Does AMD have anything special on the horizon?

[Viditor]

Originally posted by: harpoon84

Originally posted by: Viditor

Originally posted by: cmdrdredd
There's no reaspn to think that Intel is going to do 45nm and NOT increase the performance at the same time

True, but there's no reason that they will either...they didn't with the 65nm node on Netburst.

True, but 45nm C2D will feature 6MB L2 cache, 1333MHz FSB and SSE4, so it should be at least 5% faster clock for clock compared to current C2Ds. SSE4 benefits are still unknown at this stage so I won't bother speculating.

Of course with a die shrink there is also the advantage of lower power consumption, which in turn generally leads to higher clockspeeds. Intel is aiming for 3.5 - 4GHz on 45nm DC and 3.46 - 3.73GHz on 45nm QC according to the latest roadmap.

True enough...the question is how much will this help?
1. The increase in cache will have very limited benefits. It will bring the memory latency closer (but not equal) to AMD's ODMC for the majority (but not all) apps. It will also use up the lion's share of the real estate gained from the shrink (meaning that they won't get as big of a cost savings or room for additions).
2. A 1333 FSB will help the most, but even that has limited gains (as we've seen with Woodcrest). I would agree though that this will be the majority of any gain and may indeed increase performance by 5% (don't know yet...).
3. SSE4 will be on the K8L as well, but that's a future story as it requires software to be written that uses it (and we all know that this can take a year or more from other SSE releases).

As to Intel's goals for clockspeeds, I have seen you say that before, but I have to hear Intel say anything of the kind...do you have a reliable source for the rumour?

 

[Arkaign]

Hehehe

I don't have anything firm to go on regarding why I have predicted what I have on the near and intermediate future of AMD, it *is* definitely just my gut feeling and merely conjecture.

I'm looking between the lines at a few recent occurrences, and sort of draw my ideas from them.

(1)- AMD's very quiet now on the whole 4x4 front. I'm guessing that they have discretely benched them vs Kentsfield and aren't about to publicize an expensive, possibly slower (in 50% or more?) at some major apps, that sits on a questionable platform. I know I'd rather have something with a long expected lifespan, and if Viditor's AM2+ info is correct, then CURRENT AM2 boards may have a longer viable time frame Of course, a few nuts will buy 4x4, and for those that enjoy the particular advantages that may be found, good for them.

(2)- An old (and very suspect) link on possible delays : http://www.overclockers.com/tips00993/

(3)- Both AMD and Intel constantly miss deadlines, and have launch issues. Notable problems for both have been well documented.

Intel : P3-1.13 Recall, yield issues getting volume against readily-available AMD 1Ghz parts, recalls on Intel chipsets with Sdram/Rdram MTH, weak initial P4 performance with Willamette, the short-lived socket 423, spastically expensive rdram prices affecting anyone wanting a fast Intel box, godawful .90 issues covering the whole gamut, major heat/power problems to the consumer, sliding solidly behind AMD64 for the duration of the Netburst era.

AMD : Significant K7 launch delays, numerous compability/stability issues on the Irongate (and a few VIA) chipsets related to certain video cards/sound cards, short-lived Slot-A interface, extremely fragile initial Socket-A procs, ultra heat problems (with the AMD retail heatsink) with the final Thunderbirds, total lack of competitive offering against northwood 800fsb (eg; P4 2.4C vs 2400+ was about even, but by the time 3.2C vs 3200+, it was no contest, and was laughable unless the AMD was on the outstanding NF2 chipset), slow initial Athlon64 launch, the short-lived socket 754 socket (for anything but budget use)

Anyway, I look at it as somewhat unlikely that AMD will magically release K8L on time and with price/performance/availability/stability being sufficient to overtake Intel at all levels initially. I do think they can and *WILL* overtake Intel again, but more likely to be a couple revisions of K8L down the line, or the next major architechture overhaul entirely.

The fate of 4x4 will tell us volumes.

Just my thoughts, anyway, we'll see in a few months either way.

