Intel to beat AMD by 20%

gerbiaNem

2[H]4U
Joined
Mar 6, 2005
Messages
2,169
According to the front page, Intel estimates that its next generation will beat AMD's offering by 20%. I sure hope so, maybe we can get some lower prices and faster chips soon.
 
Well, considering the 900 series beats AMD buck for buck, and the upcoming Conroe will take the desktop crown for sure (everyone known Intel already has the mobile crown with the P-M and Yonah), that isn't surprising.
 
I wonder if anyone has taken a look at the performance of these chips and compared them to amd's offering? Its easy for Intel to push out a pr item stating it will blow the doors off amd but much harder for them to deliver on this promise.
 
Well, we don't have another source so far...and there's probably going to be some amount of bias in that statement, but Intel did blow the doors off of AMD for...thirty years, who says they can't do it again? ;)

 
gwai lo said:
Well, we don't have another source so far...and there's probably going to be some amount of bias in that statement, but Intel did blow the doors off of AMD for...thirty years, who says they can't do it again? ;)


Very true. They have money and resources AMD can only dream of. You know why Intel didn't turn around and go after AMD from a performance perspective? I'll tell you why. They didn' thave to. Their sales were strong and they are making more money than ever before, despite AMD's growth in the market.

Now, they need a new cash cow to keep that momentum going. Plus, AMD is doing more damage to their bottom dollar than they expected.
 
Only 20%? That seems very small for a new arch. I seriously expecting 30% increase over A64's.

Plus AMD roadmap suggest that will hit PM tdp with the 90nm process. There is supposedly a X2 that will have a tdp of only 30w and this on the 90nm process. AMD is matching tdp on a larger process.
 
Wait, Intel is saying their next chips are going to faster than the competition?
Stop the presses!!!!111elven11!!!.

Believe none of the PR you read, and about half the benchmarks you see.
 
I wonder how they measure the performance. Is it straight performance, or the performance-per-watt Intel keeps trumpeting with its next generation?
 
I'll believe a 5% performance increase when I see it, in benchmarks, done by a third competent and reliable party, like the [H], for example. Believe it when you see it.

Until then this is Intel's PR department creating flamebait all around the internet. :p
 
The real battle is on the server's market....and Opteron beats Xeon all the way to the bank :)
 
It's alright, Apple will hype that 20% up to 1 million percent for their macintels. :rolleyes:

Realistically, a price per performance/watt is the comparison that needs to be made... and not just the cpu itself, but supporting hardware. Admittedly, once AM2 comes out, there will be less differences between the systems, b/c the same ram can be used for both. And once ddr2-800 is available, that may boost it up in AMDs favor as well, though Intel will probably support it once it's readily available anyway.

I'm not an AMDroid by any means... I just like supporting the little guy. Makes the competition work harder. Same reason I won't shop at wal-mart.
 
Logan321 said:
It's alright, Apple will hype that 20% up to 1 million percent for their macintels. :rolleyes:

Realistically, a price per performance/watt is the comparison that needs to be made... and not just the cpu itself, but supporting hardware. Admittedly, once AM2 comes out, there will be less differences between the systems, b/c the same ram can be used for both. And once ddr2-800 is available, that may boost it up in AMDs favor as well, though Intel will probably support it once it's readily available anyway.

I'm not an AMDroid by any means... I just like supporting the little guy. Makes the competition work harder. Same reason I won't shop at wal-mart.

Price per performance/watt is like a ricer saying that his Honda 2000 makes 100HP/liter as an excuse for why he got beat by a 3Mpg 502 Chevy Big block....it's a lame argument unless it's still faster overall performance/watt regardless.

We know it's possible intel could do it. They've done it in the mobile market. If they do it in desktop they stand the chance of winning me back to their side. But for the moment I'm not holding my breath.
 
If as you say intel has done it on the mobile front . Why would you think they won't deliver on the desktop. I have seen what yonah does @ 2.7 and I am very sure the Conroe cpu will bitch slap yonah merom.
 
FreiDOg said:
Wait, Intel is saying their next chips are going to faster than the competition?
Stop the presses!!!!111elven11!!!.
...
After your panties have dried up, and you've settled down somewhat, you can bet your ass Intel has what it takes to bring down any competitor. For a long period of time they had marketing monkeys dictating where the Intel ship should be going, but now that the real captains have taken over, I'm pretty confident they know what they're doing. ;)
 
SKy042 said:
Price per performance/watt is like a ricer saying that his Honda 2000 makes 100HP/liter as an excuse for why he got beat by a 3Mpg 502 Chevy Big block....it's a lame argument unless it's still faster overall performance/watt regardless.

For home users I agree. I think if there were a large disparity between power usage of two comperable chips, people would tend to migrate towards the cheaper of the two, not the lower power of the two. (Though vendors might want the lower power, smaller PSU and easier to cool means cheaper computers).

In the corperate world performance per watt is a big deal.



1c3d0g said:
After your panties have dried up, and you've settled down somewhat, you can bet your ass Intel has what it takes to bring down any competitor. For a long period of time they had marketing monkeys dictating where the Intel ship should be going, but now that the real captains have taken over, I'm pretty confident they know what they're doing. ;)

Calm down and read the last sentance in my first post. I'm not bashing intel, I'm bashing marketing departments. And my soiled panties have little to do with either. :)
 
According to Tom's Hardware (eh..), socket AM2 with DDR2-667 is no faster (maybe a little slower) than what is out now. This is with unproven hardware though, so take it with a grain of salt.
 
serbiaNem said:
According to Tom's Hardware (eh..), socket AM2 with DDR2-667 is no faster (maybe a little slower) than what is out now. This is with unproven hardware though, so take it with a grain of salt.

