Disappointed by the Conroe launch

[dexvx]

Originally posted by: liebremx

Originally posted by: Quinton McLeod

Originally posted by: dexvx

Anyone who thinks the identification string of a BIOS has any relevance to performance is stupid.

Quit kidding yourself. The bios takes more than just identification for a processor. It also checks the default multipliers from the processor. It also checks and sees what it supports (Cool 'n Quiet and etc.). If you honestly think your performance wouldn't change if your bios couldn't identify your processor, then you're seriously in the dark. I also wouldn't want you working on the general public's computers. You might blow them up.

*Woot* Now you see why the string was irrelevant? The only relevant thing in that new BIOS regarding the benchmarks I'm pretty sure is the crossfire fix. They could have left the FX-60 fix and IT WOULD STILL BE PROPERLY CONFIGURED even if the string says otherwise.

Kinda ironic. The fanbois are just digging themselves a hole. The Conroe system was actually using 5-5-5-15 instead of 4-4-4-14. But in reality, the best DDR2 operates at 3-2-2-12 (I mean, it'd only be fair since the FX system used the best DDR available). Then the fanbois were crying that the Intel supplied demos were "biased". Low and behold Anand tried new demos and found that the Intel supplied demos were actually conservative. So basically, Quake4 went from 20% Conroe favored to 30%. Besides the FEAR fluke, which Anand did on mistake, Conroe's performance delta actually increased since yesterday. What will they think of next?

But yea, I don't know ANYTHING about computers. I might blow them up because I think the BIOS identification string has no relevance to performance.

 

[Quinton McLeod]

Originally posted by: dexvx

Originally posted by: Quinton McLeod

Originally posted by: dexvx

Anyone who thinks the identification string of a BIOS has any relevance to performance is stupid.

Quit kidding yourself. The bios takes more than just identification for a processor. It also checks the default multipliers from the processor. It also checks and sees what it supports (Cool 'n Quiet and etc.). If you honestly think your performance wouldn't change if your bios couldn't identify your processor, then you're seriously in the dark. I also wouldn't want you working on the general public's computers. You might blow them up.

You sir, just got pwned.

http://www.anandtech.com/tradeshows/showdoc.aspx?i=2716&p=2

If the BIOS cannot use the CPU properly, it would not even boot up. Like I said, all thats missing from the old DFI board was a microcode update for the FX-60. All those "bug" fixes from the DFI RD480 are just minor glitches that has almost no relevance to performance.

shock
denial <-- You are still here.
depression
anger
acceptance

--------------

Quotes from Anand Followup:

With SMP enabled we see that Conroe holds an even larger 31% performance advantage and with it disabled, the unreleased CPU was 29% faster. If anything, Intel?s own demo was a little more conservative on Conroe and definitely not optimized to make AMD look bad.

We inadvertently left the AMD system at a higher resolution (1280 x 960) instead of the default resolution (1024 x 768) when you select the ?High? Graphics defaults. The oversight was entirely our own doing as Intel was not running the benchmarks or configuring them, it simply happened while we were setting up both systems at the same time.

As we mentioned earlier, the Conroe system was supposed to be running at 4-4-4-15 timings but was instead configured at 5-5-5-15, giving Conroe a bit of a disadvantage. We re-ran all of our tests with the timings correctly set to 4-4-4-15, the results are below:

While the faster timings gave Intel a small performance boost in the games, it?s nothing to write home about.

Well, maybe you should read this:

I guess there is such a thing as a dumb question. Any of the new processors have their unique identifiers setup regardless of overclocking. The bottom line is DFI said that the FX-60 will not operate at full potential with the installed bios.

Link: http://www.blogger.com/comment.g?blogID=17375431&postID=114184911431790820

Straight from the mouth of Rahul Sood...

Besides, Anatech pointed out that, upon a Bios update, many important things were fixed that allowed the FX-60 to perform much better!

1. Fix memory Set 2-1-1-1-1 and 4-1-1 mode wrong.
2. Set Cool'n'Quiet default disable.
3. Change the description of DQDRV.
4. Fix Read Preamble Table Error.
5. Shorten the delay time during clock programming loop.
6. Add over clocks step by step.
7. Fix fill 3114 SVID&SSID under Cross fire mode.
8. Fix soft-reset hang on POST code F2h if enable USB mouse support.
9. Change CMOS used to fix some control item can?t save.
10. Add support K8 FX60 CPU.
11. Update SiI3112 Raid ROM.
12. Fix some SATA(DiamondMax 10 (6B160M0) HDD ) HDD detect fail at first time cool boot.