 

[harpoon84]

I believe we've discussed this already Viditor, I'm mainly going off the VR-Zone site. They have a lot of 'updated' roadmaps on everything AMD and Intel.

3.5 - 4GHz for 45nm seems reasonable, considering most current C2Ds can overclock to 3GHz on default voltage. 3.5GHz is often attainable with raised voltages, which of course means power consumption skyrockets. This is where 45nm comes in.

It looks like we're going to have these kind of debates for the best part of the next 9 months Viditor. It'll get tiring repeating myself after a while, I suggest we just relax and let everything run it's course.

 

[Viditor]

Originally posted by: harpoon84
I believe we've discussed this already Viditor, I'm mainly going off the VR-Zone site. They have a lot of 'updated' roadmaps on everything AMD and Intel.

3.5 - 4GHz for 45nm seems reasonable, considering most current C2Ds can overclock to 3GHz on default voltage. 3.5GHz is often attainable with raised voltages, which of course means power consumption skyrockets. This is where 45nm comes in.

It looks like we're going to have these kind of debates for the best part of the next 9 months Viditor. It'll get tiring repeating myself after a while, I suggest we just relax and let everything run it's course.

Fair enough Harpoon...I just wanted to be clear that this was a guess and nothing Intel has talked about...

 

[myocardia]

Originally posted by: PingSpike
AMD are doing ok right now, but they don't have anything really competitive coming out for the high end and are suffering from the inability to make enough chips to meet demand. (which isn't that horrible of a position to be in really)

The trouble with AMD is they have no fvcking money. If they had money they probably would have blown it on R&D to stay at the top performance wise. But instead they had a ton of debt...

Agreed. Unfortunately, they decided to go into debt by buying ATI. Whether that will pay off for them (I don't see it), we'll all have to wait and see. I also think that they should have spent their money on R&D, since even though "joe sixpack" doesn't ever buy the most expensive processor, he almost always buys a processor from the company who has the fastest processor. Look at how many people out there own A64's, compared to just a few years ago, right before the A64's were the fastest thing available.

 

[myocardia]

Originally posted by: Viditor

Originally posted by: cmdrdredd
quoting rumor sites as fact is very wrong...

oh well...live the dream for now.

If you are referring to the quote from the Inquirer, then I suggest you try a google of the terms "am2+ compatible"...this is really well known now. The Inq was just one of the first that popped up for me...

I'm still waiting on you to come up with a link that explicitly quotes any AMD executive at all saying that an AM2 mobo will support any processor besides the currently available AM2's. BTW, I've been waiting for quite a few months for this link, it seems.

 

[Stoneburner]

Let's say K8L does perform, to whatever degree, better than the current core architecture... why do people believe intel can take the crown back without a new archictecture? I know intel said it would develop new architectures every 2 years or so but intel's screwed up with their promises pretty often and this would seem rather difficult regardless. I'm not sure, despite viditor's reasonable arguments, AMD will take the lead but i am far less convinced die shrinks are going to keep the crown in Intel's hands... it has failed them before.

 

[DrMrLordX]

Do we have any idea what K8L's clockspeed will be like at launch?

 

[Griswold]

Originally posted by: Duvie
Even if AMD pulls off a decent launch with its K8L and it is faster clock for clock it still loses...Why? because it will be a whole new socket and perhaps even newer memory ddr3, while for many of us who have scrapped AMD systems for C2D systems can pop a Kentsfield quad core in current systems....

What kind of humbug ist that? New socket? DDR3? Please...

 

[zsdersw]

I'd say you can't judge potential for clock speed increases on Core processors with the 65nm -> 45nm shrink by the lack of increases the 90nm -> 65nm shrink brought to Netburst.

 

[Viditor]

Originally posted by: myocardia

Originally posted by: Viditor

Originally posted by: cmdrdredd
quoting rumor sites as fact is very wrong...

oh well...live the dream for now.

If you are referring to the quote from the Inquirer, then I suggest you try a google of the terms "am2+ compatible"...this is really well known now. The Inq was just one of the first that popped up for me...

