Slashdot Mirror


Dual Cores Taken for a Spin in Multitasking

Vigile writes "While dual cores are just now starting to hit the scene from processor vendors, PC Perspective has taken the first offering from Intel, the Extreme Edition 840, through the paces in single- and multi-tasking environments. It seems that those two cores can make quite a difference if you have as many applications open and working as the author does in the test." It's worth noting that each scenario consists of only desktop applications, and it'd still be interesting to see some common server benchmarks, such as a database or web server.

221 comments

  1. Newsflash... by jawtheshark · · Score: 5, Funny
    SMP system performs better when applications are multithreaded...

    (Dual core is the same as an SMP system, except the cores can communicate a bit faster with each other)

    --
    Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
    1. Re:Newsflash... by EpsCylonB · · Score: 1

      yeah it is a bit dull to anyone who has been following dual core or even just SMP.

      Cue loads of stupid people saying this technology is crap because most apps aren't multithreaded and the clockspeeds are lower than their current cPU.

    2. Re:Newsflash... by beezly · · Score: 4, Insightful

      That is not necessarily true.

      I know this article is talking about Intel dual core chips, but for well-designed CPUs with integrated memory controllers (Power5, Ultrasparc IV, Opteron), the difference between a single dual-core CPU and two single-core CPUs is significant.

      On chips with built in memory controllers, as you increase the number of cores on a chip the memory bandwidth per core decreases, however as you increase the number of chips in a system, the memory bandwidth per core remains the same and the number of cores increases.

      That can amount to a big performance difference when running memory-intensive jobs.

      Intel seem to be really losing the plot here at the moment. In multichip configurations, Intel's memory bandwidth already sucks compared to Opteron. Multicore per chip is only going to make it FAR worse.

    3. Re:Newsflash... by The+New+Andy · · Score: 5, Informative
      ... assuming the OS chooses which threads are executed on which core well enough. If two threads depend on each other heavily and they are running on different cores, you can get really crappy performance.

      So the obvious answer would be to move one of the processes to the other core. However, this isn't trivial. You either have one scheduler per core or one scheduler per operating system. (You can't have a single thread sent to both cores easily - if both cores run the same thing at the same time there will be chaos)

      If you have one per core, then the scheduler trying to get rid of the thread will have to synchronise with the other core, waiting for the other scheduler to come into context, it then has to tell it to add this new process. Obviously, there is a fair bit of overhead, and if my memory serves me correctly, each core in the current chip has its own cache - so now all the stuff which was cached has to be sent to memory (since it is in the wrong cache) and now there is nothing in the cache, making every memory access slow for the next little while. End result - you can transfer a thread between CPUs, but it is costly.

      It is possible to have a single scheduler which can then just dispatch threads to each core as it gets run by each core. The big one here is making the scheduler threadsafe - both CPUs could run the scheduler at the same time, so you have to make sure they don't crap on each other. This is a problem which we have solved already with common synchro-primitives. But, if you just lock the list of threads to run (*), then you will get a whole lot of CPU time wasted just waiting to run the scheduler. It might be acceptable for 2 cores, but it doesn't scale at all.

      (*) You may realise (just as I realised) that a scheduler is more than just a list of threads to run (it is typically implemented as a couple of lists for each priority). The same problem still occurs with more than one list of threads, it is just a bit harder for me to express (proof by bad English skills).

      Finally, I'm expecting someone to tell me that I'm wrong about something I just said. That person is probably correct. My only experience with this stuff is a 3rd year undergrad operating systems course where we played around with OS161 (a toy operating system basically). But, hopefully the end conclusion will be the same: twice the number of processors won't equal twice as much performance, and it is tough to get a fast algorithm that will scale.

    4. Re:Newsflash... by jawtheshark · · Score: 1
      After reading the article, I realised that the frontside bus was shared. I didn't expect that. It seems to be a transitory solution in order to have the "first dual-core" CPUs on the market. When AMD releases theirs I expect them to have a superior solution.

      My main point was that anybody who has used SMP systems knows that they perform better in a multithreaded and/or multitasking environment. I have an AMD Athlon MP system as my main system. The CPU's aren't the newest ones (about two years old), but I have used AMD Athlon XP systems that are faster (accoring to their rating) but feel less responsive. My system is probably slower overall (266Mhz FSB isn't helping there), but it feels faster. Subjective? Probably...

      --
      Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
    5. Re:Newsflash... by beezly · · Score: 1

      AMD's are released already in the Opteron range. Hopefully I have some heading my way right now (if I'm lucky!).

    6. Re:Newsflash... by jawtheshark · · Score: 1
      twice the number of processors won't equal twice as much performance

      You will notice that I didn't claim that. Overall, what you explain is what I remeber from my classes too.

