Intel Details New Processor Generations

http://www.hardocp.com/news.html?news=MjQ5MDgsLCxobmV3cywsLDE=


That's some interesting stuff. Anyone care to figure out any real world estimates?
40 answers Last reply
More about intel details processor generations
  1. Its good to see that Penryn is not going to be just a "dumb" shrink afterall.

    Can anyone tell me if Penryn will be compatible with current 775 mobos?
  2. There are rumors of it being compatible with 680 and the P965 boards, but those are only rumors as far as I've checked. no official statements that I've seen. They will be compatible with the P35/X38 chipsets though.

    So basically I haven't answered the question. Damn the lack of information..
    And bring on the one star!!!
  3. 8O

    AMD should be worried.

    I'm both happy and sad that we're back to the days of seeing massive CPU advances every year. I'm excited at the prospect of the applications we'll be able to run thanks to these huge leaps in processing power, but I'm less happy about the fact I can't afford to upgrade my CPU every year :(
  4. http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=2955&p=2
    Quote:
    Comparing a 3.2GHz Penryn (1.6GHz FSB) to a 3.0GHz Conroe (1.33GHz FSB), Intel has measured more than 20% increase in gaming performance (with no code changes). For video encoding applications, if SSE4 is utilized, the same Penryn vs. Conroe comparison can offer more than a 40% increase in performance.

    Finally, Intel mentioned that in the server space, the fastest quad core Penryn available (>3GHz) vs. a 2.67GHz quad core Xeon resulted in a greater than 45% increase in performance in "bandwidth and FP intensive applications". It's incredibly vague (and oddly similar to AMD's claims of Barcelona vs. Xeon performance), but Pat mentioned that STREAM and certain benchmarks in SpecFP could be considered to be "bandwidth and FP intensive".
  5. AMD better come out with its new proc.
  6. My blood itches.....


    *twitch*


    First of all, Nehalem can contain up to eight cores per die. Combined with 2-way Simultaneous Multi-Threading (SMT or Hyper-Threading), you'll have the ability to execute up to 16 threads on one chip!

    Quote:
    Even more revolutionary is the fact that Nehalem will have an integrated memory controller (IMC) and that the number of serial interconnects is variable (Intel's version of "HyperTransport").
  7. Quote:
    http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=2955&p=2
    Comparing a 3.2GHz Penryn (1.6GHz FSB) to a 3.0GHz Conroe (1.33GHz FSB), Intel has measured more than 20% increase in gaming performance (with no code changes). For video encoding applications, if SSE4 is utilized, the same Penryn vs. Conroe comparison can offer more than a 40% increase in performance.

    Finally, Intel mentioned that in the server space, the fastest quad core Penryn available (>3GHz) vs. a 2.67GHz quad core Xeon resulted in a greater than 45% increase in performance in "bandwidth and FP intensive applications". It's incredibly vague (and oddly similar to AMD's claims of Barcelona vs. Xeon performance), but Pat mentioned that STREAM and certain benchmarks in SpecFP could be considered to be "bandwidth and FP intensive".


    As with AMD and it's Barcelonotthere claims, I'll believe it when I see numbers coming from somewhere other than the source. :) But it is nice to see some details coming out...
  8. Quote:
    Even more revolutionary is the fact that Nehalem will have an integrated memory controller (IMC) and that the number of serial interconnects is variable (Intel's version of "HyperTransport").


    I wouldn't call this revolutionary - I'd call it evolutionary for Intel..
  9. Quote:
    Even more revolutionary is the fact that Nehalem will have an integrated memory controller (IMC) and that the number of serial interconnects is variable (Intel's version of "HyperTransport").


    I wouldn't call this revolutionary - I'd call it evolutionary for Intel..


    Definitely revolutionary, this like the renaissance all over again Intel emerges from the dark age of Netburst and into the light of mutiple instructions per clock cycle and IMC. And now a moment of silent prayer for AMD who gave their blood and possibly their life to get us here.
  10. Quote:
    http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=2955&p=2
    Comparing a 3.2GHz Penryn (1.6GHz FSB) to a 3.0GHz Conroe (1.33GHz FSB), Intel has measured more than 20% increase in gaming performance (with no code changes). For video encoding applications, if SSE4 is utilized, the same Penryn vs. Conroe comparison can offer more than a 40% increase in performance.