Yeah...I get that strange feeling that the production parts will be significantly better.
 
HmmmDonut said:
Only 20%? That seems very small for a new arch. I seriously expecting 30% increase over A64's.

Plus AMD roadmap suggest that will hit PM tdp with the 90nm process. There is supposedly a X2 that will have a tdp of only 30w and this on the 90nm process. AMD is matching tdp on a larger process.
You're absolutely right. Many/most of the single-cores, and even an X2 3800+, will be around 30W on AM2 when it launches. And this is without needing to migrate to 65nm, let alone 45nm.
 
mavalpha said:
You're absolutely right. Many/most of the single-cores, and even an X2 3800+, will be around 30W on AM2 when it launches. And this is without needing to migrate to 65nm, let alone 45nm.
um. Keep in mind that the Conroe series has 4x the cache that the AMDs have and have 4 issue core instead of 3. Also, you are comparing AMD's LV cores with Intel's high voltage cores. When they release LV Conroe cores, then compare them with AMD LV cores.
 
Duke3d87 said:
um. Keep in mind that the Conroe series has 4x the cache that the AMDs have and have 4 issue core instead of 3. Also, you are comparing AMD's LV cores with Intel's high voltage cores. When they release LV Conroe cores, then compare them with AMD LV cores.

Nonetheless...
http://www.xbitlabs.com/news/cpu/display/20060215195420.html

One of the most interesting things about the new processors is their heat dissipation. Dual-core CPUs with F core stepping and up to 2.6GHz frequency and 2x1MB L2 cache will boast maximum 89W TDP. The today’s processors with similar technical specifications demonstrate 110W TDP at 2.2-2.4GHz core clock rate.

AMD managed to reduce the power consumption by optimizing the transistor leakage current.



Who needs 65nm? :D
(Up to 2.6GHz, that's FX-60 to ya)
 
Raudulfr said:
Nonetheless...
http://www.xbitlabs.com/news/cpu/display/20060215195420.html

Who needs 65nm? :D
(Up to 2.6GHz, that's FX-60 to ya)
ok and? They are still hitting a process limitation and have to work a lot harder to get it to work well. When you start adding more cache, you use more die space and then it costs more to produce your processor. Shrinking the die is a lot more about performance. It's also a big business move.
 
Take it for what its worth...
My friend at Intel told me they have allowed 75W to acheive the desired mghz. They should sample this week if all goes well. He also said Conroe is a bad mofo, and will be performance per watt king.
Again this is what I was told so believe it or not... whatever.
 
savantu said:
Because Intel can get with a 65nm cpu at 65w the performance of a AMD FX62 at 125w....

Besides , 65nm brings a lot of cost savings for Intel..
They can also make a processor with 16 MB of L3 cache (dual core) based on netburst operate within 140 watt territory. That's 16 MB total at 140 watts. I believe that's a bit more then AMD's cache. And this is netburst we're talking about.
 
lol, and i'd believe an increase in performance around there... the pipes on it are 12-16 task...... how many does the X-2 have? please tell me? 17. Conroe is also supposed to come out with the top end at 1333mhz fsb and 3.3ghz clock... higher clock than any X-2 and less pipes than that X-2
 
StealthyFish said:
lol, and i'd believe an increase in performance around there... the pipes on it are 12-16 task...... how many does the X-2 have? please tell me? 17. Conroe is also supposed to come out with the top end at 1333mhz fsb and 3.3ghz clock... higher clock than any X-2 and less pipes than that X-2

I think it has more Pipes or 14 for Conroe and 12 for X2. I might be wrong but I wouldn't be surpised if AMD didn't add at least two pipes for K10.
 
Donnie27 said:
I think it has more Pipes or 14 for Conroe and 12 for X2. I might be wrong but I wouldn't be surpised if AMD didn't add at least two pipes for K10.

What would be the point? It seems like goin the larger pipe route would just kill the latency advantage AMD has built up.
 
At least it doesn't have a 31 stage pipe. That's like trying to drink a Big Gulp through a 3 foot straw: It takes forever to get to your mouth.
 
Buckus said:
At least it doesn't have a 31 stage pipe. That's like trying to drink a Big Gulp through a 3 foot straw: It takes forever to get to your mouth.

But once it comes you start choking on how much you get :p
 
I suggest that we should have a sticky topic for Intel microarchitecture tutorials :cool:

I see a lot of mis-informed souls.
 
serbiaNem said:
According to the front page, Intel estimates that its next generation will beat AMD's offering by 20%. I sure hope so, maybe we can get some lower prices and faster chips soon.
I will belive it when I see it.
 
everyone stop waisting ur breath, and just wait 6 months, then we will know.
 
we'll see when it happens.. .but its sure going to be exciting... competition is great for this market, cuz we the users benefit
 
empoy said:
I suggest that we should have a sticky topic for Intel microarchitecture tutorials :cool:

I see a lot of mis-informed souls.

Feel free to try and make one :)
 
serbiaNem said:
What would be the point? It seems like goin the larger pipe route would just kill the latency advantage AMD has built up.

Yes to a point. Two more stages wouldn't kill off any advantage, hell Northwoods have 20 stages and lower latency than the Hammers. P3 has fewer Pipes (10) than Dothan 12 or 13 and then Conroe has 14. Nope, two more pipes would allow K10 to gain an easy 400MHz with a few prime cuts with as much 800MHz Gains. Less stress combind with faster Transistors and Improved Process could also see something like 3.8GHz when all is said and done. Pure Speculation on my part, ZERO Proof.

BTW, Latency was added to Venice to improve stability.
 
I think Intel caught up nicely. AMD really pulled ahead but Intel's newest chips are really competitive. I buy what performs the best though and I guess we'll just wait and see.
 
Back
Top