Dear forum members,
We are delighted to inform you that our forums are back online. Every single topic and post are now on their places and everything works just as before, only better. Welcome back!
Discussion on Article:
IDF Spring 2006 Coverage: Day 1
Would you say this when the nvidia 7800GTX 512 came onto the scene? Would you say ATi would become a runt brand? No, because ATi struck back and released the X1900XTX which toppled it.
When Intel introduced the P4, they were countered by the AMD Athlon 64, superior in ALL respects. When a cheap $150 3000+ can beat a P4-EE that costs 9x as much, you know they've met their match. You know the P4 met its match when it lost market share rapidly to the A64 and the entusiast community adopted and embraced socket 939. Go to a LAN, and you'll see most people using AMD's. Hell, 62 out of 64 in our last one here in Seattle were AMD systems.
Intel responded by taking AMD's strategy of low-power, low-TDP, and more performance per clock cycle. A value brand? Their new AM2 socket is coming out along with Socket F in June, when Intel releases their news processors as well. You think that AMD will keep socket 939 to compete with Intel's next-gen? No, they're working on their next-generation line as well. If anything, it's AMD that set the pace by kicking the P4 in the arse and forcing Intel to adopt THEIR style of architecture.
I'm sorry, but I can't stand such ignorant statements. It's a game of tug-of-war for these companies. AMD is now a competitor to Intel and will stay that way for a while. A value brand? A runt? I think you're a total crackhead!
You really don't understand much about processors, do you? Comparing graphics processors to CPUs shows a lack of knowledge, since they have shorter generational cycles.
No one said the P4 was any good, although the Presler is competitive with the A64 due to finer lithography being used.
Do you do ANY research before spouting your stupidity online? Look at the benchmarks for the AM2 out there currently. They are slower than the existing product. The released product should be better, but keep in mind it is coming out in a few months. The published benchmarks for the Conroe at 2.66 GHz show it ass-raping the Athlon 64 running at 2.8 GHz. Plus, the Conroe will be out in an EE version at 3.3 GHz. This is also a pre-release product so should so a lot of improvement between now and release. On top of this, it uses less power.
I knew the Conroe would be better than the Athlon 64, even though a lot of dolts here kept insisting it wouldn't be. Now we have results in, and it is the biggest disparity that has existed since the K6/Pentium III days. It is an enormous blowout, the Athlon 64 is so outclassed it will take a miracle for AMD to sell this degenerate product against the magnificent Conroe.
Again, keep in mind, the Conroe is still pre-release and should improve far more than the Athlon 64. It also uses less power. Intel's manufacturing is much better. They compared a slow 2.66 GHz version against AMD's best, not the 3.3 GHz EE processor.
1 There is no 3.3GHz EE (Emergency Edition) processor, it’s at 3.0GHz.
2 The top is 2.66GHz, there will be however lower speed versions.
3 It does not consume less power, where did you see those numbers, the only numbers I have seen is on server, Intel does 5% lower, just 5% lower. AMD .90 and DDR1 intel .65 and DDR2.
4 Intel bench must be TRUE and not like Apple ones where is all much faster and then when some one does real things discover they are fake (most of the times). I think they are true or Intel would be in big trouble by making false claims (with this they hit the dirt and don’t get up again).
5 This is not yet a released product. So for at least 6 months AMD is still in the lead with an “old” processor and manufacturing process.
6 Where are the synthetic benchmarks? Is Intel losing those? Or no differences between current platforms?
7 How can Intel win games benchmarks by such a difference when most of the times games are GPU limited?
8 AMD current AM2 processors are cut in half in the bandwidth, that explains the lower performance. AMD when give the K8 for partners where locked at 800Mhz, while the K7 was already at 1.6Ghz.
9 Intel likes to show off a lot, let’s hope they are not fooling anyone this time,
I must say I’m impressed with the performance but really disappointed with the power consuming numbers less than 40% of the P4 that means equal to current old AMD systems…
I have come to expect more from you than a post like that. You sound like an AMD gayblade, which is normally beneath you.
1. The EE is widely rumored to be available at 3.3 GHz. They already have 3 GHz Woodcrest in silicon, and in the months until the release, getting to 3.3 GHz should be very easy. But, we'll have to wait and see on this one. Both are possibilities.
2. You are dead wrong about 2.66 GHz, there will be 2.8 GHz version available that is not EE according to information on the web. More importantly, they overclocked the Athlon to 2.8 GHz, which is actually doing AMD a favor.
