Conroe debunked

Status
Not open for further replies.
problem is that no matter how it's looked at, 4mb of cache is undoubtably a good thing if latencies are kept really low :p

but yes, he does have some good points.

also keep in mind one other thing: it appears that conroe is gonna suck at FPU intensive tasks. luckily for intel, there's SSE.
 
Dillusion said:
God, more fuel for the immature AMD fan club, this is never going to end.
Click to expand...
God be with us.

I love my AMD, but I also loved my P4C during its heyday. Sometimes being a fair weather fan is a GOOD thing, as in the computer industry.

Personally, I doubt seriously that Intel will drop the ball again. They know what's at stake and what was wrong with Netburst, and the Conroe numbers have been scrutinized a LOT.

The next year or so will be very cool for geeks I'm thinking.
 
The thing this article didn't address was the gaming benchmarks. Intel allowed Anand to run his own gaming benchmarks on the Conroe processor, and Conroe did very well. These benchmarks were conveniently ignored by the blogger. Frankly, who here cares if Conroe "plays" ScienceMark slower than A64? If it can kick the snot out of the same processor in UT2007? That is what matters to me, and is why I am eagerly awaiting final silicon and final benches.
 
Mayhs said:
can someone summarise it please :p the artilce wont show up
Click to expand...
basically saying that conroe's main advantage is with programs that fit within it's unified 4mb cache. he points out how the benches in sciencemark, the ones that use a very small amount of memory, conroe wins easily, but the ones that use more, the X2 catches up and surprasses.. it also explains why conroe is able to get an exceptionally fast superPI 1m time, since 1m seems to allocate just over 8mb.. 1/2 of that can be thrown in the cache at any given time, and with decent prefetching, a cache miss is gonna be pretty rare.

ontop of this, it seems that with anything that is multithreaded and likes lots of cache, there's potential negative sides to the unified cache that might pop up, similar to what happened with hyperthreading.. :(

how this will translate into real world performance, we'll have to wait and see. 4mb is a considerable amount of cache.... and games did seem to do really well in the anand test, most likely due to three things:
cache size
integer power
exceptionally optimized SSE core


however, two other interesting things:
1. how do they compare in 64bit mode? i'm using xp-64 almost exclusively now, and use 64bit apps whenever possible. it really saddens me how little progress we've made into the x86-64 world thus far, especially since all new CPUs have x86-64 capability. along with going multithreaded, this is something that programmers REALLY need to focus on, especially since it should be easier to optimize things for 64bit than making it multithreaded :p

2. conroe is apparently exceptionally similar to K8, most likely aside from a few things such as how the ALU/FPU units are set up (it seems that intel went for a ton of ALU power while scaling back the FPU a bit and focused on SSE instead.. ironic, considering the K8L is supposed to double the number of FPU units :confused: ), and the SSE execution engine, which is an area where conroe definitly holds a tremendous advantage in atm.
 
(cf)Eclipse said:
also keep in mind one other thing: it appears that conroe is gonna suck at FPU intensive tasks. luckily for intel, there's SSE.
Click to expand...
If you compare the BLAS score in Sciencemark which uses x87, Conroe is superior.
 
I knew it, there is no way I would take Intels word that thier "new" CPU is that fast.

I'm wondering if it is possible to run these benchmarks with 1MB of code (to stay within the A64 cache) and see what results they both produce...
 
Like I have always said..

AMD > Intel - In terms of being an ethical (non-lying) company, imho... Intel = SONY to me.
 
(cf)Eclipse said:
basically saying that conroe's main advantage is with programs that fit within it's unified 4mb cache. he points out how the benches in sciencemark, the ones that use a very small amount of memory, conroe wins easily, but the ones that use more, the X2 catches up and surprasses.. it also explains why conroe is able to get an exceptionally fast superPI 1m time, since 1m seems to allocate just over 8mb.. 1/2 of that can be thrown in the cache at any given time, and with decent prefetching, a cache miss is gonna be pretty rare.

ontop of this, it seems that with anything that is multithreaded and likes lots of cache, there's potential negative sides to the unified cache that might pop up, similar to what happened with hyperthreading.. :(

how this will translate into real world performance, we'll have to wait and see. 4mb is a considerable amount of cache.... and games did seem to do really well in the anand test, most likely due to three things:
cache size
integer power
exceptionally optimized SSE core


however, two other interesting things:
1. how do they compare in 64bit mode? i'm using xp-64 almost exclusively now, and use 64bit apps whenever possible. it really saddens me how little progress we've made into the x86-64 world thus far, especially since all new CPUs have x86-64 capability. along with going multithreaded, this is something that programmers REALLY need to focus on, especially since it should be easier to optimize things for 64bit than making it multithreaded :p

2. conroe is apparently exceptionally similar to K8, most likely aside from a few things such as how the ALU/FPU units are set up (it seems that intel went for a ton of ALU power while scaling back the FPU a bit and focused on SSE instead.. ironic, considering the K8L is supposed to double the number of FPU units :confused: ), and the SSE execution engine, which is an area where conroe definitly holds a tremendous advantage in atm.
Click to expand...

