Sign in with
Sign up | Sign in
Your question

AMD HAS COST TECHNOLOGY ADVANTAGES OVER INTEL

Last response: in CPUs
Share
March 15, 2006 4:41:54 PM

http://www.forbes.com/markets/emergingmarkets/2006/03/14/intel-amd-0314markets07.html

Quote:
The analyst estimates servers will comprise 30% to 40% of Intel revenues and 50% of profits, with AMD respectively at 40% to 50% and 50%.

More about : amd cost technology advantages intel

March 15, 2006 5:28:11 PM

Quote:
http://www.forbes.com/markets/emergingmarkets/2006/03/14/intel-amd-0314markets07.html

The analyst estimates servers will comprise 30% to 40% of Intel revenues and 50% of profits, with AMD respectively at 40% to 50% and 50%.


I hate to do this as I do like my AMDs (I don't own an Intel system)

Your interpretation is backwards.

If Something counts to 30%-40% of my revenue, but 50% of my profits
And your competitors count for 40%-50% of revenue, and 50% of profits.

The competitor has a more costly production process.

The statement you quoted clearly shows the Intel processors the cheaper of the 2 peices of silicon to produce.

Boo to that but hey you can't argue the numbers.
March 15, 2006 5:31:02 PM

yep your correct. blind fanboyism does nothing when the fanboy can't even see the obvious.
Related resources
March 15, 2006 5:38:29 PM

Tsk Tsk Tsk

Should have paid more attention to detail man. Penguin is correct. You have to keep in mind that producing 65nm is cheaper for Intel so, of course it only costs less to manufacture their chips. Again, one of the advantages for Intel and switching to 65nm technology.

Um, made correction for PenGuin. Sorry dude! LOL :oops: 
March 15, 2006 5:41:50 PM

Quote:
Penquin


excuse me sir...

Penguin

:p 
(Please realize this post is in jest, irony on attention to detail :p [/b])
March 15, 2006 6:12:04 PM

Quote:
http://www.forbes.com/markets/emergingmarkets/2006/03/14/intel-amd-0314markets07.html

The analyst estimates servers will comprise 30% to 40% of Intel revenues and 50% of profits, with AMD respectively at 40% to 50% and 50%.


I hate to do this as I do like my AMDs (I don't own an Intel system)

Your interpretation is backwards.

If Something counts to 30%-40% of my revenue, but 50% of my profits
And your competitors count for 40%-50% of revenue, and 50% of profits.

The competitor has a more costly production process.

The statement you quoted clearly shows the Intel processors the cheaper of the 2 peices of silicon to produce.

Boo to that but hey you can't argue the numbers.

Well, a true fanboy would look at those numbers and say that the rest of AMDs product line is more profitable than Intel's non-server.

But your argument is actually flawed. For instance, if the rest of AMDs product line had a 1% profit margin, but equal revenue to their server chips, then server chips would account for something like 90% of profits. Has their manufacturing tech changed? Nope. These are not comparative measures
March 15, 2006 6:25:27 PM

Quote:
http://www.forbes.com/markets/emergingmarkets/2006/03/14/intel-amd-0314markets07.html

The analyst estimates servers will comprise 30% to 40% of Intel revenues and 50% of profits, with AMD respectively at 40% to 50% and 50%.

Well, this article is pretty much retarded. Intel has key price advantages in certain markets, such as the cheapest dual cores, the PD805 & PD820. Bang for buck however, AMD absolutely dominates.
-------
Conroe will change all of this; Intel is supposed to drop prices 50% on some Pentium D's, I expect AMD to do the same on FX, AMD64's and X2's...can anybody say PRICE WARS?
Not only that, but some value Conroe's should be coming out...hopefully same with AM2.
March 15, 2006 7:58:10 PM

Get a new keyboard!
March 15, 2006 8:01:05 PM

Action man is here to save the day!

We've come for your caps lock key chuck! LOL
March 15, 2006 8:58:39 PM

Quote:
http://www.forbes.com/markets/emergingmarkets/2006/03/14/intel-amd-0314markets07.html

The analyst estimates servers will comprise 30% to 40% of Intel revenues and 50% of profits, with AMD respectively at 40% to 50% and 50%.