I'm still waiting on you to come up with a link that explicitly quotes any AMD executive at all saying that an AM2 mobo will support any processor besides the currently available AM2's. BTW, I've been waiting for quite a few months for this link, it seems.

Sorry, I didn't realise you were waiting...
Glad to help though, it's in the same press release where they announced the AM2+ socket.
BTW, don't bother looking because they haven't announced that yet either...the source for all of the articles on Google is the same source that was cited for AM2+ in the first place, namely HKEPC.

 

[Viditor]

Originally posted by: zsdersw
I'd say you can't judge potential for clock speed increases on Core processors with the 65nm -> 45nm shrink by the lack of increases the 90nm -> 65nm shrink brought to Netburst.

Ummm...why not?

 

[Brunnis]

Originally posted by: DrMrLordX
Do we have any idea what K8L's clockspeed will be like at launch?

Yep, 2.9GHz is the rumored max frequency at launch. Considering that the new CPUs won't be out for a pretty long time, it's pretty safe to assume that this may change several times until then. It will probably depend heavily on what Intel have at the time and also how much AMD can squeeze out of their 65nm process.

Originally posted by: DrMrLordXEven if AMD pulls off a decent launch with its K8L and it is faster clock for clock it still loses...Why? because it will be a whole new socket and perhaps even newer memory ddr3, while for many of us who have scrapped AMD systems for C2D systems can pop a Kentsfield quad core in current systems....

DDR3 will most definitely not be used and there seems to be a pretty high probability of AM2+ chips working just fine in AM2 boards. So, I wouldn't say that things look that bad really...

 

[zsdersw]

Originally posted by: Viditor

Originally posted by: zsdersw
I'd say you can't judge potential for clock speed increases on Core processors with the 65nm -> 45nm shrink by the lack of increases the 90nm -> 65nm shrink brought to Netburst.

Ummm...why not?

1. They're different architectures.

2. 65nm came at the end of Netburst's life. Had Core not been just around the corner, Presler/Cedar Mill could've seen small clock increases.

 

[Viditor]

Originally posted by: zsdersw

Originally posted by: Viditor

Originally posted by: zsdersw
I'd say you can't judge potential for clock speed increases on Core processors with the 65nm -> 45nm shrink by the lack of increases the 90nm -> 65nm shrink brought to Netburst.

Ummm...why not?

1. They're different architectures.

OK...can you give me an example of ANY architecture that improved performance from a node change and without architecture changes?

2. 65nm came at the end of Netburst's life. Had Core not been just around the corner, Presler/Cedar Mill could've seen small clock increases.

I agree, but that's to be expected from both...especially since AMD is releasing the split power planes with K8L (architecture change). I assume that they will stay pretty close to equal there...

 

[zsdersw]

Originally posted by: Viditor
OK...can you give me an example of ANY architecture that improved performance from a node change and without architecture changes?

It may end up that the 65-45 shrink for Core won't yield chips binned at higher speeds, but that won't be because the 90-65 shrink for Netburst didn't do much for clock speeds.

All I'm saying is that you cannot use one to predict another, given the differences I've cited.

I agree, but that's to be expected from both...especially since AMD is releasing the split power planes with K8L (architecture change). I assume that they will stay pretty close to equal there...

Who's talking about AMD? All I'm talking about is the 65-45 shrink for Core and the 90-65 shrink for Netburst. The particulars of AMD chips and processes have nothing to do with that.

 

[Regs]

Performance is really only a small problem for AMD. It's the price war and inventory that's killing them. Plus the merger with ATI. When you have two company's that have consistent track records and product releases and slam them together for a derailment it will make stock holders a little worry. This is why AMD is trying their best to let them know what the combined track is for the upcoming years now that expectations are lowered.

As for performance...who cares? I know I do but the majority of people who buy them care more about what retail stores or vendor its coming from and the price more so than the performance.