    Finally, Intel mentioned that in the server space, the fastest quad core Penryn available (>3GHz) vs. a 2.67GHz quad core Xeon resulted in a greater than 45% increase in performance in "bandwidth and FP intensive applications". It's incredibly vague (and oddly similar to AMD's claims of Barcelona vs. Xeon performance), but Pat mentioned that STREAM and certain benchmarks in SpecFP could be considered to be "bandwidth and FP intensive".


    Man AMD is in deep deep shit 8O
  11. Quote:
    Its good to see that Penryn is not going to be just a "dumb" shrink afterall.

    Can anyone tell me if Penryn will be compatible with current 775 mobos?


    The intel bearlake motherboards are 1333Mhz FSB / VRM 11 motherboards - and Penryn is suppoesd to work with them OK...
  12. Quote:


    Man AMD is in deep deep **** 8O


    Very much, who said that Penryn was delayed? :)

    AMD has no answer to Intel's 65 nm and is clueless about 45 nm. AMD is going to be 1.5 process technology nodes behind by Q4 2007 8O
  13. Ive seen this move in wrestling I dont know what its called but its where one wrestler puts the other wrestler on the top corner and climbs up and does a suplex from there... I think this is what Intel wants to do to AMD... lol

    Edit:I should add that at that point it seems like the match is over ! AMD needs to go for the crotch punch to block ! lol Note: Crotch punch = legal action... lol

    Edit to the edit: See computer chips can be exciting :) lol
  14. Intel is indeed keeping up a very aggresive timetable.

    The rumours appear to have been mostly correct then.

    The 8 cores x 2 threads per core => 16 threads is mildly different than the 4 cores x 4 threads per core => 16 threads that was ~~thought~~ to be the case.

    The multible CSI links is basically confirmation that Intel will use the full boat IMC architecture from the murdered Alpha EV8, so no news here either.

    The integrated GPU is indeed news, I guess Intel wants it all with Laramee doing discrete and integrated on chip being done on the same die. (Integrated Memory Controller + Integrated graphics + CPU on one chip, is this "Son of Timma..?") -if they use Rambus XDR I guess the Timma engineers will have to sue for copyright infringement or intelectual property violation or something (or at least get royalties)

    How "real" these claims of up to 40% (similar to AMD's 40% claims) remains to be seen.

    You can always find certain instruction sequences when a new specialized SSE "N+1" instruction will show huge jumps in speed, but how broadly based these gains will be is always the 64,000 megahertz question.

    I am surprised at the suggested modest speed bumps clock wise.

    Most core2duo will go deep into the 3.xx range already, so I would have expected 4.xx or more from a 45 nano part, especially if all the leakage cliams from high/low K are true.

    I guess we will see in 9 months :)
  15. Does anyone know of any currently released motherboard that will support all (or most) of the features of the Penryn? Do you imagine the 965/975 chipsets could support Penryn?

    Or must we wait for Bearlake?
  16. omgg! , what the hell is AMD doing?? , they are about to be in some very serious trouble,

    Intel is attacking them in all corners of the market , and really trying to kill them off lol.

    Anyone know whats happening to Barcelona ? R600?
  17. Quote:
    Does anyone know of any currently released motherboard that will support all (or most) of the features of the Penryn? Do you imagine the 965/975 chipsets could support Penryn?

    Or must we wait for Bearlake?


    See this article.

    Quote:
    Intel Digital Enterprise Group chief Steve Smith, speaking to journalists today, didn't address mobile processor speeds, but he did say that 45nm desktop and server chips will be made available clocked to 3GHz and above. Xeon DPs will see their frontside bus speed upped to 1600MHz.

    Even so, the new mobile, desktop and server chips will fit into motherboards designed for today's Core 2 Duo, Core 2 Extreme, Core 2 Quad and Xeon mobos, BIOS updates aside. The desktop and server CPUs will operate within today's thermal and power envelopes, Intel said: 65W for desktop dual-cores; 95W and 130W for desktop quad-cores; 40, 65 and 80W for dual-core servers; and 50, 80 and 120W for quad-core Xeons.