I hate to do this as I do like my AMDs (I don't own an Intel system)

Your interpretation is backwards.

If Something counts to 30%-40% of my revenue, but 50% of my profits
And your competitors count for 40%-50% of revenue, and 50% of profits.

The competitor has a more costly production process.

The statement you quoted clearly shows the Intel processors the cheaper of the 2 peices of silicon to produce.

Boo to that but hey you can't argue the numbers.

Wouldn't your analysis mean that the rest of Intel's product line is less profitable than servers, taking "revenues" as being sales? For instance, if Intel sells ~33% servers and makes half of all profits on this, that means the other ~67% is responsible for the other 50%, thus meaning these other sectors are less efficient (i.e., less profitable). There may be many reasons for this lack of profitability, including inefficiency of production processes, marketing costs per unit, etc. AMD seems to have better balance, with ~50% of total sales producing half of their total profit.
March 15, 2006 9:56:10 PM

Ahh yes, this is true...

However, I wasn't going to do the job of an analyst, I was just pointing out that is immeadiate jump to conclusions was faulty.

If you really wish to delve into it, the reason they pull this off is that the profit margins on intel server chips is higher than the PC chips. it costs roughly the same to produce either chip.

But since AMD has had the dark cloud of "stability issues" hanging over it, Intel has been able to maintain a higher margin on it's server chips due to "better reliability".

Anyone who has a head on their shoulders knows that AMD's stability issues are about as prevalent as condors nowadays.

But it's amazing what one big screw up (K6 and related processors) can do to the image of a comapny.
March 16, 2006 12:15:20 AM

Quote:
Ahh yes, this is true...

However, I wasn't going to do the job of an analyst, I was just pointing out that is immeadiate jump to conclusions was faulty.

If you really wish to delve into it, the reason they pull this off is that the profit margins on intel server chips is higher than the PC chips. it costs roughly the same to produce either chip.

But since AMD has had the dark cloud of "stability issues" hanging over it, Intel has been able to maintain a higher margin on it's server chips due to "better reliability".

Anyone who has a head on their shoulders knows that AMD's stability issues are about as prevalent as condors nowadays.

But it's amazing what one big screw up (K6 and related processors) can do to the image of a comapny.

I absolutely disagree. One big screw up can be made up very quickly with a competitive line of powerful chips next time around. Look at Nvidia. People though they were cooked after the release of the 5 series! Then take a look at the 6 series...absolutely dominated, and they totally revived themselves.
The point is, even if you screw up, you can still regain the lost ground with a powerful release. Only complete morons would think of the K6 when they think of AMD, and if you want to talk about screw-ups, people will look at the P4 Prescott when they think "Intel", and thus half the people on this forum blindly believe any Intel runs extremely hot.
March 16, 2006 12:20:12 AM

Quote:
Then take a look at the 6 series...absolutely dominated


I think that's a bit of an exaggeration.
March 16, 2006 12:40:51 AM

Okay, overstated. But still, 6 series was extremely successful. Think of all the 6800gt's and 6600gt's that sold...crazy.
March 16, 2006 12:45:45 AM

Indeed it was extremely successful.
March 16, 2006 12:45:51 AM

Quote:
he meant pentiums duh

65nm Pentium 4's are actually cooler than A64', I hate to break it to you.
March 16, 2006 1:20:07 AM

Quote:
65nm Pentium 4's are actually cooler than A64', I hate to break it to you


Where in the world a P4 on 65nm is cooler than AMD's Athlon 64 on 90nm??

Show me a single bench that will back up your FUD, since all of them tend to contradict you. :wink:
March 16, 2006 1:30:39 AM

Quote:
65nm Pentium 4's are actually cooler than A64', I hate to break it to you


Where in the world a P4 on 65nm is cooler than AMD's Athlon 64 on 90nm??