3. Intel has been saying all along that their processors are lower power than the Athlon 64, and they keep saying it. The actual numbers might not be huge. DDR2 has nothing to do with the CPU power usage. SOI helps AMD save power as well, so throw that into the mix.
4. The other sites were able to work with the machines for an hour, and were actually able to run their own benchmarks (most applications were preinstalled by Intel, not all) on the machines. Apple never allowed this.
5. No one ever disputed AMD has the advantage until these come out, but not a very big one. The Presler is competitive, and beats the Athlon 64 in some benchmarks. So, yes, they have an advantage, but a smaller one than they have had.
6. This is bizarre. You are clinging to synthetic benchmarks are more important than real life ones? They probably didn't include synthetics because they are essentially meaningless outside of a very limited scope.
7. Obviously, they are not GPU limited, especially with the video card configurations they installed on those machines. What you said here strikes me as borderline nonsense, except I think you make an excellent point. For most people, most games will not show a big difference with normal video card configurations, so in that sense, I agree. But, this to emphasize CPU performance so an super powerful video configuration is appropriate.
8. I am not even sure what you are saying here, but it probably has something to do with memory. Keep in mind, the benchmarks for the Conroe were made with DDR2 667 MHz, not DDR2 800 MHz. They will be what is used at release.
9. I don't think Intel is deceiving anyone, or that they could be. They let people put their hands on these machines and run their own benchmarks. The power envelope for their high performance parts are excellent. I don't know why you take such a negative view of it. Also, keep in mind they can easily offer a lower clock speed processor at very low power that would still have extraordinary performance since it performs so well clock normalized.
1. Intel is at 3.0Ghz with Woodcrest, assuming that it’s the top of the line in the server market, probably is also the top for desktop (EE). I call it Emergency Edition because I don’t know if you remember but the 1st EE Intel processor was a Xeon based northwood with 3MB L3 if I’m not mistaken, it was only released to be on top of some benchmarks charts, allowing them not losing all benchmarks to AMD...
2. 3. 4. 5. OK
6. I don’t like them too, just to get the all picture...
7. Yes that’s right, I don’t know about what it will happen on normal conditions, I’m not a big fan of SLI. Some people say its technology advancements, for me is just COPY -»PASTE technology.
8. Yes, I’m saying that current processor send to partners are "CASTRATED" in memory bandwidth, cut in HALF to be more precisely. That’s why DDR2-667 is slower than DDR1-400.
9. Yes that’s right, but AMD already have that too. In fact all processors lines have that, lower speed -» lower consuming, higher speed -» higher consuming. I will be marveled if Intel can bring 2.0Ghz beating AMD 2.6Ghz.
I'm sure the sheep will run right out and buy Intel products so people can see how naive they are. When the dust settles and Intel actually delivers their over-hyped products, you'll see they are still quite a ways behind AMDs products. AMD is not sitting around waiting for Intel to catch up.
Comparing Intel's future products - that you can't buy now, with AMD's current products, not the ones that will be available if and when Intel ever delivers there hype, shows Intel's desperation. They know the sheep have no clue and will believe the hype. Once they've blown their wad on an inferior Intel product then they won't be able to afford a superior AMD product. Ya gotta love the insanity of it all.
Anyone with a clue would compare the actual products in the marketplace at the time of purchase instead of buying the hype pumped out by intel a year in advance. But hey the hype has made Intel, ATI, Nvidia, Asus and a lot of other unscrupulous companies very wealthy. It don't take much to deceive the gullible.
Totally clueless post. Get off the anti-big money rants, they look silly after a while.
These chips are in silicon, and according the published benchmarks from another site they BURY the Athlon 64. Not beat, BURY. This is pre-release too.
Pulling an Ostrich and acting like the near future doesn't matter isn't very bright. You balance the two.
But you can’t still buy one can you?
I would like you to convince some one with you reply to Randy to buy that Intel processor today..... The only thing he will get is the P4...
What Randy was saying that you didn’t read is that Intel is comparing with AMD current products with their FUTURE products. Is not right next month…
Where was Intel when AMD invited Intel to do a server processor benchmarks and power consuming?
You talk very well TA152H, but seem to forget very well the present and the past, and talk only about the future, It’s very easy for me to talk about one GPU from X with 64 pipelines, and that it will be faster, I still don’t get the point by saying that, because when that time comes, we don’t know what will happen with it and what the others offer or will offer.
