Before Intel Core 2 Duo Conroe came out

PeteRoy

Senior member
Jun 28, 2004
958
2
91
www.youtube.com
In 2005 before Conroe came out, I remember reading all those "analysts" saying that Intel doesn't have a chance to beat AMD unless they integrate memory controller to the CPU die like AMD did.

This is just an example that sometime analysts are analysts and they don't know everything there is to know about CPU, software or whatever else it is they analyze.
 

Cookie Monster

Diamond Member
May 7, 2005
5,161
32
86
Hmm, i remember that to. The most sad thing about it is that what AMD gained during the netburst era was lost in one god damn QUARTER. "The Intel Strikes Back" as i call it.

But you know what this means.. we can expect "The Return of the AMD" :D
 

Borealis7

Platinum Member
Oct 19, 2006
2,901
205
106
I dont know what you are complaning about. just because AMD doenst hold the perfomence crown doesnt mean its bad across the board.

Need i remind you the XbitLabs article that clearly showed that after the april price cuts, Intel and AMD had virtually identical cost/performence ratio in every price bracket except the sub 100$ where AMD still rules?

AMD hasnt gone anywhere, its just making less noise.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,248
16,108
136
And as cores increase, Intel will hurt more....It was a great step, but they won a battle. The war is not over yet....
 

dmens

Platinum Member
Mar 18, 2005
2,275
965
136
Originally posted by: myocardia
Was there even a point to this post? If so, I obviously missed it.

Yeah the point is that analysts and most hardware site writers don't know anything (hi Kyle). The IMC being the only logical way to go was gospel for about 2 years, and still is in many circles. The concept of design tradeoff was apparently a novel idea to these people.
 

myocardia

Diamond Member
Jun 21, 2003
9,291
30
91
Originally posted by: dmens
Originally posted by: myocardia
Was there even a point to this post? If so, I obviously missed it.
The IMC being the only logical way to go was gospel for about 2 years, and still is in many circles.
You and the OP had better get to work on informing Intel of that, then, since their next big cpu design will have an IMC. Haven't you two heard of Nehalem?
 

harpoon84

Golden Member
Jul 16, 2006
1,084
0
0
Originally posted by: myocardia
Originally posted by: dmens
Originally posted by: myocardia
Was there even a point to this post? If so, I obviously missed it.
The IMC being the only logical way to go was gospel for about 2 years, and still is in many circles.
You and the OP had better get to work on informing Intel of that, then, since their next big cpu design will have an IMC. Haven't you two heard of Nehalem?

Well, rumor has it there will be two versions of Nehalem. The desktop version would still use FSB, but the server version uses an IMC. This is just a rumor of course, but many sites have reported this.
 

DrMrLordX

Lifer
Apr 27, 2000
22,905
12,975
136
Bollocks, nobody in their right mind thought Intel "needed an IMC". All you had to do was look at the performance of Dothan versus single-core K8 (even San Diego chips) or Yonah versus dual-core K8 (even Toledo chips) to see what Intel could do with a non-Netburst design. Overclocked Dothans and Yonahs were quite competitive despite being mobile chips with some performance problems stemming from design sacrifices made to keep the chips' power consumption levels low.

Integrated memory controllers are good things in many ways, but Intel showed us that they could compete without them even before Core 2 hit the scene.
 

JackPack

Member
Jan 11, 2006
92
0
0
It's called memory hierarchy.

A large L2 means your processor doesn't need to hit the main memory as often. An ODMC is a design tradeoff since you use die space for the controller instead of cache.
 

betasub

Platinum Member
Mar 22, 2006
2,677
0
0
Originally posted by: JackPack
It's called memory hierarchy.

A large L2 means your processor doesn't need to hit the main memory as often. An ODMC is a design tradeoff since you use die space for the controller instead of cache.

Nice summary. Deserves a :cookie:
 

f4phantom2500

Platinum Member
Dec 3, 2006
2,284
1
0
yeah, but also wasn't intel throwing benchmarks and such about conroe all over the place? amd is being rather tight-lipped about barcelona. if barcelona was the core-killer that we all hope it is, i think amd would be throwing benchmarks everywhere. plus people are getting info from industry insiders that barcelona isn't all we're hoping for. but i won't be ready to accept either as a winner until i read some actual reviews using the same chips that we will be able to buy.
 

Lonyo

Lifer
Aug 10, 2002
21,938
6
81
Originally posted by: JackPack
It's called memory hierarchy.

A large L2 means your processor doesn't need to hit the main memory as often. An ODMC is a design tradeoff since you use die space for the controller instead of cache.

They also gave Conroe some sweet memory access, which means it takes barely any more time to access main memory than AMD's IMC chips, which means they had no real need for IMC, since they had some aspects of IMC performance already (i.e. the reduced access time to main memory)

The point to the Latency discussion is that, as expected, AMD has much more opportunity for performance improvement with memory speed increases in AM2. Intel will eventually reach the point, if the lines were extended, where they would have to move to an on-chip memory controller to further improve latency. This is not to take anything away from Intel's intelligent design on Core 2 Duo. They have found a solution that fixes a performance issue without requiring an on-chip controller - for now.
 

Genx87

Lifer
Apr 8, 2002
41,091
513
126
Originally posted by: dmens
Originally posted by: myocardia
Was there even a point to this post? If so, I obviously missed it.

Yeah the point is that analysts and most hardware site writers don't know anything (hi Kyle). The IMC being the only logical way to go was gospel for about 2 years, and still is in many circles. The concept of design tradeoff was apparently a novel idea to these people.