Show me a single bench that will back up your FUD, since all of them tend to contradict you. :wink:
I hate to say this, but it's basically impossible to find a review of the 65nm single core chips on google. Anandtech did one, but they didn't mention temperatures =(
All I can say is, people using the stock Intel cooler are reporting very low temperatures on stock speed/voltage.
March 16, 2006 1:35:46 AM

Quote:
I hate to say this, but it's basically impossible to find a review of the 65nm single core chips on google. Anandtech did one, but they didn't mention temperatures =(
All I can say is, people using the stock Intel cooler are reporting very low temperatures on stock speed/voltage.


The same thing goes for AMD since the Venice core was introduced. :wink:
March 16, 2006 1:39:31 AM

Quote:
I hate to say this, but it's basically impossible to find a review of the 65nm single core chips on google. Anandtech did one, but they didn't mention temperatures =(
All I can say is, people using the stock Intel cooler are reporting very low temperatures on stock speed/voltage.


The same thing goes for AMD since the Venice core was introduced. :wink:
Point is, if Intel wasn't retarded and released 65nm Netburst chips 3 years ago, we would've never ran into the heat fiasco.
March 16, 2006 1:50:18 AM

Since you couldn't post any benchmarks, I'll do it for'ya:
http://www.gamepc.com/labs/view_content.asp?id=pd900&page=4&cookie%5Ftest=1
Quote:
Nevertheless, these numbers are greatly improved over the Pentium-D 800 series, which typically utilized over 300W when the processors were maxed out, in comparison to the ~235-245W numbers we're seeing with the 900 series. Impressive for Intel, but AMD is still the leader here. AMD's Athlon64 X2 chips still run cooler and consume less power compared to the Intel Pentium-D 900 series.



Quote:
Point is, if Intel wasn't retarded and released 65nm Netburst chips 3 years ago, we would've never ran into the heat fiasco.

Yes. You have a point there.
AMD suffered the same problem with the Througbred processor. If they knowed about SOI before, the Athlon XP core would've scaled a lot beter than it did.
March 16, 2006 2:06:05 AM

Quote:
Since you couldn't post any benchmarks, I'll do it for'ya:
http://www.gamepc.com/labs/view_content.asp?id=pd900&page=4&cookie%5Ftest=1
Nevertheless, these numbers are greatly improved over the Pentium-D 800 series, which typically utilized over 300W when the processors were maxed out, in comparison to the ~235-245W numbers we're seeing with the 900 series. Impressive for Intel, but AMD is still the leader here. AMD's Athlon64 X2 chips still run cooler and consume less power compared to the Intel Pentium-D 900 series.



Quote:
Point is, if Intel wasn't retarded and released 65nm Netburst chips 3 years ago, we would've never ran into the heat fiasco.

Yes. You have a point there.
AMD suffered the same problem with the Througbred processor. If they knowed about SOI before the Athlon XP core would've scaled a lot beter than it did.
I'm not a fan of SOI, leaks more than Strained Silicon and it has cold boot problems...note no Intel chips have that cuz Intel uses Strained Silicon on all chips, while AMD only does on FX chips. K8 is what saved AMD.
March 16, 2006 2:15:01 AM

Quote:
I'm not a fan of SOI, leaks more than Strained Silicon and it has cold boot problems...


Well, it's far superior to Intel's strained silicon since it gives more performance while lowering power consumption. I haven't seing anything about cold boot problems with it.
March 16, 2006 2:41:13 AM

I'm surprised you haven't heard about the "cold bug". I'm sure someone with more AMD overclocking experience can tell you more, but it's supposedly a problem with SOI and the memory controller which doesn't run stable under 0C which prevents extreme overclocking. I'm not sure if it's still around, because it isn't publicised as much, but it might explain why some chips overclock better than others.

Anyways, on another note, I'd love to hear your take on AM2 only offering about a 3-5% performance increase over S939 using DDR2 800. AMD said those are final revisions so that's pretty much what your getting. Even if the final figures double that and show a 10% increase the FX-62 would still be 10% slower than the 2.67GHz Conroe. Admittedly those Intel provided systems may not be the most accurate, but the difference seems very large, especially with whatever Extreme Edition Intel is planning sitting in the background.

http://www.theinquirer.net/?article=30290
March 16, 2006 2:51:02 AM

I know you've probably answered this before but I wanted to ask you. What is your take on AMDs 90nm cache size factor and how far can AMD go in your opinion, with their current technology?

