Thursday, June 22, 2006

Quiz: who said this?

I'm following the George Patton philosophy. We're not going to defend [trash], we're just going to attack, attack, attack. We're going to attack ...everywhere. We're not interested in defending. Somebody asked [Patton], "Hey, should we dig a trench or a hole?" He said, "You do that, you die." Same thing here.

19 Comments:

Blogger Ryan Ho said...

hint... go to google, enter

"I'm following the George Patton philosophy"

click "I'm feeling lucky"

No wonder Google is so many people's best friend...

9:13 AM, June 22, 2006  
Blogger asssssssssss said...

Hector obviously:)

9:17 AM, June 22, 2006  
Blogger Sharikou, Ph. D said...

click "I'm feeling lucky

Hmm, that was too easy.

9:26 AM, June 22, 2006  
Anonymous Anonymous said...

Sorry, this is off topic but am really curious to get the Dr.S opinion...

If yields for Intel's "Core" micro architecture are as bad as you have stated, why would they release their highest clocked and highest FSB chip (Woodcrest) first?

Wouldn't it make sense to release the lower quality chips first, the ones with imperfections, vs. the ones that have to be almost perfect?

Or does this mean Intel is gonna flood the market with Conroes on July 23?

Thanks.

12:29 PM, June 22, 2006  
Blogger Ajay S. said...

whoops, looks like good news day for AMD at the inquirer

i like the title of this one "Help! My MacBook goes moo", bascially Apple having problems cooling the core duo

http://www.theinquirer.net/?article=32582

AMD dell hookup much bigger than expected

http://www.theinquirer.net/?article=32536

65nm opterons improve on power efficency, according to AMD

http://www.theinquirer.net/?article=32585

2:09 PM, June 22, 2006  
Blogger Sharikou, Ph. D said...

If yields for Intel's "Core" micro architecture are as bad as you have stated, why would they release their highest clocked and highest FSB chip (Woodcrest) first?

Woodcrest will clock at various speeds, as low as 1.8GHZ. The fact Intel release Woodcrest first proves that yield is low, so Intel has to do the low volume chips first.

2:37 PM, June 22, 2006  
Blogger Mad Mod Mike said...

The reason for Woodcrest release is also the official statement from Intel that says "AMD beat our @$$ on the Server Front and we need to make a new product to beat their 5 year old product, but we hope it can beat it because so far, it isn't"

2:47 PM, June 22, 2006  
Anonymous sdba said...

AMD launch "Anti-Hyper-Treading" ?

http://www.xbitlabs.com/news/cpu/display/20060622143710.html

Well, the performance war is not settled yet..

3:32 PM, June 22, 2006  
Anonymous Anonymous said...

How about a comment or 2 about what you've said.

AMD over Intel architecture

'Intel CPUs have none of the above. In Intel architecture, all communications happen on an external shared bus controlled by an external chipset manufactured on 130nm process. The fastest future Intel bus has a bandwidth of 10.6GB/s, less than the bandwidth required for DDR2 800MHZ(12.8GB/s).'


Is that why they tried to partner up w/ Rambus to solve the bandwidth/bus bottleneck issue?

4:39 PM, June 22, 2006  
Anonymous Anonymous said...

"Woodcrest will clock at various speeds, as low as 1.8GHZ. The fact Intel release Woodcrest first proves that yield is low, so Intel has to do the low volume chips first."

This statement shows you have no comprehension of Silicon manufacturing - stick to 64bit analysis.

5:35 PM, June 22, 2006  
Anonymous Anonymous said...

The reason for Woodcrest release is also the official statement from Intel that says "AMD beat our @$$ on the Server Front and we need to make a new product to beat their 5 year old product, but we hope it can beat it because so far, it isn't"


This is a far better analysis than sharikou - the biggest performance delta is in servers). I would add 2 things:

1. Server products generally have better margin so it helps the bottom line quicker, too...

2. The release dates are really only 1-2 months apart for the various products so it really doesn't make a huge difference (you don't fix a yield issue in 1-2 months as it takes more than that amount of time to get a single lot out of the fab which is another reason why the "bad yield" theory is crap)

6:09 PM, June 22, 2006  
Anonymous Anonymous said...

