K10 will be 60% faster than Intel in IPC
George Ou is such a retard. That idiot is screaming over the difference between 3% or -6%.
The whole point made by AMD was that the 2.8GHZ Opteron and 3GHZ Woodcrest and neck to neck on integer performance within margin of error.
What will kill Intel is the K10. K10 will be 42% faster than Xeon 5355, which runs at 2.66GHZ. But AMD's top dog will be at 2.3GHZ. Which means in in terms of IPC, AMD will be 60% faster.
BTW: It seems Nvidia's IGP is in trouble.
17 Comments:
Intel Quad core is over 2.5x faster than AMD's dual core. AMD is finished. Clovertown is so advanced that it has pre-fragged Barcelona and all AMD's other up-coming architecture.
Intel has new DX10 IGP coming out soon. Even S3 will beat AMD to DX10. R600 requires another silicon respin, that's why it was delayed again. By the time it's done the Geforce 9 will be out.
AMD BK Q2'08.
Sharikou PhD = Piled High and Deep!
Did you know it doesn't matter what the IPC is, all that matters is what the final CPU frequency / Power is and what the final system level performance is. Mock what you want benchmarks at a given frequency etc. etc. In the end all that matters is how many CPUs you sell and how fast they are.
Actually it not even matters how fast they are it only matters how many you sales. Even in the dog day sof 2005 and 2006 when Netbust couldn't win a benchmark INTEL made billions and the poor men in green were so green with envy... little cry babies calling foil, cheating.. but the bottom line is they couldn't make money or make a signficant dent on INTEL when they were ahead in IPC, benchmarks or eveyrthign else. They go not chance now, INTEL crushes them across the board.
Barcelona is all FUD, not a benchmark or engineering platform demo'd anywhere for any independent evaluatoin. I know why.. they are so afriad of hte performance that any early annoucement will give INTEL ample time to flood the channel with a faster Core2 and crush them. Theyr roadmap is history.
Then in 2nd half 2007 45nm Penrym is here with huge quadcore caches... In 2008 Nehalem ramps with even more surprise coming from 4 45nm factories..
It isn't even a debate. IN 2005 it was interesting as all the performance numbers were in AMDs favor and they still couldn't make money.. what chance to they have now now that they got NOTHING.. NO Chance..
Go back to school and get a real PhD
Hey Lex.
to quote you..
"Actually it not even matters how fast they are it only matters how many you sales."
right... so here is the punchline
1. so it doesn't matter what tweaks on old tech that intel comes out with.. since performance doesn't matter.
2. so market share matters most, yes? by your own admission, AMD has been kicking intel's ass, since intel has alot less market share and alot less sales than they had in the not too distant past...
oh.. and AMD had a record amount of market share last report and next report it will be even higher.
enjoy yourself while you can before AMD's K10 Kung Fu kicks Intel's Piii Voo Doo
I"m not even going to comment on the Intel VS AMD anymore, itsss seems Sharikou has a random number generator when it comes to "knowing" the performance of k10.
As for intel 6150 IGP how can it be in trouble, did you not know you can overclock it as well. The 690G is a faster IGP but really for being so late to the game im' suprized its not faster, and even loses to the 61X0 IGP's.
Once again Nvidia's or ATI(AMD now) cash cows are not IGP performance, their looking to enable some light gaming that woudnl't be possible on other integrated graphics, and compatibility too.
PLease get a damn life Sharikou these posts are getting horrible.
Care to explain this-----> Sharikou, Ph. D said...
Intel is a disgrace to the civilized world and the whole IT industry. How narrow minded these people are! Henri Richard told stories about Intel exes refused to attend events where AMD is on stage. ATI has not merged with AMD yet, and Intel ban ATI from their event.
Any way, I reiterate my projection that Intel will post operating losses in 3Q06 and seven quarters afterwards. Conroe volume is too late and too few. By the time Conroe ramps to 25% of Intel's units, AMD's K8L will be out. It will take Intel at least three more years to catch up with K8L.
Friday, July 28, 2006 12:27:00 AM
Intel did not post losses. Conroe volume is ramping up and we STILL don't have K8L/K10 or whatever its supposed to be called.
Complete total Fud
AMD is running out of money. What do you think of that AMD fanboys? Intel has denied AMD all profit since they can produce CPUs much more cheaply. AMD doesn't have the money to transfer any of it's fabs to 45nm. AMD is finished. When AMD goes BK Intel and Nvidia will peck at the scraps of AMD. Hector Ruiz and Henri Richard will have no beg Paul Otellini for a job. They will be working for Intel as janitors!!
AMD is a year behind Intel in manufacturing technology. No quad cores for AMD. AMD's IGPs are only shader model 2 compliant, making them completely useless. R600 is delayed again, probably because of its insane power requirement and absurd length. By the time R600 rolls out, Nvidia will frag AMD with G90.
AMD BK Q2'08
This comment has been removed by the author.
It appears that your prediction were mostly correct, Sharikou!
BK in Q2’08
C2D fragging
The only thing you got wrong is the name of the company:
“AMD Warns: Q1 2007 revenue miss!”
LOL!
Roborat, Ph. D.,
AMD:
-as bought one of the best GPU company (worth bilions). That company does not cost the same of Uli for example.
-Build another FAB.
Intel on the other hand:
-is firing people
-selling factories
-selling divisions of the company
Who will show negative results?
Dont need to answer.
"AMD's IGPs are only shader model 2 compliant, making them completely useless."
Useless is having some product that have the features, but you cant run them because they really don’t work.
Uhhhh. Intel is posting billions in profits. AMD is the one posting losses. Now a revenue miss? Not having R600 out there is killing Ati and AMD is losing valuable server market share to Intel because of quad core CPUs.