 

[MrX8503]

Originally posted by: Duvie
Even if AMD pulls off a decent launch with its K8L and it is faster clock for clock it still loses...Why? because it will be a whole new socket and perhaps even newer memory ddr3, while for many of us who have scrapped AMD systems for C2D systems can pop a Kentsfield quad core in current systems....

So your saying that you didnt have to buy all new stuff when you went from AMD X2 to Intel C2D?

 

[DrMrLordX]

Originally posted by: Brunnis
[

Originally posted by: DrMrLordXEven if AMD pulls off a decent launch with its K8L and it is faster clock for clock it still loses...Why? because it will be a whole new socket and perhaps even newer memory ddr3, while for many of us who have scrapped AMD systems for C2D systems can pop a Kentsfield quad core in current systems....

Eh? That's not my quote there. That was posted by Duvie.

Thanks for the figure on K8L's projected launch speed, though. If K8L won't be exceeding 2.9 ghz stock, I'm thinking Intel may have the performance edge in clockspeed alone. For awhile, at least.

But as you said, stuff like that is subject to change.

 

[JEDIYoda]

Originally posted by: Viditor

Originally posted by: cmdrdredd
quoting rumor sites as fact is very wrong...

oh well...live the dream for now.

If you are referring to the quote from the Inquirer, then I suggest you try a google of the terms "am2+ compatible"...this is really well known now. The Inq was just one of the first that popped up for me...

always been a die hard supporter of AMD...huh??

 

[BigDH01]

I'm getting the impression that AMD will try to get close to C2D performance but probably won't overtake it. In all honesty, I don't know if they are really too concerned.

Since they purchased ATI, what has AMD been focused on? Platforms. Torrenza, Fusion, you name it. AMD is going to do what is best to make them and the stockholders money... focus on the server segments, budget segments, and mobile segments. Although we might care about the enthusiast/performance segments, there is more money to be had elsewhere. The server segments will offer them margin and the budget/mobile segments will offer them volume. It makes perfect sense.

The Fusion platform will make them very competitive in the mobile/budget segments and alternative platforms like Torrenza may make them very competitive in the server segments where they are already growing. If you look at the roadmaps, what does AMD emphasize? You'll see new names for the revisions, but the roadmaps highlight platform improvements. DDR3, HT3.0. They seem very concerned with their chips ability to scale performance based on number of CPUs. That's where AMD is going... for good or bad. For the stockholders, this is probably a very good thing as there is profit in these markets. For us enthusiasts, this is probably a bad thing as enthusiast performance parts are probably going to take a back seat. However, it's the stockholders that own the company, not us.

I'm guessing AMD will be happy if K8L comes close to the performance of Core. They will remain competitive in the enthusiast market but probably won't end up on top for a very long time. And why would they care? They are selling out chips to HP, Dell, and other retailers and there aren't many CPUs left for channels usually belonging to enthusiasts like us. AMD is not stupid, they've looked ahead at Intel's aggressive schedule and decided it wasn't worth being in that fight when they have a chip that will scale very well in the server market where the margins are larger.

The ATI acquisition didn't make sense to me until I thought of it in these terms. They've gained access to a platform that they can gear to the budget/mobile market. They've also gained access to powerful GPUs that can be retooled to perform scientific calculations not involved in realtime 3d rendering (like games). Then they announced Torrenza and Fusion and now it makes perfect sense. I just think everyone should be prepared that AMD might be exiting the high performance desktop market.