    That's awesome. I could go for a Penryn CPU for sure. Or I could remain satisfied with my E6300 until Nehalem next year! :)
  18. Quote:
    Even more revolutionary is the fact that Nehalem will have an integrated memory controller (IMC) and that the number of serial interconnects is variable (Intel's version of "HyperTransport").


    I wouldn't call this revolutionary - I'd call it evolutionary for Intel..


    Definitely revolutionary, this like the renaissance all over again Intel emerges from the dark age of Netburst and into the light of mutiple instructions per clock cycle and IMC. And now a moment of silent prayer for AMD who gave their blood and possibly their life to get us here.

    Methinks thou hast regarded, peradventure, too many Lord Of The Rings movies.
  19. Quote:
    There are rumors of it being compatible with 680 and the P965 boards, but those are only rumors as far as I've checked. no official statements that I've seen. They will be compatible with the P35/X38 chipsets though.

    So basically I haven't answered the question. Damn the lack of information..
    And bring on the one star!!!

    I know you did a lot of research on this before.
  20. Quote:
    Does anyone know of any currently released motherboard that will support all (or most) of the features of the Penryn? Do you imagine the 965/975 chipsets could support Penryn?

    Or must we wait for Bearlake?

    Ask the ninja or jack
  21. Quote:
    Its good to see that Penryn is not going to be just a "dumb" shrink afterall.

    Can anyone tell me if Penryn will be compatible with current 775 mobos?


    Penryn with have +/- 21 million additional LOGIC (not cache) transistors versus Conroe, there is a LOT going on besides a die shrink

    :wink:
  22. Now, what both Intel and AMD need, more and more, and it will be just increasing....is much more demanding software!

    This is the only thing that can make all the progress (evolutionary is correct useage of course), pay off.
  23. Quote:
    http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=2955&p=2
    Comparing a 3.2GHz Penryn (1.6GHz FSB) to a 3.0GHz Conroe (1.33GHz FSB), Intel has measured more than 20% increase in gaming performance (with no code changes). For video encoding applications, if SSE4 is utilized, the same Penryn vs. Conroe comparison can offer more than a 40% increase in performance.

    Finally, Intel mentioned that in the server space, the fastest quad core Penryn available (>3GHz) vs. a 2.67GHz quad core Xeon resulted in a greater than 45% increase in performance in "bandwidth and FP intensive applications". It's incredibly vague (and oddly similar to AMD's claims of Barcelona vs. Xeon performance), but Pat mentioned that STREAM and certain benchmarks in SpecFP could be considered to be "bandwidth and FP intensive".


    Man AMD is in deep deep **** 8O

    Isn't this just kinda putting peryne on par with where barcalona is supposed to be? And plus thats at 3.2ghz. They might mean those in actual performance or performance per clock, as we have yet to determine which it will be from either company.

    Anyway, i still don't see the need for AMD is doomed posts...
  24. Quote:
    http://www.anandtech.com/cpuchipsets/intel/showdoc.aspx?i=2955&p=2
    Comparing a 3.2GHz Penryn (1.6GHz FSB) to a 3.0GHz Conroe (1.33GHz FSB), Intel has measured more than 20% increase in gaming performance (with no code changes). For video encoding applications, if SSE4 is utilized, the same Penryn vs. Conroe comparison can offer more than a 40% increase in performance.

    Finally, Intel mentioned that in the server space, the fastest quad core Penryn available (>3GHz) vs. a 2.67GHz quad core Xeon resulted in a greater than 45% increase in performance in "bandwidth and FP intensive applications". It's incredibly vague (and oddly similar to AMD's claims of Barcelona vs. Xeon performance), but Pat mentioned that STREAM and certain benchmarks in SpecFP could be considered to be "bandwidth and FP intensive".


    Man AMD is in deep deep **** 8O

    Isn't this just kinda putting peryne on par with where barcalona is supposed to be? And plus thats at 3.2ghz. They might mean those in actual performance or performance per clock, as we have yet to determine which it will be from either company.