I've read many times about this and it seems to be an ongoing issue. I apologize in advance if this appears to be a repetitive question however, you seem very knowledgable.

Thank you in advance and keep up the great posts. I thoroughly enjoy reading them. :wink:
March 16, 2006 2:53:57 AM

Quote:
I'm not a fan of SOI, leaks more than Strained Silicon and it has cold boot problems...


Well, it's far superior to Intel's strained silicon since it gives more performance while lowering power consumption. I haven't seing anything about cold boot problems with it.
Well, it definately does make less energy problems. However, at -10c, some SOI processors will not boot at all! Freaking gay if you ask me. Also, you get less out of overclocks and clock speeds since it really does leak more speed than Strained Silicon.
March 16, 2006 2:57:31 AM

Quote:
Explain please? The current data does not support your statement.


OK, here you go:

Digitimes

Digitimes again...

Quote:
Soitec claims significant advantages for SOI. It offers greater performance at significantly lower rates of power consumption. That translates into faster processing while producing less heat. That also translates into longer battery life for mobile devices. Charge leakage is reduced to a minimum as also are transistor cross-talk and latch-up, allowing greater transistor proximity. As designs move ever further into deep submicron, these are all significant advantages.
March 16, 2006 2:58:04 AM

Quote:
Point is, if Intel wasn't retarded and released 65nm Netburst chips 3 years ago, we would've never ran into the heat fiasco.


This is not actually true, the density of transistors that are on or off scale linearly with transitor count and as you decrease litho node, gate lengths and thickness scales too. Both AMD and Intel have approached gates so thin, than quantum mechanical tunneling is driving high gate leakage currents. Simply doing 65 nm over 90 nm does not buy you that back that much in this regard -- there are other factors though, but don't count on SOI to be the end all beat all.

Case in point, the current Pressler and Cedarmill chips are essentially Prescott cores and they are at around the same TDP, slightly lower but nothing earth shattering and at similar speeds.

That's not totally true. The 65nm chips Intel released are far more capable and power efficient than the 90nm chips, and do not have the same massive power problems. Sure, the architecture itself isn't fantastic, but if you compare the overclocks you can get on a Prescott stock cooler versus a Cedar Mill same conditions I think it serves my point. The reason Intel doesn't release them at 4.0ghz is basically because on the Preslers, heat from 2 high-frequency cores is still very difficult to control, albeit better than the Smithfields. Intel said it is a very real possibility they will release 4+ ghz Cedar Mill chips however; they certainly can with the current 65nm technology. Notice Anandtech's article on the Cedar Mill's release; a 661 was overclocked to 4.5 on stock cooling without a massive voltage increase and uncontrollable heat you would need on a Prescott.
March 16, 2006 3:00:41 AM

Quote:
Explain please? The current data does not support your statement.


OK, here you go:

Digitimes

Digitimes again...

Quote:
Soitec claims significant advantages for SOI. It offers greater performance at significantly lower rates of power consumption. That translates into faster processing while producing less heat. That also translates into longer battery life for mobile devices. Charge leakage is reduced to a minimum as also are transistor cross-talk and latch-up, allowing greater transistor proximity. As designs move ever further into deep submicron, these are all significant advantages.

Let me ask you something: why do you think AMD uses Strained Silicon on its FX series? They know power consumption isn't a problem here; and they want people to get the best possible overclocks. All the AMD record setters for superpi/3dmark are on LN2 cooled FX55's/FX57'S with speeds that are absolutely impossible for an A64.
March 16, 2006 3:16:19 AM

Quote:
Anyways, on another note, I'd love to hear your take on AM2 only offering about a 3-5% performance increase over S939 using DDR2 800.


Even if it's a 5% gain with DDR2-800, the process itself would increase the processor performance to 25% over socket 939 processors.
As I said before, I'd wait for the thing to be out and fully benched by everyone.
March 16, 2006 3:19:25 AM

Quote:
Anyways, on another note, I'd love to hear your take on AM2 only offering about a 3-5% performance increase over S939 using DDR2 800.