I'm sorry to say, but that's a very important update for the FX-60! I bet you don't even know what half of those updates do.

 

[Quinton McLeod]

Originally posted by: dexvx
Kinda ironic. The fanbois are just digging themselves a hole. The Conroe system was actually using 5-5-5-15 instead of 4-4-4-14. But in reality, the best DDR2 operates at 3-2-2-12 (I mean, it'd only be fair since the FX system used the best DDR available). Then the fanbois were crying that the Intel supplied demos were "biased". Low and behold Anand tried new demos and found that the Intel supplied demos were actually conservative. So basically, Quake4 went from 20% Conroe favored to 30%. Besides the FEAR fluke, which Anand did on mistake, Conroe's performance delta actually increased since yesterday. What will they think of next?

But yea, I don't know ANYTHING about computers. I might blow them up because I think the BIOS identification string has no relevance to performance.

The fact of the matter is, Conroe doesn't need timings. Timings don't mean much to Intel. Intel needs Bandwidth. So, that argument is pretty much moot.

Also, none of us can use these benchmarks are a way to glorify and crown Intel.

1) Intel provided controlled machines in which many others were not even allowed to change.

2) The processor is being compared to old technology and is slanted to be released in 6 months. You would have to be completely brain dead to expect AMD to not release anything by that time. Intel needs to compare apples with apples. Right now, comparisons are meaningless because NONE of us can even touch the processor until later this year.

3) Anandtech tried demos on an Intel controlled machine. To make an unbias benchmark, they need to test it in their own labs. This benchmark is meaningless at this point in time.

 

[dmens]

Timings don't mean much to Intel. Intel needs Bandwidth.

Uh.... bandwidth is both latency and width, and with can be addressed by wider and/or multiple busses. Latency can only be addressed by timing, or having HT/controller or CSI/TT. Latency is merom's weakness.

Intel provided controlled machines in which many others were not even allowed to change.

Bitch and whine. The fact that Anand was allowed to make a mistake by messing up FEAR resolution means he had plenty of control for a public pre-release demo.

The processor is being compared to old technology and is slanted to be released in 6 months. You would have to be completely brain dead to expect AMD to not release anything by that time

You'd have to be braindead to think DDR2 will be able to close the gap. No one has even heard a word on K10. What will they have in 6 months? Oh right, DDR2, plus speedsteps.

This benchmark is meaningless at this point in time.

Suit yourself. I wonder if your fanatic skepticism would be duplicated when K10 is demo'ed before release.

 

[Quinton McLeod]

Originally posted by: dmens

Timings don't mean much to Intel. Intel needs Bandwidth.

Uh.... bandwidth is both latency and width, and with can be addressed by wider and/or multiple busses. Latency can only be addressed by timing, or having HT/controller or CSI/TT. Latency is merom's weakness.

No, you're completely wrong.
Bandwidth is how much data can be passed within a given time frame.
Timings are associated to the CPU clock cycle in which data is allowed to move to and from.

Intel provided controlled machines in which many others were not even allowed to change.

Bitch and whine. The fact that Anand was allowed to make a mistake by messing up FEAR resolution means he had plenty of control for a public pre-release demo.

No they did not have plenty of control. They did not have control over the motherboard, the graphics cards used, the processor.. Heck, the entire system! Intel didn't even allow Anandtech to reinstall Windows after the bios upgrade. They obviously didn't have enough control to make this an accurate Benchmark.

The processor is being compared to old technology and is slanted to be released in 6 months. You would have to be completely brain dead to expect AMD to not release anything by that time

You'd have to be braindead to think DDR2 will be able to close the gap. No one has even heard a word on K10. What will they have in 6 months? Oh right, DDR2, plus speedsteps.

First of all, AMD has only said AM2 will support DDR 2. They did not say much about anything else. To even draw a conclusion on information you don't know much about is foolish.

This benchmark is meaningless at this point in time.

Suit yourself. I wonder if your fanatic skepticism would be duplicated when K10 is demo'ed before release.