Intel compared their product to an overclocked AMD processor, not a currently released one.
This whole discussion is about a timeframe in the near future, not currently. No one is saying that the Athlon 64 is worse than the Presler. No one is saying that the current AMD chips are inferior. I am saying there is not a chance in Hell AMD will have a processor equal to the Conroe when it comes out.
So, while the Conroe is not available, it does exist. It currently blows away the Athlon 64. Both exist today, and the Athlon loses badly. Considering the maturity of the chips, the disparity will almost certainly grow as the Athlon is an aging, limping mule, while the Conroe is a shiny new warhorse just getting fitted for its armor.
AMD needs to start talking about their next generation chip soon, and you watch and see, they will. There is a new game in town, and the K8 can't play in it.
There will be some "new" K8 that will boost performance of current core by 10%. There will be some “big” news in the server market, all those pins on the socket are meant for something. If it’s for what I’m thinking there will be big news again on the server market. However, this is not technology that can be used on the desktop or mobile market.
But I guarantee to you that at least in the server market Intel will still keep losing market.
My big fear is you are right that there isn't anything new for AMD, I just keep hoping there is. They have to be working on something, they have to be. The K8 is so outclassed, it better be at least fairly close to completion too. If not, AMD is heading into the abyss.
If they lose market share in the server market, it will only be in the very near future. The Yonah is already a better processor in certain segments, and the Woodcrest blows the Opteron out of the water so will no doubt end any hope AMD has to gain market share once it is available.
So, if you are saying in the near term, I'd be inclined to agree with you. If you think it will when the Woodcrest comes out, I'd have to disagree with you completely. It's plain inferior and plus it's from a less reputable and less supportive company. It doesn't make much sense for it to win considering those conditions.
there is somethings that i didnt understand with intel preview:
-Why dont they compare the new product with their existing offer?
(I would have even looked much better vs the P4 instead the FX)
-Why they changed the drivers of the Ati cards to recognize the processor?
-What would happen if the processor didn’t get recognized?
-Was the results achived by drivers optimizations?
-Why didnt we see also a NVIDIA Geforce card there?
-I have read that HT will return on the next version of conroe, the question is not why will it will return, but why did intel disabled it on the Pentium D?
I understand HT not working (disabled) on the celeron, but on the Pentium D?
-We are all looking at a huge 4MB cache processor, should the cache give it some (big) lead? I mean I remember the Northwood L3 3MB cache was “much faster” than the 512KB version.
I can't answer those questions for sure, but I can give opinions.
I would have been more surprised if Intel did compare it to the Pentium D. The reason I say that is the Conroe isn't out yet, and by seeing it annihilate the Pentium D, and it would have been gruesome, they would have had an even more difficult time selling these foul processors. So, I'm guessing that they didn't want to cannibalize sales of the Pentium D for the six months or so that it would be the only processor available.
I'm not sure why they would change the drivers specifically, but my guess is it was to use some of the new instructions available on the processor, or in some way optimized for the terrific performance it offers. If it is a generally released driver, or will be, I don't see this as being misleading. One advantage Intel has enjoyed is better software support, and that advantage will get larger rather than smaller going forward.
I suspect Intel was most interested in choosing as close a platform as they could between themselves and AMD. For AMD, people can whine all they want, but the chipset simply doesn't matter that much anymore. The memory controller is on the processor and the performance between chipsets is much closer because of that. Chipsets on Intel processors are different though, since they can have a much greater impact because they have the memory controller. So, I would guess Intel wanted to pick a platform that would show their processor in a good light, and then pick a corresponding AMD platform that was as close as possible.
With regards to Nvidia, my suspicions are that Intel has a better and broader relationship with ATI, and seeing as how it was an ATI motherboard, I think it was a reasonable choice. Also, if they chose Nvidia, you'd probably be asking why they didn't choose ATI.
Hyperthreading doesn't work as well as you add more processors. There is still the cost of it, but often times you don't need the additional logical units because you have more to begin with. How often are four threads running on a computer at the same time? I don't mean existing as background tasks, but I mean actually needing CPU time. For most people, not too often. If you looked at some of the tests for the EE edition of the Pentium D, you saw a lot of cases where Hyperthreading showed serious performance problems and was quite undesirable. With the incredible ILP of the Conroe, I'd be even more surprised if they added Hyperthreading, particularly since they plan on adding even more cores in the near future. The advantage of using it on one processor rapidly dissipate as you add more and more cores. I would guess Hyperthreading is dead, having been usurped by multiprocessor cores. Of course, I could be wrong on this.