      --
      Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
    7. Re:Newsflash... by jawtheshark · · Score: 1

      Lucky you: my current finances don't allow it. Besides, for now I'm happy with my current SMP system. I'll only be screwed when 32-bit OSes die.

      --
      Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
    8. Re:Newsflash... by fabs64 · · Score: 1

      probably not subjective, being that in everyday desktop usage, you are always going to be running more than one app at once. And if they're completely different processes, they will be shared between the CPU's well, hence more responsive. the multithreaded thing comes into play when you have one app that you need to go ueber fast, you'll find on your MP system that if you run a game it will use 100% of one CPU and 0% of the other, this is what people mean by needing a multithreaded environment.

    9. Re:Newsflash... by The+New+Andy · · Score: 0
      I reread my post after submitting and wished I wrote something similar to that - sorry if it came off the wrong way.

      I just want to make sure people don't think that multicore is a simple thing as your (correct) post makes it out to be.

    10. Re:Newsflash... by rodac · · Score: 1

      If your os has a scheduler that moves a process back and forth between different cpus/cores with no understanding if the cost of invalidating and reloading the cache then you sould upgrade to an os that does not have a broken scheduler. any scheduler that can not handle this is broken. do not use broken os or broken schedulers if you have more than 1 core/cpu.

    11. Re:Newsflash... by jawtheshark · · Score: 0

      No probs... After all I was mainly going after Funny mods :-) Not Insightful or Informative as I got.

      --
      Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
    12. Re:Newsflash... by hyc · · Score: 2, Informative

      The fact is, when you have only one problem to solve, a single fast CPU is always better than an equivalent count of slower CPUs. One 1GHz CPU is better than 2 512MHz CPUs of the same design, for solving a single problem.

      It's another fact that it's easiest for humans to analyze only one problem at a time. So the most straightforward way to handle any computing task it to crunch at it linearly from beginning to end.

      Unfortunately, gains in raw CPU speed have always come slower than the demand for number-crunching power, which is why there have been so many parallel supercomputer designs over the past couple decades. And again unfortunately, the only way to get good performance out of those designs is by explicitly coding for parallel processing. And that's hard for humans.

      The bottom line is that when you throw two 1GHz processors at a problem, your time to solution is not twice as fast as using one 1GHz processor. There are overhead costs associated with reallocating the problem and managing the increased number of processors. But, it *is* still faster than using one 1 GHz processor.

      Scale the numbers up as appropriate.

      Still, it was fun watching the blinking lights on our Connection Machine, a decade or so ago...

      --
      -- *My* journal is more interesting than *yours*...
    13. Re:Newsflash... by Anonymous Coward · · Score: 0

      Intel has always used a shared-bus solution, even on IA-64. Its been a major flaw for quite some time, as competitors have much more advanced bus architectures. Intel got away with it because they are cheap, increasing the frequency / dual-pumped, and most systems never scaled past 2-4 processors.

      Hopefully multi-chip packaging will finally give Intel an incentive to start creating a point-to-point bus protocol like HyperTransport.

    14. Re:Newsflash... by jawtheshark · · Score: 3, Insightful
      when you have only one problem to solve

      That highly depends on the problem. If your problem is highly parallizable and the application that resolves your problem has been written (correctly) in a multithreaded way, then two CPU's will perform better. (As you say, it doesn't scale in a linear way)

      Of course, you might just say that a parallizable problem is not one problem, but many small problems that need to be solved separately ;-)

      --
      Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
    15. Re:Newsflash... by !the!bad!fish! · · Score: 0, Offtopic
      -- Remember Christine Watkins

      1st person eaten in Jaws.
      What do I win?

      --
      Kids today are tyrants. They contradict their parent, gobble their food, and tyrannize their teachers. - Socrates 400 BC
    16. Re:Newsflash... by jawtheshark · · Score: 1

      My congratulations, because most people don't get it :-) Actually, I think you're the first to find it out without my help.

      --
      Ahhh...the great dumpster continuum. Many a free computer will be found there. -- sowth (748135)
    17. Re:Newsflash... by beezly · · Score: 1

      That's quite true. My background is in High Performance Computing so the majority of applications I see get a significant benefit from multiple CPUs/cores.

      However, I believe we will see a change in development practices. Applications will become more multi-threaded as the number of CPUs per die increase. I think AMD have already announced that eventually they will only sell multi-core CPUs (in this market).

      Of course, this means that programmers will have to learn how to write parallel code :)

    18. Re:Newsflash... by Senor_Programmer · · Score: 2, Insightful

      Is it just me or did the performance of the single core AMD relative to the Intel dual core in those benchmarks just scream out..."I want the AMD dual core!"?