Even if it's a 5% gain with DDR2-800, the process itself would increase the processor performance to 25% over socket 939 processors.
As I said before, I'd wait for the thing to be out and fully benched by everyone.
With all respect, that 3-5% increase is for real, even for the processor. If you sit down and think about it, the only thing innovative about AM2 is DDR2, just a re-release of 939. Don't get me wrong, I haven't benched the thing or anything, but honestly, how can a little memory bandwidth lead to far better performance? AMD needs a lot more than that, and I really hope it is more than just 939 plus a pin plus DDR2.
March 16, 2006 3:31:38 AM

Quote:
With all respect, that 3-5% increase is for real, even for the processor. If you sit down and think about it, the only thing innovative about AM2 is DDR2, just a re-release of 939.



ak47is1337 :
How many times do I have to tell you that it's not only DDR2 that will help the processor but it's the new and improved process that will give a theoretical 20% performance gain. DDR2-800 will only give AMD a 5-10% increase in performance, but once again, the process is what's going to help these processors to achieve a 20% more performance.

Did I make my self clear this time?? :wink:
March 16, 2006 3:37:54 AM

Quote:
http://www.forbes.com/markets/emergingmarkets/2006/03/14/intel-amd-0314markets07.html

The analyst estimates servers will comprise 30% to 40% of Intel revenues and 50% of profits, with AMD respectively at 40% to 50% and 50%.

Well, this article is pretty much retarded. Intel has key price advantages in certain markets, such as the cheapest dual cores, the PD805 & PD820. Bang for buck however, AMD absolutely dominates.
-------
Conroe will change all of this; Intel is supposed to drop prices 50% on some Pentium D's, I expect AMD to do the same on FX, AMD64's and X2's...can anybody say PRICE WARS?
Not only that, but some value Conroe's should be coming out...hopefully same with AM2.

To be more precise, AMD currently onlly dominates in the bang/buck catagory in gaming. If you are more audio/video inclined, Intel is the better buy.
March 16, 2006 11:29:04 AM

Quote:
With all respect, that 3-5% increase is for real, even for the processor. If you sit down and think about it, the only thing innovative about AM2 is DDR2, just a re-release of 939.



ak47is1337 :
How many times do I have to tell you that it's not only DDR2 that will help the processor but it's the new and improved process that will give a theoretical 20% performance gain. DDR2-800 will only give AMD a 5-10% increase in performance, but once again, the process is what's going to help these processors to achieve a 20% more performance.

Did I make my self clear this time?? :wink:
What "process"? They aren't making 65nm till 2007, the are still using the same K8 overall architecture and their 90nm chips cannot be scaled well at all anymore given that they will top out at 3ghz. I have a feeling AM2 will seriously disappoint, becuase once AMD finally releases its chips with 65nm, Intel will already have 45nm in place =/
March 16, 2006 2:51:09 PM

Quote:
What is your take on AMDs 90nm cache size factor and how far can AMD go in your opinion, with their current technology?

Hmm, I probably need to be careful with my tone to avoid antagonizing people further. The questions you ask are the million dollar questions of course.

I regards to cache sizes, I've made these comments many times before, but sadly I can never find the previous threads where I made them. In any case, larger L2 cache sizes really don't benefit the K8 architecture because of how AMD designed it. The K8 uses an exclusive architecture which means that the data in the L1 cache isn't duplicated in the L2 cache. As I understand it, this was more critical before because it offered the flexibility that the L2 cache did not need to be larger than the L1 cache and could be any size and it increase the total cache size since the 2 can be added together.

The other method of course is inclusive cache which Intel uses, which duplicates the L1 cache in the L2. In an exclusive cache, when an L2 cache line is copied into the L1 cache, an existing line in the L1 cache needs to be copied out into the L2 cache to make run for it. This of course adds latency and as a point of detail a victim buffer is used to increase performance. In an inclusive cache, when an L2 cache line is moved into the L1 cache, an existing L1 cache line doesn't need to be written to L2 because it is already duplicated there. This of course increases performance.