I wonder if your fanatic skepticism would be duplicated when the K10 is demo'ed before release

 

[Markfw]

Quinton, You are in left field dude, you are totally wrong. Please don't act like a fanboi....

 

[dexvx]

Just dont give up eh?

Anand quotes from the followup:

With the new BIOS installed we confirmed that Cool?n?Quiet was disabled, so that was not impacting the performance results at all. The new BIOS also correctly identified the Athlon 64 FX-60 processor, although as you can see from the results above, the proper detection of the CPU didn?t translate into greater performance.

Originally posted by: Quinton McLeod
The fact of the matter is, Conroe doesn't need timings. Timings don't mean much to Intel. Intel needs Bandwidth. So, that argument is pretty much moot.

Nope, quote from Anand: "While the faster timings gave Intel a small performance boost in the games, it?s nothing to write home about."

 

[bstowe94]

Its just good business. The line from Intel is that conroe is a better product and people buy better products should be a sign that they are taking the AMD challenge seriously and we as consumers will be better off for it. Sure some folks may hold off a new PC upgrade for 6 months but Intel has just fired a warning shot across the bow and it is up to AMD to respond with a better product.

If you look at the position Intel has been in... losing market share, tons of rumors of Dell adopting AMD processors, missing earnings forecasts, and not having a competitive product thier hand was forced. This is a good move on thier part to calm fears that they were far behind in the performance category and it looks like they have done that.

AMD now has 6 months to make improvements to thier current line to bring them more in line with these benchmarks. higher clocks faster memory perhaps the L3 that has been rumored. They don't have to be number one in this area but they do have to compete and they will probably be within 5% of the performance of Intel's top product and that should be good enough.

 

[dmens]

Timings are associated to the CPU clock cycle in which data is allowed to move to and from.

What are you talking about? The CPU or northbridge works with set memory constraints, not the other way around. What the hell is a "CPU clock cycle" ? The I/O doesn't go anywhere near as fast as the core clocks.

First of all, AMD has only said AM2 will support DDR 2. They did not say much about anything else. To even draw a conclusion on information you don't know much about is foolish.

With their timetable, they won't be able to do too much to their existing design. I don't know what they are doing exactly, but I can sure as hell make an educated estimate. It would only be foolish to speculate if you have absolutely no clue on what is involved (hiya).

 

[liebremx]

Originally posted by: Quinton McLeod
Link: http://www.blogger.com/comment.g?blogID=17375431&postID=114184911431790820

Straight from the mouth of Rahul Sood...

So you don't have any opinions of your own now that you realize you don't know what you're talking about? Rahul is only coming to those conclusions based on the BIOS changelog. Ask him whether he has seem the source code or read the BIOS developer's guide or talked to the guy that implemented those fixes.

1. Fix memory Set 2-1-1-1-1 and 4-1-1 mode wrong.
3. Change the description of DQDRV.
4. Fix Read Preamble Table Error.

All this are all memory related fixes not tied to any specific A64 stepping (take a look at the BIOS developer's guide yourself... if you can understand it).

2. Set Cool'n'Quiet default disable.

Whether is supported by the processor or not is decided by the BIOS through model+family NOT CARING about the FULL CPU signature). Being disabled by default has nothing to do with the model unknown issue in discussion here.
5. Shorten the delay time during clock programming loop.
6. Add over clocks step by step.

7. Fix fill 3114 SVID&SSID under Cross fire mode.

This I thought was the problem, missed the 3114 number (SIL SATA controller) so it's not related to the CPU nor video, nor PEG ports.. etc.

8. Fix soft-reset hang on POST code F2h if enable USB mouse support.
9. Change CMOS used to fix some control item can?t save.
11. Update SiI3112 Raid ROM.
12. Fix some SATA(DiamondMax 10 (6B160M0) HDD ) HDD detect fail at first time cool boot.

Irrelevant, unless they were using RAID in the setup and even if that was the case it turned out to make no difference in those benchmarks making lots of HD accesses.

10. Add support K8 FX60 CPU.

Yeah.. something like:

mov eax, 1
cpuid
...
A64:
and eax, 0xF ; Mask stepping
..
cmp eax, EQU_FX_60_STEPPING
jne unknown
mov eax, EQU_FX_60_SIGNATURE
jmp display
unknown:
mov eax, EQU_UNKOWN_SIGNATURE
display:
call display_the_darn_cpu_name_please ; No stack here, call is just used for illustrative purposes

hehehe..