4x4 is also a mystery to me. I thought it was a mistake when Intel cannibalized their server line to release a Xeon for the desktop. As an enthusiast, it was neat to see (althought quite desperate), but as a shareholder I thought it was a mistake. I feel the same way about 4x4. 4x4 will just be using rebadged Santa Rosas that will be sold at a huge discount. Those are valuable chips that could be sold in the server segment and make more money for AMD. I don't think AMD should cannibalize their server segment to try to compete with Kentsfield. And I think the execs at AMD have come to this realization. 4x4 was definitely hyped, but it appears the original release date was missed and things have been pretty quite with regards to 4x4 even though a release must be imminent. Usually, review sites are dropping hints about performance this close to release, or at least mentioning they have a new toy to play with. I just haven't found anything like this and I'm guessing it's because review sites simply don't have their hands on a reviewable product. The INQ blames the delay on a shortage of CPUs. Let me ask you this... if AMD is in short supply of Socket F Santa Rosas because of high demand, will they really be in a hurry to release this chip to a desktop market at a discount? Think like a stockholder or exec. Bad idea. This is especially true as it looks like the motherboard alone will cost $400-500 (based on comparable Intel nVidia chipset) and AMD will have to sell Santa Rosas at a huge discount if they want to compete with Kentsfield that can drop into much cheaper motherboards (such as my DS3). Remember, AMD is trying to sell the 4x4 platform and not just the CPUs. The profit incentive just isn't there when AMD will have a native quad core next year. I see the 4x4 maybe coming out before the end of the year, and even then, in very limited quantities.

In the end, AMD is about making money and the high performance/enthusiast desktop market is just not #1 priority right now and perhaps not in the foreseeable future.

 

[Arkaign]

Originally posted by: BigDH01
I'm getting the impression that AMD will try to get close to C2D performance but probably won't overtake it. In all honesty, I don't know if they are really too concerned.

Since they purchased ATI, what has AMD been focused on? Platforms. Torrenza, Fusion, you name it. AMD is going to do what is best to make them and the stockholders money... focus on the server segments, budget segments, and mobile segments. Although we might care about the enthusiast/performance segments, there is more money to be had elsewhere. The server segments will offer them margin and the budget/mobile segments will offer them volume. It makes perfect sense.

The Fusion platform will make them very competitive in the mobile/budget segments and alternative platforms like Torrenza may make them very competitive in the server segments where they are already growing. If you look at the roadmaps, what does AMD emphasize? You'll see new names for the revisions, but the roadmaps highlight platform improvements. DDR3, HT3.0. They seem very concerned with their chips ability to scale performance based on number of CPUs. That's where AMD is going... for good or bad. For the stockholders, this is probably a very good thing as there is profit in these markets. For us enthusiasts, this is probably a bad thing as enthusiast performance parts are probably going to take a back seat. However, it's the stockholders that own the company, not us.

I'm guessing AMD will be happy if K8L comes close to the performance of Core. They will remain competitive in the enthusiast market but probably won't end up on top for a very long time. And why would they care? They are selling out chips to HP, Dell, and other retailers and there aren't many CPUs left for channels usually belonging to enthusiasts like us. AMD is not stupid, they've looked ahead at Intel's aggressive schedule and decided it wasn't worth being in that fight when they have a chip that will scale very well in the server market where the margins are larger.

The ATI acquisition didn't make sense to me until I thought of it in these terms. They've gained access to a platform that they can gear to the budget/mobile market. They've also gained access to powerful GPUs that can be retooled to perform scientific calculations not involved in realtime 3d rendering (like games). Then they announced Torrenza and Fusion and now it makes perfect sense. I just think everyone should be prepared that AMD might be exiting the high performance desktop market.

4x4 is also a mystery to me. I thought it was a mistake when Intel cannibalized their server line to release a Xeon for the desktop. As an enthusiast, it was neat to see (althought quite desperate), but as a shareholder I thought it was a mistake. I feel the same way about 4x4. 4x4 will just be using rebadged Santa Rosas that will be sold at a huge discount. Those are valuable chips that could be sold in the server segment and make more money for AMD. I don't think AMD should cannibalize their server segment to try to compete with Kentsfield. And I think the execs at AMD have come to this realization. 4x4 was definitely hyped, but it appears the original release date was missed and things have been pretty quite with regards to 4x4 even though a release must be imminent. Usually, review sites are dropping hints about performance this close to release, or at least mentioning they have a new toy to play with. I just haven't found anything like this and I'm guessing it's because review sites simply don't have their hands on a reviewable product. The INQ blames the delay on a shortage of CPUs. Let me ask you this... if AMD is in short supply of Socket F Santa Rosas because of high demand, will they really be in a hurry to release this chip to a desktop market at a discount? Think like a stockholder or exec. Bad idea. This is especially true as it looks like the motherboard alone will cost $400-500 (based on comparable Intel nVidia chipset) and AMD will have to sell Santa Rosas at a huge discount if they want to compete with Kentsfield that can drop into much cheaper motherboards (such as my DS3). Remember, AMD is trying to sell the 4x4 platform and not just the CPUs. The profit incentive just isn't there when AMD will have a native quad core next year. I see the 4x4 maybe coming out before the end of the year, and even then, in very limited quantities.