You have to remember that in all computing, hardware is always ahead of software. Remember the hype last year when dual-cores premiered? Everyone said that this was the future of gaming. Yet, apart from a few patches here and there, no game has shown dual-core optimizations. Or, how about the new PPU released by Ageia? I wonder how long we'll be waiting until a game company decides to put that to use. 64-bit tech is older than dual-cores, for sure, but it's still experimental technology. I'm sure you've seen the various bugs, glitches, and other signs of instability that x64 currently brings. I'm sure software companies are just waiting for a stable release of a 64-bit OS before they jump onto the bandwagon.

Also, as far as this article is concerned, I was initially inclined to dismiss it entirely because it was so obviously layered with pro-AMD bias. Ironic, considering my last 2 machines have both been AMD, isn't it? Anyway, I decided to give this guy the benefit of the doubt and re-read it. The point that the 4mb L2 cache is helping Conroe is not entirely proven, but it is certainly worth exploring. If that's his only gripe, then Conroe is definitely going to wipe the floor with AMD. Considering this is an engineering sample, the real Conroe processors will probably be better. And, Anand's latest AM2 review (posted today) shows little/no gain over the Socket 939 design. Frankly, I was impressed with the sheer efficiency of the memory controller on the K8 processors. If AMD could do the same, or even close to the same, with AM2, we'd have a real winner.
 
that article is far from being neutral and subjective...the bias is pretty heavy. And a lot of the points are just plain stupid. For example:

Conroe can just run off the cache with very high speed -- another cheap gimmick at the expense of very large die size.
Click to expand...

large cache is a cheap gimmick? How dare intel try to make their cpu's faster!!! give me a break. The writer is acting like the majority of computer programs are giant in size and the point of the large cache is solely to artificially inflate benchmark scores. And that is ridiculous. Fact is that the vast majority of code in the cpu is reused several times and that most programs and data are designed to fit into cache. What, does he honestly want intel to purposely create a small cache so that benchmarks don't fit into it and the cpu is slower? What he's saying is comparable to saying "the only reason a viper is fast is because they gave it a giant displacement engine, which is a cheap gimmick at the cost of price."

Or maybe the writer should look at AMD themselves who uses cache size as a differentiating feature between their low and high end cpus.

Overall I think almost all of his points are either pointless or just common sense. He makes a huge deal about how the performance of conroe trails off once a program is larger than the L2 cache. Really? Wow! Who would have thought? First he criticizes conroe for having a huge L2 cache, and then criticizes it for not performing well enough without it. I try to look at things subjectively, but in all honesty that article is pretty much just a great big giant beam of sunshine up AMD's ass.

K8 is a great architecture...no doubt about that. AMD did an amazing job with it. But intel has likewise produced a real gem with Core and this author needs to accept that instead of hiding in denial and trying to poke holes in intel's work.
 
Bona Fide said:
Frankly, I was impressed with the sheer efficiency of the memory controller on the K8 processors. If AMD could do the same, or even close to the same, with AM2, we'd have a real winner.
Click to expand...
yeah, i'm keeping my fingers crossed, but it's gonna be mighty tough for just a memory controller change to catch up to what conroe is doing.

SLee said:
If you compare the BLAS score in Sciencemark which uses x87, Conroe is superior.
Click to expand...
i went to look, but XS is down again. :rolleyes:
 
I'll wait for Conroe to become commercially available and then judge it for myself. We don't really have a whole heck of alot to go on right now (and how reliable is the current info?) and that leads to theorizing.
 
thnx eclipse....hmm this new cpu war is getting a lot more interesting than i first conceived it to be :)
 
I love how its ALWAYS AMD biased.. but never ever ever could there be Intel biasness happening right?
 
Tetrahedron said:
I love how its ALWAYS AMD biased.. but never ever ever could there be Intel biasness happening right?
Click to expand...
i am amd biased by definition :(
(admitting there is a problem is the first step to fixing it :D)
 
I love how the author uses esoteric benchmarks to extrapolate how conroe will do in real world tests.
Does FEAR fit in a 4mb cache?

I also like how he compares the Conroe previews to FUD, and then goes on to engage in some highly speculative FUD making of his own.