I'm sorry to say, but that's a very important update for the FX-60! I bet you don't even know what half of those updates do.

Wait, let me ask Rahul, maybe he knows!

Bottom line sir, don't use adjectives unless you're pretty sure about the stuff you're talking about.

 

[TekDemon]

lol it's not a paper launch. What it really is, is just Intel serving notice to AMD (and AMD fanbois everywhere apparently) that their time on top is about to come to an end, because the king is on his way back.

And it's not like I haven't been cheering on AMD all this time. Heck, I wanted a K5 processor but there was lousy availability (which was what really hurt AMD back then).

But come on, we all knew this was going to happen-Intel isn't exactly some amatuer company.

Heck, the architecture in Conroe still has a lot in common with the P6 that was thought up years and years ago. Pentium Pro became Pentium II became Pentium III became Pentium M became Core Duo became Conroe. So it's pretty obvious that Intel knows plenty well how to design a badass CPU-they had just invested most of their engineering into IA-64 and got caught off-guard by AMD and their x86-64 push. So they did what they had to do-tried to band-aid the Pentium 4 to remain at least somewhat competitive with AMD while putting a lot more effort into retuning the technology they already had into a badass 64-bit x86 CPU.

So...this is just notice to AMD that it should be worried. Very worried. AMD stock holders seem to be fairly worried too since the stock pretty much dumped 10% when Intel showed off Conroe.

Now, if Intel actually manages to launch a 3.33Ghz XE part in 6 months it will be VERY hard for AMD to come up with something competitive by then. It'll be 50% faster than the overclocked FX-60 in the benchmarks, so unless AMD manages to do something really miraculous with AM2 they've basically just been told that they're about to get pwned.

Heh I wonder how the folks over at AMD are feeling now. It's probably a bit of an ego blow =p

 

[harobikes333]

TekDemon Nice post I'm still getting a 3800 X2 in two weeks though^ ^

 

[Keysplayr]

Originally posted by: plus
FX60 is a 1mb cache part at 2.8ghz.

X2 4600 is a 1/2mb cache part at 2.4ghz

If it's an FX60 overclocked, than it has 1mb cache.

They could overclock the X2 4600 to 2.8ghz, but it wouldn't run with the FX60.

I'm just trying to note what details are actually provided, since there are so few, and this test is so important to Intel.

Anand could clear all this up pretty easily by providing more detail.

Plus

And after Anand provides the detail you are looking for, you will find something else to crap on. So, it wouldn't do us any good because we would still have to listen to your bile.

 

[Keysplayr]

Originally posted by: plus
Were you guys concerned in the least that the bios couldn't recognice the FX60?

Probably normal right?

This probably has come up in a post after this, but Anand went back and Flashed the Bios on that RD480 mobo and it properly recognized the FX-60. Guess what. No difference. So, cool n quiet was not running. The second core WAS recognized. Conroe's memory timings were in fact 5-5-5 and they fixed it to 4-4-4 where they said it was in the first testing day. Conroe actually gained ground on the "FX-62" albeit very little. Anand made a mistake on the FEAR bench. They used 1024x768 on the Intel machine and 1280x960 on the FX62 machine. They fixed this the second day. Conroe still remained 20% faster than AMD FX-60 o/c'd to FX-62 speeds. To add insult to injury, this 2.66 Conroe is to be released as the performance "mainstream" part with higher clocked parts up to 3.0 GHz and an EE version at 3.33GHz. Unless AMD pulls an absolute miracle, they are FUBARED come July. They will still have terrific chips, but will have to slice their prices in half or better just to sell them.

 

[IntelUser2000]

You see, Rahul, is an idiot. He claimed on one of his blogs that Yonah is a 53W part and he claimed he proved it. He then takes Anand's numbers from the first Yonah review and since X2 system consumed 36W more than Yonah's system while X2 TDP is 89W, Yonah's TDP would be 89W-36W. WHAT AN IDIOT!!! We know that X2's don't reach 89W, more like 50-60W. Also know that 89W TDP means max and max will be really hard to reach. And some of you believe him!! Link: http://sharikou.blogspot.com/2005/11/intel-has-efficiency-problem-in-yonah.html

He also claims that Intel's confirms by saying Yonah's datasheet says its 67W TDP. Link claiming Intel's datasheets say its 67W: http://sharikou.blogspot.com/2006/02/intels-last-hope-marketing.html

According to Intel's datasheet, its 31W, saying it said it was 67W BY INTEL, immediately invalidates whatever he said.