I was a bit concerned on seeing an INDEPENDENT review of Conroe vs AMD's chips at this website...
Conroe seems to be beating AMD by a mile.

http://www.pcper.com/article.php?aid=265&type=expert&pid=13

Do you guys see something wrong in these results? IF they're true, AMD may not retain its performance per watt lead for much longer.
PS: I'm not an Intel fanboy, btw.

9:53 PM, June 22, 2006  
Anonymous Edward said...

"I was a bit concerned on seeing an INDEPENDENT review of Conroe vs AMD's chips at this website..."

Is there a reason for them not to use the same memory modules for testing?

Other than that, we see Conroe outperforms the same class of A64 about 15%, which is 1) what I've been predicting since beginning, 2) way below the initial round of Intel-prepared benchmarks.

"PS: I'm not an Intel fanboy, btw."

Yeah, I believe that, but you better know that somebody wouldn't take your words not matter what you say, as long as you don't praise AMD.

11:59 PM, June 22, 2006  
Anonymous Anonymous said...

I was a bit concerned on seeing an INDEPENDENT review of Conroe vs AMD's chips at this website...
Conroe seems to be beating AMD by a mile.

http://www.pcper.com/article.php?aid=265&type=expert&pid=13

Do you guys see something wrong in these results? IF they're true, AMD may not retain its performance per watt lead for much longer.
PS: I'm not an Intel fanboy, btw.


How dare you say that ? Don't you know that PCPer.com is also an Intel paid "pumper" ? Everybody is a paid Intel "pumper". Only Sharikou knows the truth.

My advice for him and his followers is to create a broterhood and commit mass suicide when Conroe wipes the floor with AMD.

1:56 AM, June 23, 2006  
Anonymous Anonymous said...

Conroe won this around. No doubt.

AMD has really good yield
http://www.virtual-hideout.net/articles/AMD_AM2_3500+_Overclocking/index5.shtml

Look at how AMD 3500+ being overclocked to FX-57 level. I need to hunt down a ML-44 and undervolt it to MT level!

3:06 AM, June 23, 2006  
Blogger DBA said...

The coming month will be very interesting.

1. Intel introduce Woodcrest on 23 July
2. AMD debuts 4X4 socket soon(?) and "Reverse-Hyper-Threading" AM2 on 24 July.

In the signle-thread application world, Woodcrest would struggle to match(if not no match) the "Reverse-Hyper-Threading" AMD64 X2 which executes a thread with 2 cores.

In the multi-thread application world, Woodcrest is not in better situation. Even if Woodcrest can beat AMD64 X2, it can hardly match the 4X4 plateform.

I guess Woodcrest is not a bad CPU after all, but AMD is simply changing the rules of the game from CPU/core to system.

Oh, poor Intel..

6:08 AM, June 23, 2006  
Anonymous Anonymous said...

Do you guys see something wrong in these results? IF they're true, AMD may not retain its performance per watt lead for much longer.
PS: I'm not an Intel fanboy, btw.


That test looks ok to me. Conroe is 65nm and will have the performance/watt lead until k8 rev g is coming later this year.

9:16 AM, June 23, 2006  
Anonymous Anonymous said...

And who coined the phrase "With friends like this, who needs enemies.."

That's what Dell is thinking of the channel stuffer Intel.

9:18 AM, June 23, 2006  
Blogger Eddie said...

todospara1/chicagraf0 here

About Patton and Attacking:

"I don't want to get any messages saying that we are holding our position. We're not holding anything, we'll let the Hun do that. We are advancing constantly, and we're not interested in holding onto anything except the enemy. We're going to hold onto him by the nose, and we're going to kick him in the ass. We're going to kick the hell out of him all the time, and we're going to go through him like crap through a goose"

That is the speech quoted in Patton, the movie, given in 1944 in England.

You may see it here

9:44 AM, June 23, 2006  

Post a Comment

Links to this post:

Create a Link

<< Home