I think calling a CPU feature a "gimmick" is an arguement of last resort. "Sure, it's faster, but it's just a gimmick." You could say that about any CPU feature that increases performance, including an on-die memory controller. Die shrinks simply increase the CPU maker's transistor budget. How they use those transistors to best increase performance is up to them. I'll be anxious to see how AMD uses the extra transistors at 65nm to increase it's performance relative to Conroe.

When the dust settles and there is a performance/price leader that will provide a large enough increase over my 170 @ 2.8 GHz, I'll get it, irrespective of brand.
 
Conroe is only about 14x mm^2 by current die size estimates and that is with the "enormous" 4MB cache, thanks to 65nm fabrication technology, and Intel's exceptionally good cache density, Intel can afford to throw in this muc cache on the 65nm and still price it reasonably aka 316US for the E6600 model.

It isn't Intel's problem that AMD lags behind Intel on process technology, and has worse cache densities then Intel does. Hence AMD designed their architecture to not rely on cache too much, they simply can't afford to put that much cache on in the first place.
 
In the other thread (XS or the intel forum?) where VictorWang's conroe results showed the only benchmark that conroe lost was sciencemark, someone mentioned that sciencemark will become the most important benchmark now. :p The link in the first post proves it.
 
pxc said:
In the other thread (XS or the intel forum?) where VictorWang's conroe results showed the only benchmark that conroe lost was sciencemark, someone mentioned that sciencemark will become the most important benchmark now. :p The link in the first post proves it.
Click to expand...

it will replace 3dmark as our new favortie game
 
SLee said:
Here are a couple links to posts made by one of the main Sciencemark programmer (with scores):

http://www.aceshardware.com/forums/read_post.jsp?id=115160317&forumid=1

http://aceshardware.com/forums/read_post.jsp?id=115160639&forumid=1
Click to expand...

Man I have just about stopped going to aces. If you have been there much you would know why....nothing against Brian/Johan/et al it is some of the guys who post who are so far beyond being f.b.'s that make the place unreadable at times. Wonder where WIG went.
 
coldpower27 said:
Conroe is only about 14x mm^2 by current die size estimates and that is with the "enormous" 4MB cache, thanks to 65nm fabrication technology, and Intel's exceptionally good cache density, Intel can afford to throw in this muc cache on the 65nm and still price it reasonably aka 316US for the E6600 model.

It isn't Intel's problem that AMD lags behind Intel on process technology, and has worse cache densities then Intel does. Hence AMD designed their architecture to not rely on cache too much, they simply can't afford to put that much cache on in the first place.
Click to expand...


all ill say is i work in a fab, saying amd doesnt have 2x2 mb caches yet simply cuz of price is laughable

i cant remember if conroe is due out in the fall or the summer( no time to keep up on other company's products and shit, and not due for a comp upgrade til later) so anyway, it takes up to 3 months to manufacture chips from start to finish, so the samples anad got to benchmark is basicaly final silicon (if its coming out when i think)

and keep in mind intel was saying ~20% gain over fx 60 which had been out for a lil while, purly comical. I guess theres plenty of dell buyers out there that will buy into the hype though so i guess its kinda ingenious of intel.
 
powerade said:
all ill say is i work in a fab, saying amd doesnt have 2x2 mb caches yet simply cuz of price is laughable

i cant remember if conroe is due out in the fall or the summer( no time to keep up on other company's products and shit, and not due for a comp upgrade til later) so anyway, it takes up to 3 months to manufacture chips from start to finish, so the samples anad got to benchmark is basicaly final silicon (if its coming out when i think)

and keep in mind intel was saying ~20% gain over fx 60 which had been out for a lil while, purly comical. I guess theres plenty of dell buyers out there that will buy into the hype though so i guess its kinda ingenious of intel.
Click to expand...
Interesting that you say you work in a Fab and say why AMD doesn't have 2X2Mb because of cost is silly, well care to shed some light on why AMD didn't throw on 2x2Mb cache? I gave more then just cost reason why AMD didn't do it if you read my posts.

I think cost and yield do play an important role in determining how much cache can be put on a processor Windsor the Socket AM2 Athlon 64x2 has a Die Size of 220 mm^2 with 2x1MB and that is much larger then Conroe is. Hence why you don't see AMD throwing more cache onto it as they can't afford to at the moment not on 90nm DSL SOI, or won't because they want to enjoy the margins they have had the luxury of charging when Intel made this mistake.