You can't really compare the Northwood with the Conroe in terms of cache. Increasing cache size isn't always good, since the bigger the cache, the slower the cache. The way around this is to add another level of cache, which is what the Northwood did. The bad part is you get more duplication of data, but the good part is since you didn't increase the L2 cache, you don't have to increase the latency to it. So, the L3 cache of the Northwood didn't cause any performance impact on the L2 cache. Increasing the L2 cache size would have. As you go to multiprocessor configurations, cache becomes very important because it not only increases the performance of the processor with the cache, it indirectly increases the performance of other processors in the computer since you have less contention for the memory bus. So, it can be pretty important in servers with multiple CPUs.
If your contention is that this gives Intel an undue advantage, I think we need to keep in mind that AMD has an onboard memory controller, whereas Intel doesn't. I put up a post on one of these threads a while back about that, explaining that the memory controller isn't always good. Here is an example of where Intel could use the transistors saved on implementing a memory controller on-board and instead increase the cache size. So, I guess I am saying that if you say it is unfair for Intel to have a big L2 cache, it is equally unfair for AMD to have an on-board memory controller. Actually, I don't believe either is, they made their design trade-offs and we should compare them as they are offered. AMD has the edge in the memory controller, Intel in the cache size.
AMD to sell this degenerate product
Woodcrest raped it
Not beat, BURY
Whom is "spouting your stupidity online" ?
Your comunication skills leave alot to be desired and speaking like you have just done is really not going to make anyone take your points .. even if some of them are valid. Try it one day, people tend to respond a tad better or do we really know who "isn't very bright" ?
Speaking about communication skills, do you know the meaning of run-on sentence? Good grief, learn some grammar you simpleton.
Everything I said is backed by information, I don't post links only because they would go to other sites and I don't think that is fair to this site.
You really didn't understand any of those lines? Or you did, but didn't like it? You know, I get a malicious pleasure out of watching all the rabid AMD backers whine now that the writing is on the wall. It's funny, I like AMD as a company, or at least did until Hector Ruiz took over, but the AMD backers that post here can make you want bad things to happen for AMD.
I'm guessing you're one of them. The Merom, et al, is so much better than the Athlon 64, even long before its release, it is frankly astonishing. There is an enormous disparity, and on top of this you have to factor in Intel's excellence in manufacturing, better support with compilers, motherboards and chipsets, and better support to their customers.
I have no idea how AMD is going to sell their lousy processors after these are released. They're slow and use more energy, and this with an on-board memory controller. Plus, compilers come out to support the processor, it's only going to get worse since people don't write programs to execute well on the little guy in preference to the one with the big sales. On top of this, the little guy will get a lot smaller since they have a non-competitive product. Even in games, AMD's heretofore stronghold, the Athlon 64 gets brutally beaten. This type of difference is incredible, and shows that design was not nearly so mature as many people (including myself) had thought and somehow Intel greatly improved IPC while also lowering power.
This is a fantastic processor that has rightly put all the lousy x86 processors like the Athlon 64 and Pentium 4 in their proper place as rubbish. Even Intel says this is the best processor they have ever made, which might not mean anything except when you consider they also make the Itanium and probably shouldn't be discrediting it. Anyone seeing the benchmarks has to be in awe of it, especially several months ahead of release. It's plain shocking.
AMD had better have something ready, and soon, to replace the degenerate Athlon 64. It's looks so outdated now, AMD will have a hard time surviving if the next design isn't ready soon. When Intel has a lousy product, they lose market share and make less money. When AMD has a lousy product, they lose massive amounts of money are risk going out of business. We better hope they have something out before 2008 that can compete with these incredible chips, or they'll be buried along with the miserable Athlon 64. As much as I'd like to hear the whining and crying here, I wouldn't like having to pay huge amounts of money for processors because Intel has no competition. It's almost worth it though.
Conroe and all this stuff is a preview, a64 is shipping and kicks every Intel chip. Hard.
Let's see what happen after AM2 intro and Intel Core
I have to satisfy my costumers in the next 6 months with a good product, I will sell them paper specs? Benchmarks? Futurology? I don’t think so.
Next 6 months AMD will keep it’s market share increase, and I doubt Intel will right in 6 months will be with high availability of all those systems.
Add your Comment
Enter your username and e-mail address. Password will be sent to you.