perform. Each one can be better or worse than the opposing design at a set task. If there was a way to get brilliant all-round performance, they'd be doing it by now.

As I say, I'm not biased either way. I'll go with what's best for me.

[jagabo]

Rallynavvie has a benchmark set up here at VideoHelp - why don't we all submit non-overclocked, totally-standard-settings results to that and see how it goes?

Unfortunately his benchmark doesn't account for a few settings in TMPGEnc that have a large effect on rendering time.