In the end, AMD is about making money and the high performance/enthusiast desktop market is just not #1 priority right now and perhaps not in the foreseeable future.

Good stuff

 

[JumpingJack]

Originally posted by: Dimkaumd
I recently bought shares of AMD because I thought the stock dropped significantly and had room for a rise. However, since then I am down about 5% and upon further review I am noticing that AMD doesnt have anything special on the market right now to compete with Intel. Intels Core 2 Duos and the new Quards seem to be taking market share away from AMD again.

Anyone think that AMD can regain its strength through upcoming products? Will 2007 be dominated by Intel chips or AMD?

AMD will release 65 nm in December of this year; however, initial launch speeds will not clock higher than their 90 nm parts. The highest clocked 65 nm CPU will come in at 2.5 to 2.6 GHz and they will push their 90 nm CPUs to 3.0 GHz, there is a shrink involved obviously, but no major architectural changes therefore the 65 nm product clocked at 2.6 GHz will perform the same as the 90 nm product clocked at 2.6 GHz. This is why the 65 nm products are taking up the 4400, 4600, 4800, and 5000 model numbers.

AMD is targetting mid 2007 for K8L, or some have called Barcelona. A heavily reworked revision to the current K8 core but not a completely new architectural approach. The list of details AMD has shared appears they are focusing mostly on bandwidth improvements and some on IPC improvements. On paper it looks like a move in the right direction, you, me and many others are anxiously awaiting test results to see how these improvements translate into performance.

Beyond that, it is a little merky but details are leaking out... you can go to AMDs website or scower DailyTech to find various roadmaps and timelines.

 

[Viditor]

Originally posted by: BigDH01
I'm getting the impression that AMD will try to get close to C2D performance but probably won't overtake it. In all honesty, I don't know if they are really too concerned.

Since they purchased ATI, what has AMD been focused on? Platforms. Torrenza, Fusion, you name it. AMD is going to do what is best to make them and the stockholders money... focus on the server segments, budget segments, and mobile segments. Although we might care about the enthusiast/performance segments, there is more money to be had elsewhere. The server segments will offer them margin and the budget/mobile segments will offer them volume. It makes perfect sense.

The Fusion platform will make them very competitive in the mobile/budget segments and alternative platforms like Torrenza may make them very competitive in the server segments where they are already growing. If you look at the roadmaps, what does AMD emphasize? You'll see new names for the revisions, but the roadmaps highlight platform improvements. DDR3, HT3.0. They seem very concerned with their chips ability to scale performance based on number of CPUs. That's where AMD is going... for good or bad. For the stockholders, this is probably a very good thing as there is profit in these markets. For us enthusiasts, this is probably a bad thing as enthusiast performance parts are probably going to take a back seat. However, it's the stockholders that own the company, not us.

I'm guessing AMD will be happy if K8L comes close to the performance of Core. They will remain competitive in the enthusiast market but probably won't end up on top for a very long time. And why would they care? They are selling out chips to HP, Dell, and other retailers and there aren't many CPUs left for channels usually belonging to enthusiasts like us. AMD is not stupid, they've looked ahead at Intel's aggressive schedule and decided it wasn't worth being in that fight when they have a chip that will scale very well in the server market where the margins are larger.