    Anyway, i still don't see the need for AMD is doomed posts...

    same here...we cant confirm anything untill the real benchies come...but still i feel that its hightime amd did come up with something special for real
  25. For all we know Barcelona isn't even as fast as Clovertown. :roll:
  26. AMD is not doomed......they still kick VIA's ass. :lol:
  27. Quote:
    Well, it appears that Intel is pulling out some old tricks with the Radix-16 algorithm to speed up the FPU, apparently they are using IP from the purchase of DEC:

    http://www.patentstorm.us/patents/5023827.html


    Since Intel is copying AMD, which copied Intel, we should expect any innovation to be copied [or "followed" instead of "copied" if you prefer!] as soon as possible, which usually would mean that in time you end up with Coke and Pepsi, that is, two main brands, one better advertised. But, why shouldn't AMD get smart eventually about advertising? Not that they will! But it's possible. To me AMD's advertising (which only seems to be their logo and "smart choice"), the little bit there is, seems amateur.

    btw, nice research on that link! I'd vote it 5 stars, but the votes don't mean anything, since they seem pretty random and have little relation to the quality of a post. So I won't degrade your post with 5 stars :wink:
  28. Quote:
    Well, it appears that Intel is pulling out some old tricks with the Radix-16 algorithm to speed up the FPU, apparently they are using IP from the purchase of DEC:

    http://www.patentstorm.us/patents/5023827.html


    Used in every AMD chip since the origanal Athlon!!
  29. Quote:
    Well, it appears that Intel is pulling out some old tricks with the Radix-16 algorithm to speed up the FPU, apparently they are using IP from the purchase of DEC:

    http://www.patentstorm.us/patents/5023827.html


    Used in every AMD chip since the origanal Athlon!!

    This is something I was unware of, a detail I lacked knowledge on .... yep... a little diggin'....

    Thanks,
    Jack 8O :o .....He is only human. :wink: :D
  30. Quote:
    Well, it appears that Intel is pulling out some old tricks with the Radix-16 algorithm to speed up the FPU, apparently they are using IP from the purchase of DEC:

    http://www.patentstorm.us/patents/5023827.html


    Since Intel is copying AMD, which copied Intel, we should expect any innovation to be copied [or "followed" instead of "copied" if you prefer!] as soon as possible, which usually would mean that in time you end up with Coke and Pepsi, that is, two main brands, one better advertised. But, why shouldn't AMD get smart eventually about advertising? Not that they will! But it's possible. To me AMD's advertising (which only seems to be their logo and "smart choice"), the little bit there is, seems amateur.

    btw, nice research on that link! I'd vote it 5 stars, but the votes don't mean anything, since they seem pretty random and have little relation to the quality of a post. So I won't degrade your post with 5 stars :wink:I really have to laugh at all this.."Intel's copying AMD" excuse, that gets bantied around far to often. "If it weren't for AMD(K8 ), Intel wouldn't have been forced to make Core2Duo. So you can thank AMD for your C2D."
    If people really want to go down that road, you can thank Intel for your AMD(company).period! There would likely be no AMD, if it weren't for Intel. IIRC some engineers(or an engineer) left Intel and joined a few other engineers, and dreamt up AMD. :wink:
  31. As cool as this is, I have a hard time justifying a processor upgrade at the moment.

    Most of my games are still GPU bottlenecked (Even with an 8800 GTS), and anything productive I do is bottlenecked at the hard drive.

    For now, this is cool technology, but I'm apt to drop $400 into 2 15k drives before dropping a new proc in my PC.

    I think it's just getting to the point where CPUs far exceed the requirements for productivity software (Photoshop, Visual Studio, Office), and Games are generally stopped at the GPU. Other than encoding and CADD, I'm not sold on all of the new power.

    Even if one or two games come out that benefit from the extra horsepower (such as Alan Wake), that's just not enough to justify an upgrade.
  32. Quote:
    As cool as this is, I have a hard time justifying a processor upgrade at the moment.

    Most of my games are still GPU bottlenecked (Even with an 8800 GTS), and anything productive I do is bottlenecked at the hard drive.

    For now, this is cool technology, but I'm apt to drop $400 into 2 15k drives before dropping a new proc in my PC.