Really the benches of the E6700 2.66GHZ have been shown to be on avg of 20% faster then a Athlon FX 60 overclocked to 2.8GHZ, soem cases better some cases worse. There will be an 3.00GHZ+ Extreme Edition of Conroe to combat the future Athlon FX 62. So the margin of victory will be further extended by Intel not degrade.
 
coldpower27 said:
Interesting that you say you work in a Fab and say why AMD doesn't have 2X2Mb because of cost is silly, well care to shed some light on why AMD didn't throw on 2x2Mb cache? I gave more then just cost reason why AMD didn't do it if you read my posts.

I think cost and yield do play an important role in determining how much cache can be put on a processor Windsor the Socket AM2 Athlon 64x2 has a Die Size of 220 mm^2 with 2x1MB and that is much larger then Conroe is. Hence why you don't see AMD throwing more cache onto it as they can't afford to at the moment not on 90nm DSL SOI, or won't because they want to enjoy the margins they have had the luxury of charging when Intel made this mistake.
Click to expand...

Comparing a die size of a 65 nm process to a 90nm process is, at best, moronic. Not for just for coldpower, but also powerade. If we assume that conroe is 14mm on a side at 65, we'll just go linearly to 21mm on a side at 90nm, which magically takes us to 440?sqmm.. more than twice the size of the X2. Kiddies need to learn how to play nice, and to do a little simple math before they get in big fights on the intarweb.

(btw, even if the scales aren't linear between generations, I don't care, this is utterly stupid)
 
So the margin of victory
Click to expand...

wait a minute! aer they handing out medals for this victory? if they aer, i bet amd just doesnt know about it yet and thats why they are lagging behind!!

look out hector!! they a3r here to steal you medals!! dont let them have teh 1337n3ss!!
 
baldrik said:
wait a minute! aer they handing out medals for this victory? if they aer, i bet amd just doesnt know about it yet and thats why they are lagging behind!!

look out hector!! they a3r here to steal you medals!! dont let them have teh 1337n3ss!!
Click to expand...

We should make spellcheck mandatory.
 
i would just like to say <nelson voice>ha ha </nelson voice>

And that's directed not at intel or amd camp; but at anyone who takes this with more than a single grain of salt
 
mwarps said:
Comparing a die size of a 65 nm process to a 90nm process is, at best, moronic. Not for just for coldpower, but also powerade. If we assume that conroe is 14mm on a side at 65, we'll just go linearly to 21mm on a side at 90nm, which magically takes us to 440?sqmm.. more than twice the size of the X2. Kiddies need to learn how to play nice, and to do a little simple math before they get in big fights on the intarweb.

(btw, even if the scales aren't linear between generations, I don't care, this is utterly stupid)
Click to expand...
Comparing Conroe die size to Windsor is very legitimiate, as they will be out in the same time frame, as AMD will be still on the 90nm DSL SOI process when Conroe ships, AMD has no one to blame but itself for this. It's not Intel's problem that it is 9-12 months ahead of AMD on the 65nm transistion.

Conroe for instance is about 145mm^2 for the 65nm node, however if it were built on the 90nm node it's die size would be 241mm^2, your math is off by huge margins as you obvious don't know to do it right. I wouldn't be so quick to criticize when your own math skills are quite lacking.

Let me show you why I am right.

Newcastle die size = 144mm2 on 130nm.
Winchester die size = 84mm2 on 90nm

Prescott-2M die size = 135mm2 on 90nm
Cedar Mill die size = 81 mm2 on 65nm

so 84/144 = 58.3% so 0.5833

so 81/135 = 60% so 0.60

So in order to extrapolate Conroe's die size you do the following:

145mm2/0.60 = 242mm2 on the 90nm node
To reverify, 145/242 = 59.9% which is about right for an optical shrink

The comparison though of 242mm2 is irrelevant as Conroe will never be built on 90nm technology. While Windsor will be built on 90nm technology.

So make sure indeed that your own mathskills are up to par before you go flamboyantly criticizing others.

And we are comparing the Windsor core 220mm2 die size vs Conroe which is ~145mm2 die size. So of course Intel will have the die size advantage as they were first to shift to 65nm technology and reaping the benefits of it, this is reality. AMD will have to put up with their larger dies for the time being until they themselves can shift over to 65nm.

The comparison is legtimate becuase AMD will still be using 90nm technology when Intel is using 65nm such is the problem when your slower then your competitor on process transitions.
 
baldrik said:
let me fix that for you
Click to expand...
Rofl, is that the best you can come up with, it also not Intel's problem that AMD is poorer and is lacking in financial reosurces due to lack of a competitive product for a good portion of it's lifetime. This is AMD's own doing.
 
not Intel's problem that AMD is poorer and is lacking in financial reosurces due to lack of a competitive product for a good portion of it's lifetime.
Click to expand...
yes it is.
 
Status
Not open for further replies.
Back
Top