Back to Core: Anand ASSUMES that it will be out in 6 months, since the rumors are Q3, and Q3 is September. But the most other rumors going around is that its been FORWARDED by AT LEAST A MONTH, which makes it July/August.

So we are looking at 4-5 months here. AM2 is supposed to come in June 6, which isn't really far away from Conroe....

Plus, the competitor Conroe 2.667GHz would be competing is the mainstream versions. If anyone of you thinks the 2.667GHz that costs $537 US is gonna compete with FX's, whatever, your loss. It will compete with the 2.667GHz Athlon X2's which is lower than the 2.8GHz AT tested. Hmm.

The REAL competition of FX's would be the 3GHz and 3.33GHz EE's with 1333MHz FSB, and the gap is gonna be as wide.

BTW, only 2.4GHz and 2.667GHz Conroe is 1066MHz bus, the lower speed grades are 800...

 

[Viditor]

Originally posted by: keysplayr2003


This probably has come up in a post after this, but Anand went back and Flashed the Bios on that RD480 mobo and it properly recognized the FX-60. Guess what. No difference. So, cool n quiet was not running. The second core WAS recognized. Conroe's memory timings were in fact 5-5-5 and they fixed it to 4-4-4 where they said it was in the first testing day. Conroe actually gained ground on the "FX-62" albeit very little. Anand made a mistake on the FEAR bench. They used 1024x768 on the Intel machine and 1280x960 on the FX62 machine. They fixed this the second day. Conroe still remained 20% faster than AMD FX-60 o/c'd to FX-62 speeds. To add insult to injury, this 2.66 Conroe is to be released as the performance "mainstream" part with higher clocked parts up to 3.0 GHz and an EE version at 3.33GHz. Unless AMD pulls an absolute miracle, they are FUBARED come July. They will still have terrific chips, but will have to slice their prices in half or better just to sell them.

All of the talk about AMD being screwed unless they can match Conroe at launch seems as ridiculous to me. It's just as ridiculous as Intel being screwed (or needing to cut prices in half) when the A64 and X2 absolutely smeared them for performance for years.

What we know
1. We know that AMD has their own Nextgen chip coming out within a year
2. We know that AMD begins volume production of 65nm in August
3. AMD still has the advantage in enterprise platforms because of HT and MOESI protocol
4. AMD's power usage will drop significantly, even on 90nm

What we don't know
1. We don't know what yields/volume Intel will be capable of on Conroe
2. We don't know what the finished production Conroe will perform like (it could be better or it could have another of Intel's infamous "bugs")
3. We don't know if/when AMD will introduce Z-ram cache (which allows for 5 times the cache in the same space and with even lower power usage)
4. We don't know what tweaks AMD is making on the memory controller (though we know they're making them)
5. We don't know when Intel will be able to produce a 3GHz NGA chip
6. We don't know how well NGA performs on 64 bit

Props to Intel for fulfilling their promise of making the NGMA an excellent performing desktop...they really needed it badly, and most sane people expected no less! But, it won't shift the marketshare numbers for quite awhile (it never does) except that it will almost certainly stop the bleeding.
While many people are frustrated that AMD has kept absolutely quiet on what's coming, it's a smart strategy for them and there really is no downside. It's not like people are going to decide that they have to wait 6 months to buy a Conroe (with the exception of a few enthusiasts).

For you Intel Fanbois out there, I can only imagine the party you must be having right now after all this time...enjoy, you've earned it! Just remember to hand someone else your car keys before you pop the Champagne!

For the AMD fanbois, don't despair that you don't have a party of your own right now...it might be a surprise party, or it just might be scheduled for later. However, if you don't see an invite by Q2 07, I'd start to get nervous...