    I think it's just getting to the point where CPUs far exceed the requirements for productivity software (Photoshop, Visual Studio, Office), and Games are generally stopped at the GPU. Other than encoding and CADD, I'm not sold on all of the new power.

    Even if one or two games come out that benefit from the extra horsepower (such as Alan Wake), that's just not enough to justify an upgrade.


    Just so.

    The cpu makers need some super-fly software to come out and eat up all those excess idle chip cycles.

    I do very heavy multi-tasking all day on my x2 4200, and it averages about 30% cpu useage when all the stuff is running active. [and that is 70% idle even at the reduced speed of QuietN'Cool!!]

    But I'd love to see it become too slow someday before the dust bunnies get to it!
  33. Quote:

    Isn't this just kinda putting peryne on par with where barcalona is supposed to be? And plus thats at 3.2ghz. They might mean those in actual performance or performance per clock, as we have yet to determine which it will be from either company.

    Anyway, i still don't see the need for AMD is doomed posts...


    I agree... this is not a doomed AMD news break...

    And yes, if you take the Intel claim (which should be treated with the same grain of salt as the AMD claim), then Penryn pulls it even to Barcelona in the FP/BW arena based on the statement.

    AMD may not be doomed yet but AMD has to have a clear win with Barcelona pulling even with Penryn doesn't get them the ability to raise prices to a sustainable level. With processor parity Intel will continue to depress prices as their production costs are lower at 45nm and let AMD bleed out to become a Microsoft type monopoly.
  34. Quote:
    [

    ....Intel ..... become a Microsoft type monopoly.


    It's an interesting scenario to imagine. What would happen then?

    My scenario: China or S. Korea (via Samsung) spends $15 or $25B (or more), and Intel has a new competitor. China could put $30B into it without a sweat. Their US treasury holdings are near $1T.

    Actually, China might do something like this regardless.... I think China's deal with Airbus means China is looking at gaining commercial aircraft capability for strategic reasons. Chips could follow. In fact, China just got Intel to commit to a fab there..... Perhaps this is already in the works, unofficially, in the corridors of power there.

    Intel won't be able to have no competitor, any more than Boeing.
  35. Quote:
    My scenario: China or S. Korea (via Samsung) spends $15 or $25B (or more), and Intel has a new competitor. China could put $30B into it without a sweat. Their US treasury holdings are near $1T.

    Actually, China might do something like this regardless.... I think China's deal with Airbus means China is looking at gaining commercial aircraft capability for strategic reasons. Chips could follow. In fact, China just got Intel to commit to a fab there..... Perhaps this is already in the works, unofficially, in the corridors of power there.

    Intel won't be able to have no competitor, any more than Boeing.


    Halbhh, you are forgetting about America's greatest asset, and the reason why companies like Boeing, Intel, AMD, etc. are here in the US (and not someplace cheaper): the quality of our universities and the engineers they produce. It is true that China and other geographies (particularly 3rd world geos) churn out engineers, but there is not an apple to apple comparison when it goes to quality.

    Put another way, there is a reason why the world's best and brightest come to the US to study, and it's not for the bars or beaches (incidentally, this is why US students LEAVE the US :lol: )

    There are always exceptions. Technion in Israel, IIT in India, etc. Technion alumni designed the Core and Core 2 Duo for Intel in Haifa.
  36. Quote:
    Well, it appears that Intel is pulling out some old tricks with the Radix-16 algorithm to speed up the FPU, apparently they are using IP from the purchase of DEC:

    http://www.patentstorm.us/patents/5023827.html


    Since Intel is copying AMD, which copied Intel, we should expect any innovation to be copied [or "followed" instead of "copied" if you prefer!] as soon as possible, which usually would mean that in time you end up with Coke and Pepsi, that is, two main brands, one better advertised. But, why shouldn't AMD get smart eventually about advertising? Not that they will! But it's possible. To me AMD's advertising (which only seems to be their logo and "smart choice"), the little bit there is, seems amateur.

    btw, nice research on that link! I'd vote it 5 stars, but the votes don't mean anything, since they seem pretty random and have little relation to the quality of a post. So I won't degrade your post with 5 stars :wink:

    "Hi. I'm an Opteron."

    "And I'm a Core2."