It still is when the systems get more complex. 1-4 way Intel can get away with their chipsets doing the work. But operton does fine in the 8+ way segment vs Intels current crop of conroe based Xeons.

 

coldpower27

Golden Member
Jul 18, 2004
1,676
0
76
Originally posted by: Genx87
Originally posted by: dmens
Originally posted by: myocardia
Was there even a point to this post? If so, I obviously missed it.

Yeah the point is that analysts and most hardware site writers don't know anything (hi Kyle). The IMC being the only logical way to go was gospel for about 2 years, and still is in many circles. The concept of design tradeoff was apparently a novel idea to these people.

It still is when the systems get more complex. 1-4 way Intel can get away with their chipsets doing the work. But operton does fine in the 8+ way segment vs Intels current crop of conroe based Xeons.

I don't think there are any Core based Xeons in the 8 Socket segment just yet, Tigerton isn't due till sometime in Q3. AMD does well in segments where Core based technology hasn't reached yet.
 

F1N3ST

Diamond Member
Nov 9, 2006
3,802
0
76
ATI and AMD are both doing bad, HD2900XT is worse than the GTS 320MB, and Conroe>AMD right now. The GTS320MB is also cheaper than the 2900.
 

Alyx

Golden Member
Apr 28, 2007
1,181
0
0
AMD may have equal or higher price/performance but that doesn't mean its good for the company, just good for the consumer. AMD makes less money per processor than what intel does, this leads to lower profits over all. AMD may sell more processors by lowering price but they also don't make as much money. Intel is faster at process shrinks than AMD by about a quarter or more and AMD has historically always had difficulties with process shrinks. Not only that but intel has more fabrication facilities that are able to make processors at a smaller micron process than AMD putting AMD in a hard spot. AMD is billions in debt from the ATI purchase, AMD is loosing money because of the small profit per unit (direct result of the price war), AMD's stock is tanking/has tanked, plus they don't have enough in the piggy bank to upgrade their Fabs.
To add injury AMD does not have a processor that is near ready for production, meaning there is no golden egg to show investors to help raise capital.
 

aCynic2

Senior member
Apr 28, 2007
710
0
0
Competition is good. In this case, it made Intel get off it's fat azz and start producing something not cliche'.
 

toughwimp11

Senior member
May 8, 2005
415
0
76
Originally posted by: Cookie Monster
Hmm, i remember that to. The most sad thing about it is that what AMD gained during the netburst era was lost in one god damn QUARTER. "The Intel Strikes Back" as i call it.

But you know what this means.. we can expect "The Return of the AMD" :D

and it will end at the "return of the amd" until another star wars movie comes out. its the law
 

Noema

Platinum Member
Feb 15, 2005
2,974
0
0
Originally posted by: aCynic2
Competition is good. In this case, it made Intel get off it's fat azz and start producing something not cliche'.

Well...we can imagine that if it hadn't been for the Athlon, Intel would be launching the Pentium 5 on Netburst architecture right now, running at 10GHz, with the cheapest entry model going for $1200 and requiring Liquid Nitrogen to keep it cool and stable.

:Q :p

So I'm happy the C2D rocks. I'm certainly enjoying mine, and I do hope AMD releases a nice competitive CPU soon, both in price and performance.
 

Fox5

Diamond Member
Jan 31, 2005
5,957
7
81
Originally posted by: f4phantom2500
yeah, but also wasn't intel throwing benchmarks and such about conroe all over the place? amd is being rather tight-lipped about barcelona. if barcelona was the core-killer that we all hope it is, i think amd would be throwing benchmarks everywhere. plus people are getting info from industry insiders that barcelona isn't all we're hoping for. but i won't be ready to accept either as a winner until i read some actual reviews using the same chips that we will be able to buy.

Amd didn't throw athlon 64 benchmarks all over the place. Despite opteron being quite fast (even debuting well below 2ghz) compared to Xeons, the Athlon 64 came almost literally out of nowhere. It seemed like people had no idea what kind of performance a64 would have until right before it hit the market, and even then people thought it would only be a marginal improvement over the athlon xp.
 

bryanW1995

Lifer
May 22, 2007
11,144
32
91
Originally posted by: f4phantom2500
yeah, but also wasn't intel throwing benchmarks and such about conroe all over the place? amd is being rather tight-lipped about barcelona. if barcelona was the core-killer that we all hope it is, i think amd would be throwing benchmarks everywhere. plus people are getting info from industry insiders that barcelona isn't all we're hoping for. but i won't be ready to accept either as a winner until i read some actual reviews using the same chips that we will be able to buy.
I want to believe that barcelona will be great, but I have waited too long for amd to wake up. I do find it interesting, however, that intel is cutting prices big time and pushing penryn out now, however, whereas nvidia did nothing to fight the r600 launch except raise prices the day it was announced.
 

Borealis7

Platinum Member
Oct 19, 2006
2,901
205
106
Originally posted by: toughwimp11
Originally posted by: Cookie Monster
Hmm, i remember that to. The most sad thing about it is that what AMD gained during the netburst era was lost in one god damn QUARTER. "The Intel Strikes Back" as i call it.

But you know what this means.. we can expect "The Return of the AMD" :D

and it will end at the "return of the amd" until another star wars movie comes out. its the law

picture this:

CPU Wars: The Phenom Menace, Attack of the Cores, and Revenge of the Chip

coming in 2008 to a tech site near you! BE THERE!