 

[openwheelformula1]

I am suprised there isn't much talk about AMD's K8L, the product before K10 arrives. It's going to double Floating Point Units. I wonder if AMD can impliment 4 wide (or even more) like Conroe. Anyway AMD is very tight lipped currently as they should since they are getting 80% of the retail market. With all the Conroe excitement, I think there maybe more surprises later. (As it always has been)

 

[pedramrezai]

The way most of the people are thinking is just the way Intel has planned. I am a AMD fan but I can realize competition is good for customers. I was shocked by Core performance but using a handicapped AMD system really annoyed me. First, reviewers have already proven RD580 or solutions with dual 16x can deliver up to 10-15% more performance when paired with high-end, bandwith hungry vga cards. Second, we have been hearing of dual core optimizations in display drivers for some time but were unable to see something significant until we saw Conroe performance; I am quite suspicious over some hefty optimizations in intel-cooked display driver. Time will reveal. Third, this might be the beginning of a new SSEx game with unfair optimizations for a new technoogy.
I am surprized how people are trashing the current as well as future AMD64 technology.But remember that Core is not out yet and all these might be some optimizations that has granted it this performance level. Moreover, the current AMD64 technology is almost 3 years old and the new AM2 will update its specs. AMD did not like DDR2 high latency; What they are looking for is its higher frequency that can be paired with the new AM2 FSB.For Athlon 64 and Sempron a 333mhz FSB that paires with DDR2 666 and for the Fx parts a 400 mhz FSB pairing with DDR2 800. If DDR1 could reach these frequencies you could now see the real potential of AMD64. This kind of bandwith will give Core a hard time. Also remember that AMD is increasing cache (L2 and maybe L3). Shared cache is also something that will be seen in the future products and will bring huge performance gains. Based on the preliminary data of 200/266 async single channel bandwith of 3500mb/s a memory bandwith of >10k is expected in the final product and if Intel was going to compare its future platform, it was not fare to compare it with an infrastructure of >2 years old. I am sure the new AM2 will regain AMD reputation once again. But we all must remember that this competition between major players is good for the end users.

 

[Fox5]

Originally posted by: TekDemon
lol it's not a paper launch. What it really is, is just Intel serving notice to AMD (and AMD fanbois everywhere apparently) that their time on top is about to come to an end, because the king is on his way back.

And it's not like I haven't been cheering on AMD all this time. Heck, I wanted a K5 processor but there was lousy availability (which was what really hurt AMD back then).

But come on, we all knew this was going to happen-Intel isn't exactly some amatuer company.

Heck, the architecture in Conroe still has a lot in common with the P6 that was thought up years and years ago. Pentium Pro became Pentium II became Pentium III became Pentium M became Core Duo became Conroe. So it's pretty obvious that Intel knows plenty well how to design a badass CPU-they had just invested most of their engineering into IA-64 and got caught off-guard by AMD and their x86-64 push. So they did what they had to do-tried to band-aid the Pentium 4 to remain at least somewhat competitive with AMD while putting a lot more effort into retuning the technology they already had into a badass 64-bit x86 CPU.

So...this is just notice to AMD that it should be worried. Very worried. AMD stock holders seem to be fairly worried too since the stock pretty much dumped 10% when Intel showed off Conroe.

Now, if Intel actually manages to launch a 3.33Ghz XE part in 6 months it will be VERY hard for AMD to come up with something competitive by then. It'll be 50% faster than the overclocked FX-60 in the benchmarks, so unless AMD manages to do something really miraculous with AM2 they've basically just been told that they're about to get pwned.

Heh I wonder how the folks over at AMD are feeling now. It's probably a bit of an ego blow =p

Oh come on, we can hardly still call Conroe an upgraded P3, otherwise Athlon is an upgraded K6.
Going from P3 to Pentium M or Pentium M to Core Duo had more major changes than going from Pentium Pro to Pentium 3. The most major change the P6 core saw didn't even warrant a new model number (though fundamentally integrating the cache into the core marked going from a P2 called a P3 to an actual P3). Conroe is definetely a next gen architecture, inferior in no way to any current Intel architecture and superior in many. And I'd say it borrows more from P4 than P3 anyway, except for that it isn't focusing quite as much on clock speeds, but this is definetely still a clock speed focused architecture, it's no 1.5Ghz Itanium or Power5.

Oh, and Itanium probably didn't hurt Intel that much, the Pentium 4 was definetely a relevant architecture and a true focus on extracting the maximum possible performance out of a processor. Plus, a lot of the innovations it required to have such great performance carried over to Conroe and made it as good as it is. Pentium 4's only problems are running out of steam before expected and being an outdated architecture, the latter of which is something that every architecture eventually encounters and there is rarely a replacement ready in time.