    *somewhere in the distance you can hear Steve Jobs screaming as he's tarred and feathered*
  37. Nehalem is going to be impressive but the way I interpret it's implementation, Nehalem seems to be more of an integrated platform solution as opposed to a dedicated desktop/workstation solution. Someone care to help me make more sense of it?

    Quote:
    ...which usually would mean that in time you end up with Coke and Pepsi, that is, two main brands, one better advertised. But, why shouldn't AMD get smart eventually about advertising?

    It's funny you say that because I wonder where AMD would be if the hired those blue faced fags to push the Athlon or X2? Actually, I wonder where those blue faced pipe smokers would be if they hadn't been hired by Intel?!?!?!

    Quote:
    Even more revolutionary is the fact that Nehalem will have an integrated memory controller (IMC) and that the number of serial interconnects is variable (Intel's version of "HyperTransport").


    I wouldn't call this revolutionary - I'd call it evolutionary for Intel..

    Definitely revolutionary, this like the renaissance all over again Intel emerges from the dark age of Netburst and into the light of mutiple instructions per clock cycle and IMC. And now a moment of silent prayer for AMD who gave their blood and possibly their life to get us here.
    Puh-leez! The is a definite evolution for Intel, it was a matter of time before they would have no choice but to go with an IMC, the countdown to the demise of the front side bus began when AMD adopted the IMC. Just as Intel adopting an IMC was eventually going to happen, so it goes with CSI; and there is nothing revolutionary about that.

    I'm glad to see that this hasn't totally turned into an total Intel-Lovefest and bash AMD thread. Some interesting links and good reads. Thanks!
  38. Quote:
    Does anyone know of any currently released motherboard that will support all (or most) of the features of the Penryn? Do you imagine the 965/975 chipsets could support Penryn?

    Or must we wait for Bearlake?


    See this article.

    Quote:
    Intel Digital Enterprise Group chief Steve Smith, speaking to journalists today, didn't address mobile processor speeds, but he did say that 45nm desktop and server chips will be made available clocked to 3GHz and above. Xeon DPs will see their frontside bus speed upped to 1600MHz.

    Even so, the new mobile, desktop and server chips will fit into motherboards designed for today's Core 2 Duo, Core 2 Extreme, Core 2 Quad and Xeon mobos, BIOS updates aside. The desktop and server CPUs will operate within today's thermal and power envelopes, Intel said: 65W for desktop dual-cores; 95W and 130W for desktop quad-cores; 40, 65 and 80W for dual-core servers; and 50, 80 and 120W for quad-core Xeons.


    That's awesome. I could go for a Penryn CPU for sure. Or I could remain satisfied with my E6300 until Nehalem next year! :)

    Fantastic, indeed.

    So... does it make the most sense to get a board that has a 1333 FSB? Would 1066 boards be severely crippled in comparison?
  39. Quote:
    My scenario: China or S. Korea (via Samsung) spends $15 or $25B (or more), and Intel has a new competitor. China could put $30B into it without a sweat. Their US treasury holdings are near $1T.

    Actually, China might do something like this regardless.... I think China's deal with Airbus means China is looking at gaining commercial aircraft capability for strategic reasons. Chips could follow. In fact, China just got Intel to commit to a fab there..... Perhaps this is already in the works, unofficially, in the corridors of power there.

    Intel won't be able to have no competitor, any more than Boeing.


    Halbhh, you are forgetting about America's greatest asset, and the reason why companies like Boeing, Intel, AMD, etc. are here in the US (and not someplace cheaper): the quality of our universities and the engineers they produce. It is true that China and other geographies (particularly 3rd world geos) churn out engineers, but there is not an apple to apple comparison when it goes to quality.

    Put another way, there is a reason why the world's best and brightest come to the US to study, and it's not for the bars or beaches (incidentally, this is why US students LEAVE the US :lol: )

    There are always exceptions. Technion in Israel, IIT in India, etc. Technion alumni designed the Core and Core 2 Duo for Intel in Haifa.

    I 100% agree about America having a huge advantage. In a word, it's freedom.

    That's why so many of the brightest choose to live here.
Ask a new question

Read More

CPUs Processors Intel