Intel fired a lot of people, sold a lot of stuff, sold excess factories. All that was to get back on track. Intel's '07 will be far better (financially) than '06. Intel will regain more server market-share and it's ASP will continue to rise resulting in more profit. AMD, OTOH, is is facing a cash crisis. They're running out of money and spent billions to buy ATi when they were in the middle of a price war with Intel. Not a smart move.
Intel fired a lot of people, sold a lot of stuff, sold excess factories.
1.Most of the people fired were from the marketing department.
2.Sold a lot of stuff - that is the money losing Xscale division and they partnered with Micron for flash.They will probably sold NOR too , since it losses money.
3.They closed one FAB in Israel ( 250/180nm ) and sold one in the US ( 180nm ).These were old FABs , not worth upgrading.
At the same time , Intel is building 2 45nm FABs ( the largest in the world ) and upgrading one from 90nm to 45nm,The total investment is worth $9.5 billion.Whoever thinks Intel shrinked is an idiot.
Randy Allen said...
Hector Ruiz and Henri Richard will have no beg Paul Otellini for a job. They will be working for Intel as janitors!!
So in other words... you'll all be work buddies!
Uhhhh. Intel is posting billions in profits.
So did Enron and WorldCom... what's your point!
AMD is the one posting losses. Now a revenue miss?
What loss? Do make the distinction between acquisition write downs/offs and losses. If AMD had not bought ATI, they would have made like $400 million or something like that!
Intel fired a lot of people, sold a lot of stuff, sold excess factories. All that was to get back on track.
OH, so are you admiting that they wern't on track? This would be a first! Thought everything was so fine with Intel?
AMD, OTOH, is is facing a cash crisis. They're running out of money and spent billions to buy ATi when they were in the middle of a price war with Intel. Not a smart move.
Cash crisis? Sure they used cash, but they can replace that cash with the already-established credit margin.
And for everyone that thinks debt is bad, well think again! Unlike your debt, a companies debt is all tax deductible, while that cash they 'could' have used is earning interest. Debt is good for a company!
BUT, being a realist, I do have to admit three things:
1) AMD should have paid a lot more of the acquisition with stock rather than cash... BAD move!
2) I don't know what Hector is doing with the Barcelona code-of-silence! This is just ridiculous and is hurting the stock!
3) AMD should start looking at putting it's pride in the gutter and think about releasing two/four dies-on-chip rather than native solutions. They don't have the processing power for this behavior!
Jeach!
This comment has been removed by the author.
JEEASSSSH said!
… Intel is posting billions in profits.
So did Enron and WorldCom... what's your point!
Intel gaining revenue share at the cost of AMD who is losing money, Randy Allen pointed out the big contrast between two competing companies. Bringing up corporations who’d admitted accounting fraud into the Intel-AMD discussion is completely irrelevant and without logic.
What loss? Do make the distinction between acquisition write downs/offs and losses. If AMD had not bought ATI, they would have made like $400 million or something like that!
AMD will post a loss based on combined AMD/ATI revenue for Q1’06. Maybe you’re in denial, AMD predicted $1.8B revenue last December for this Qtr but now can’t even seem to hit $1.6B. AMD is approaching operating loss very soon. Still, what Randy Allen said was accurate. AMD still posted a big fat loss it doesn’t matter where it comes from.
Cash crisis? Sure they used cash, but they can replace that cash with the already-established credit margin.
Absolute non-sense. It’s “already established credit margin” is zero after 2 quarters of profit warnings, inaccurate forcasts, inventory build up and product delays.
And for everyone that thinks debt is bad, well think again! Unlike your debt, a companies debt is all tax deductible, while that cash they 'could' have used is earning interest. Debt is good for a company!
LMFAO. You really are going to say anything just to justify the awful position AMD is in right now. AMD has a cashflow problem. CAPEX will suck it dry this year and without enough sales, it will have to borrow money somewhere. What investors are worried about is visibility of profitability. There isn’t any. You can only borrow enough until one day no one will lend you money and your creditors will want their money back. And that’s BK for you. Seems like AMD is well on its way.
BUT, being a realist, I do have to admit three things:
1) AMD should have paid a lot more of the acquisition with stock rather than cash... BAD move!
The terms of payments wasn’t solely in AMD’s control. Surely, ATI investors wouldn’t have accepted an AMD stock only acquisition and I’m sure AMD tried their best.
2) I don't know what Hector is doing with the Barcelona code-of-silence! This is just ridiculous and is hurting the stock!
Showing a good product will hurt current sales immediately due to Osborne effect. Showing a bad product will hurt current sales as customers will slowly drop the platform while stocks plummet. Seems like keeping quiet is the best option.
3) AMD should start looking at putting it's pride in the gutter and think about releasing two/four dies-on-chip rather than native solutions. They don't have the processing power for this behavior!
They can't because of capacity and the IMC. Not like they didn’t try and got burned by the 4x4 Luke-i-am-your-QuadFather.
"They can't because of capacity and the IMC."
Just the capacity. The second die could access ram thought HT link.
And the 120W+120W would get them into the prescott problem. Not wanted.
Just the capacity. The second die could access ram thought HT link.
And the 120W+120W would get them into the prescott problem. Not wanted.
If they used two dies from an energy efficient core, running @ 2.4Ghz or 2.6Ghz that would be possible in a reasonable thermal envelope. 2x 65W = 130W. That's certainly possible. I'm not sure how they handle the IMC issue though. I guess the IMC on one die is deactivated, and the two dies are connected via a fast hyper-transport link.
Post a Comment
<< Home