Since an exclusive cache puts more emphasis on the size on the L1 cache and less on the L2 cache, increasing the L2 cache size on an exclusive architecture doesn't increase performance as much as an inclusive cache. Even disregarding process differences, the reason why AMD doesn't have large L2 caches is, because they generally don't see as much benefits. Intel of course uses large L2 caches not only to reduce being bandwidth starved, but because they architecture benefits more. The thing that Intel needs to worry about with the inclusive cache is that they have the correct ratios between the L1 cache size and the L2 cache size. Since the L1 cache is copied to the L2 cache, the L2 needs to be a sufficient size to be of any use. Intel of course messed this up with the 128KB L2 cache in the Northwood Celerons, which was far too small and crippled their performance. The current 256KB Prescett Celeron Ds perform much better in comparison despite the pipeline increase.

As to how much cache you really need, there's of course a point of dimishing returns. The 1MB L2 cache per core in the K8 should be sufficient for desktop usage although AMD could use a shared cache in order to avoid duplication and maximize cache space. I suppose if AMD doesn't move to a shared cache, doubling the L2 could beneficial to hide any latency associated with the move to DDR2 although as you move to DDR2 800 and beyond the concerns really aren't there anymore. On desktops, I'm not sure if an L3 cache would be beneficial since with an OMC, the latency to the RAM is already low so if the L3 has higher latency than the L2 the benefit becomes increasingly moot. Servers of course love larger amounts of L2 and L3 cache so it would be useful there.

When you say how far can AMD go with their current technology I'm assuming you mean their 90nm process. They appear to be introducing some form of FD-SOI into their Socket F chips so they will be stretching the process a little further. I still think the big gains will come with 65nm. They've actually been tweeking their designs with the L2 cache in the new AM2's actually taking up less space on the same 90nm process than on the S939s. Revisions and advancements in the SOI process will obviously bring power and heat lower. However, that doesn't address the cost issue. It can be argued that Intel's 65nm process may not be advanced as AMD's 90nm SOI process, but it's still cheaper for Intel to produce. As long as AMD remains on 90nm they will be hampered in how they price their products. AMD may be able to produce a quad core on 90nm with the right thermals, but the size will make it problematic.

On the current 90nm process, I really don't see AMD increasing cache sizes, at least not in the desktop or mobile market since that's not really necessary or cost-effective. It should be noted that the Turion X2s that will be released all have 512KB of L2 cache per core rather than 1MB. With the FX-62 clocking in at 2.8GHz coming in June, AMD could probably sustain 1 more clock ramp to 3GHz, but after that it'll be next year and hopefully 65nm.

Will AMD be able to beat Conroe when it launches? Personally, I don't see it as very likely. Intel's new architecture appears superior to the K8 on a per clock basis which means Intel will probably have the lead in H2 2006. Improvements in AMD's process might shrink the gap, but I haven't seen evidence of their newer processes yet. Afterall, the TDP of the FX has actually increased and now the Opteron HE TDP has increased as well to 68W. AMD did say they would eventually have an Opteron EE version at 55W, but with Socket F already launching in H2, eventually will no doubt put it into 2007 and 65nm.

http://www.theinquirer.net/?article=30349

Interestingly, AMD has said that Socket F will use the exact same process as current Opterons and not FD-SOI as originally thought.

http://www.itjungle.com/breaking/bn031506-story01.html

Quote:
First of all, these processors, which are due in the third quarter of this year, not the middle of the year, are based on the same 90 nanometer process that the current Opterons use.

In any case, AMD will certainly come out kicking in 2007 when their new K8L architecture comes out along with 65nm.

This is just my opinion of course, and I'm sure many people will disagree, but you asked for it. I'm sure MadModMike will say this is just google information, which I'm sure you can find there if you wanted.
March 16, 2006 5:05:04 PM

I'd just like to add that AMD does benefit from larger caches. Greater L1 cache in the K7-K8 core does improve perfromance a lot just as Intel benefits from larger L2 cache.

If I'm not wrong, AMD will increase the L1 cache in their next-gen processor architecture and they'll enhance cache hierarchy also.
March 16, 2006 11:19:54 PM

At least that's one point you and I can agree on. The L1 cache of the AMD K6 line of processors was double that of Intel's Pentium, and during the k7 and k8 series it was quadruple, then 8 times, then back to quadruple to that of intel's PIII, P4 williamette and Northwood, and Prescott respectively. The Intel design traded L1 size for low latency in the Pentium 4. It works well in some situations, but in retrospect AMD's "larger is better" approach seems to have been much more beneficial overall.