 

[kdrp]

So if a performance preview qualifies as a paper launch what is this then?


http://www.tomshardware.com/2006/02/21/a_look_at_amds_socket_am2_platform/

 

[Zebo]

What it really is, is just Intel serving notice to AMD (and AMD fanbois everywhere apparently) that their time on top is about to come to an end, because the king is on his way back.

LOL ...not counting the 8 months AMD had no answer to Northwood C.. how can you call Intel the "king" when AMD has lead in performance for 6 years now?

I do agree with you though..brilliant move on intels part.. They basically said "hey AMD we are going to kick your ass" and "people don't buy AMD X2's these next six months, lookie what we gots" and that marketing dweeb at AMD has said nothing to indicate AMD has any real answers come conroe just banal bamboozlements out of his mouth.

Viditor is dreaming. AMD can respond w/o giving any deatils... If AMD had the goods, all they had to say is "No, they won't" as intel just showed them up.. Thier silence speaks volumes.

 

[Keysplayr]

Originally posted by: Zebo

What it really is, is just Intel serving notice to AMD (and AMD fanbois everywhere apparently) that their time on top is about to come to an end, because the king is on his way back.

LOL ...not counting the 8 months AMD had no answer to Northwood C.. how can you call Intel the "king" when AMD has lead in performance for 6 years now?

I do agree with you though..brilliant move on intels part.. They basically said "hey AMD we are going to kick your ass" and "people don't buy AMD X2's these next six months, lookie what we gots" and that marketing dweeb at AMD has said nothing to indicate AMD has any real answers come conroe just banal bamboozlements out of his mouth.

Viditor is dreaming. AMD can respond w/o giving any deatils... If AMD had the goods, all they had to say is "No, they won't" as intel just showed them up.. Thier silence speaks volumes.

Six years? I'm not so sure about that.

And I'm not sure that AMD silence is speaking anything this time around.

 

[HotkeyCC]

I agree that the silence means nothing. You never know what AMD has in the works, and they have so much on the backburner that even if conroe does totally kill AMD, it will probably be for a maximum of 6 months before AMD takes its rightful place back on top. I for one am not going to buy into intel's PR and put off my computer purchases, because it doesn't seem worth it to me. I guess we'll just see how this whole thing plays out. The only thing I'm genuinely worried about as an AMD fanboy is AM2. I hope they have a few more improvements than theyre letting us know about, so it will at least be a viable chipset, not just used as a price reduction for 939.

 

[Keysplayr]

Originally posted by: HotkeyCC
I agree that the silence means nothing. You never know what AMD has in the works, and they have so much on the backburner that even if conroe does totally kill AMD, it will probably be for a maximum of 6 months before AMD takes its rightful place back on top. I for one am not going to buy into intel's PR and put off my computer purchases, because it doesn't seem worth it to me. I guess we'll just see how this whole thing plays out. The only thing I'm genuinely worried about as an AMD fanboy is AM2. I hope they have a few more improvements than theyre letting us know about, so it will at least be a viable chipset, not just used as a price reduction for 939.

Well, as far as AM2 goes, they have DDR2 support, and oh hey!! A new HSF!!
AMD takes it's "rightful" place? hehe. Methinks that neither company can claim that right.
Intel seems to mean business again, so AMD will surely try and fire back at it's earliest convenience. So the term rightful does not apply. They will one up each other, unless of course we find that Intel is dead set on keeping it's foot on AMD's neck with appropriate pressure. They have always had the resource to do this, it's just the will that was lacking, I think.

 

[imported_Seer]

Originally posted by: darkdemyze

Originally posted by: deeznuts
Just theoretically though. Does this make sense? That DDR2 plays to the weakness of the A64 processor?

Can you explain how else there's such a cut in performance?

Are you trying to say that memory doesnt have anything to do with performance? If that's the case then why do we even have 184-pin DIMMs in the first place? Why not just stick with 168-pin DDR133 sticks and just keep lowering latencies

Also, is that why Intel switched over to DDR2 cause I can see how much it's hurting them..

DDR Latencies are already almost as low as possible (2-2-2). The only thing faster would be 1-1-1. Or somehow inventing RAM that had an instantaneous response. why would you lower bandwidth by such huge amounts to reduce the latencies by one clock cycle? That is, if older ram can even run at latencies of 1, which I doubt it can