The ATI acquisition didn't make sense to me until I thought of it in these terms. They've gained access to a platform that they can gear to the budget/mobile market. They've also gained access to powerful GPUs that can be retooled to perform scientific calculations not involved in realtime 3d rendering (like games). Then they announced Torrenza and Fusion and now it makes perfect sense. I just think everyone should be prepared that AMD might be exiting the high performance desktop market.

4x4 is also a mystery to me. I thought it was a mistake when Intel cannibalized their server line to release a Xeon for the desktop. As an enthusiast, it was neat to see (althought quite desperate), but as a shareholder I thought it was a mistake. I feel the same way about 4x4. 4x4 will just be using rebadged Santa Rosas that will be sold at a huge discount. Those are valuable chips that could be sold in the server segment and make more money for AMD. I don't think AMD should cannibalize their server segment to try to compete with Kentsfield. And I think the execs at AMD have come to this realization. 4x4 was definitely hyped, but it appears the original release date was missed and things have been pretty quite with regards to 4x4 even though a release must be imminent. Usually, review sites are dropping hints about performance this close to release, or at least mentioning they have a new toy to play with. I just haven't found anything like this and I'm guessing it's because review sites simply don't have their hands on a reviewable product. The INQ blames the delay on a shortage of CPUs. Let me ask you this... if AMD is in short supply of Socket F Santa Rosas because of high demand, will they really be in a hurry to release this chip to a desktop market at a discount? Think like a stockholder or exec. Bad idea. This is especially true as it looks like the motherboard alone will cost $400-500 (based on comparable Intel nVidia chipset) and AMD will have to sell Santa Rosas at a huge discount if they want to compete with Kentsfield that can drop into much cheaper motherboards (such as my DS3). Remember, AMD is trying to sell the 4x4 platform and not just the CPUs. The profit incentive just isn't there when AMD will have a native quad core next year. I see the 4x4 maybe coming out before the end of the year, and even then, in very limited quantities.

In the end, AMD is about making money and the high performance/enthusiast desktop market is just not #1 priority right now and perhaps not in the foreseeable future.

That's a well thought out and written post, and I agree with a lot of it...the only problem I have (which seems to be a common one) is the line

"I'm getting the impression that AMD will try to get close to C2D performance but probably won't overtake it"

I understand that it's an impression, and I agree that it really won't effect AMD all that much (some excellent insight there BTW)... But, if you look at the technology (as we did when C2D was just an announcement) then I don't see how that impression can be true (unless, as is possible, you don't think AMD will be able to deliver what they promise).
If you look closely at the advancements in C2D, and then at the announced advancements in K8L, you'll note that the only difference between them is that C2D will be able to retire 4 ops per clock...
That would be very cool (and advantageous for Intel) if applications actually allowed you to do it, but almost none of them do.
Then you have to look at the differences between the platforms (HT vs FSB), and unless AMD screws it up, it's almost a foregone conclusion that K8L will be faster...at least that's what it shows on paper.

While Intel's refresh at 45nm will let them increase the bandwidth (FSB 1333), it still won't come close to HT 3.0.

That said, the next real advance for Intel won't be until Nehelam at the end of 2008...when they get CSI. Almost nothing is known about how this will perform (I don't think they have working silicon this far out), but I strongly doubt that Intel will let us down on it. I'm expecting it to be a blockbuster...

 

[DrMrLordX]

Originally posted by: Viditor

If you look closely at the advancements in C2D, and then at the announced advancements in K8L, you'll note that the only difference between them is that C2D will be able to retire 4 ops per clock...
That would be very cool (and advantageous for Intel) if applications actually allowed you to do it, but almost none of them do.
Then you have to look at the differences between the platforms (HT vs FSB), and unless AMD screws it up, it's almost a foregone conclusion that K8L will be faster...at least that's what it shows on paper.

Intel will probably have a clock speed advantage if K8L tops out at 2.9 ghz on release. Probably. However, if Intel rolls out a stock 3.0-3.2 ghz Kenstfield to counter K8L, expect it to be a hot chip (literally).

Whether or not said clock speed advantage will be enough to tip scales in Intel's favor remains to be seen. I imagine that it will.