EDIT: Itcommander data, I must give you props on your previous post. Very well written and highly informative.
March 18, 2006 6:27:57 AM

Empirical data seems to suggest that you opinion of SOI may be in error.
While the A64 added a graet many transistors to xp, the layout stayed fairly close. The key change was SOI. The A64s run considerably cooler.
When the A64s went to 90 nanos, there was a slight increase in TDP.
The venus added an improved SOI, and got lower TDP.
Truth is, Intel could make very good use of SOI (and, I understand will, on the 45 nano chips), but they would rather thier customers enjoy a warmer environment, rather than loose profit margin.
March 18, 2006 6:52:10 AM

You should try to not be blinded by your preference for Intel.
Whether larger cache will be usefull is entirely dependant on the size of exe files. As chips get larger caches, progs tend to require , and take more benefit from them.
Quote:
Since an exclusive cache puts more emphasis on the size on the L1 cache and less on the L2 cache, increasing the L2 cache size on an exclusive architecture doesn't increase performance as much as an inclusive cache. Even disregarding process differences, the reason why AMD doesn't have large L2 caches is, because they generally don't see as much benefits.

I always thought that Intel needed larger line 2 cache because thier L2 was reduced due to holding a copy of L1, and because thier L1 was so much smaller. Total cache equals L2 cache, while Amd has a larger L1 cache, that is actually adative to L2, for a much larger total cache.
Big problem for inclusive cache stems from a L1 call, from memory. In this instance, bothe L1 and L2 need to be purged of a space equal to the call size. Much more of a penalty than
Quote:
In an exclusive cache, when an L2 cache line is copied into the L1 cache, an existing line in the L1 cache needs to be copied out into the L2 cache to make run for it.

I really dont see any advantage to wasting the best memory in a system by having info doubled or tripled. It's just a waste.
March 18, 2006 9:04:25 AM

LOL you're one funny character. Put your glasses on and read:
"WE HAVE seen many AM2 socket CPUs. Most of them are running at 2.4GHZ and are showcased with DDR 2 800 memory. That is the only way to show any kind of performance difference from the DDR 400 based existing 939 CPUs.
The guys that ran some benchmarks on those machines confirmed that you can expect three to five per cent performance increase and that is about it"
They said you can expect a AM2 chip to deliver 3-5% more performance provided that you are using DDR2 800(that will cost a fortune for sure).

LOOK,i know it is hard for you to admit but that's the truth. Overall there is a 3-5% increase. One way or the other that's the magic number.
Untill 65nm that's due to arrive in 2007 the performance boost over a 939 CPU having the same frequency is 3-5%. No need to theorise when you have a FACT that shows it all. In fact i doubt AMD ever said that the chip will bring a hefty increase. They just mentioned that moving on to better memory will bring more performance with it. And it has, 3-5%.
Your head is full of BS. Wake up and face reality.

Larger cache? Yup i heard about L3 cache on future FX. But the one tested there was a X2 and it probably had cache equall to a 939 based X2. AMD does benefit from larger cache though the performance gains are not impressive in most apps. Then again FX 62 and X2 5000 will only have 1 mb cache. And i've yet to see a L3 cache in action.
"However, AMD will introduce an Athlon 64 X2 5000+ clocked at 2.6 GHz and equipped with 1x 1 MB L2 cache."by THG
No evolution from today's cache. Untill 2007 nothing fabolous coming from AMD just what they'll need to keep their products up to date. That might bring some performance but heck 3-5% is little no matter how you look at it. I was expecting more but you live with what you have. That being said there is a good chance AMD will loose the performance crown this year. But AMD fanboys should be happy. Future FX at X2 prices is very tempting even for Intel fans.
The problem here is that a cheap AM2 procesor in order to provide that 3-5% performance will have to come with a not so cheap DDR2 800.
Even at price performance AMD's lead might be roughly shakened.
!