      Seriously, unless you're application can run in the cache on the Intel parts, the AMD is gonna win hands down when running at the same clock rate which translates pretty closely to the same power consumption. AMD will yet be a tad lighter on power consumption just because the stuff is packed more tightly even though it has more active components. Equal 'wire' size + smaller size = shorter 'wires' resulting in reduced IR losses (for the same operating voltage).

    19. Re:Newsflash... by EpsCylonB · · Score: 1

      And again unfortunately, the only way to get good performance out of those designs is by explicitly coding for parallel processing. And that's hard for humans.

      Yes its hard now but it won't always be hard, we will develop tools and methodologies that help us, just as we have before.

    20. Re:Newsflash... by Anonymous Coward · · Score: 0

      An "Intel Emergency Edition XP Dual Core Quad Pumped V8 Extreme OverHead-Cam Super Quadro Nitro" edition space heater.

    21. Re:Newsflash... by zerocool^ · · Score: 1


      You know, in Windows, you can hit Ctl+Alt+Del to bring up the task manager, go to the processes tab, right click on a process, and go to "Set Affinity".

      Just sayin'...

      --
      sig?
    22. Re:Newsflash... by The+New+Andy · · Score: 1
      So do you really want to have to tell your os which threads to run on which CPU? It doesn't really scale very well when you have a ga-ba-gillion threads running.

      The ability to set affinity is more of a sign that the system isn't perfect yet (it is better than something I could design, but that doesn't mean that I can't say it isn't perfect)

    23. Re:Newsflash... by mi · · Score: 1
      You know, in Windows, you can hit Ctl+Alt+Del to bring up the task manager, go to the processes tab, right click on a process, and go to "Set Affinity".
      On SunOS it is called pbind and it existed since well before Windows supported SMP at all.
      --
      In Soviet Washington the swamp drains you.
    24. Re:Newsflash... by cahiha · · Score: 1

      So the obvious answer would be to move one of the processes to the other core.

      The obvious answer is that the situation isn't a whole lot different from dual processor machines...

    25. Re:Newsflash... by vasqzr · · Score: 1


      One 1GHz CPU is better than 2 512MHz CPUs

      1GHz = 1024GHz ???

      Clockspeed isn't done in powers of 2, like memory

    26. Re:Newsflash... by CastrTroy · · Score: 1, Troll

      And again unfortunately, the only way to get good performance out of those designs is by explicitly coding for parallel processing. And that's hard for humans.

      I took a Parallel programming course in university. It was hard. Most students didn't get it. I got an A+ in the course. The hard part is that you really have to forget everything you have learned about programming on a single processor. You have to use completely different algorithms. All the regular algorithms follow a straight line. Parallel processing doesn't work this way. For most applications it's more trouble than it's worth. But for those complicated problems that really need it, it can be worth it. It will probably remain away from the common desktop for a long time to come. But it does have its uses.

      --

      Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
    27. Re:Newsflash... by walt-sjc · · Score: 1

      Um, you may want to check your math. MHz != GHz.

    28. Re:Newsflash... by Anonymous Coward · · Score: 0

      Intels new chipsets use a dual-bus FSB and alleviates the memory bandwiddh problem pretty much. Look at IBM's Xeon MP TPC-C benchmark...

      Thing are going to change pretty much when FB-DIMMS gets introduced.

    29. Re:Newsflash... by walt-sjc · · Score: 2, Insightful

      While that's true, these dual core chips (especially Intel's lame single-memory bus design) really seem targeted towards the desktop market where the impact is greatest, yet cost differential is realativly small (in relation to the total price of a system including software.)

      I'm more interested in what IBM's Cell processor can do. While some problems are definately single threaded by nature, the majority are not. I have a GIS application that could definately benefit from as many processors as I can throw at it (although there comes a point where you need more memory, disk I/O etc. too so you still need to spread them out over a cluster to some degree.)

    30. Re:Newsflash... by beezly · · Score: 2, Interesting

      Dual Bus memory still doesn't cut it.

      Each Opteron has a dual bus memory controller on board (granted, only DDR400)... but as I increase the number of CPUs, the number of memory busses increases.

      The 4 CPU boxes I'm working on have 8 independent DDR400 memory busses.

      The only obvious way to get your memory bandwidth to scale is to have your memory controllers per CPU (or even better, on the CPU itself).

    31. Re:Newsflash... by InvalidError · · Score: 1

      Same clock rate?

      The fastest AMD chip for the near future is 2.6GHz while the slowest P4 for the last year or so is 2.8GHz. If you want to compare based on clock speed, the Pentium-M (aka P3-v2) is a much fairer comparison. It has been a well known fact since the P4's launch that the P4's IPC (instructions per clock) sucks when compared to the P3's. (And even more so when compared to the PM's.)

      I have both an A64-3000+ and a P4-3G. My typical workloads usually contain a number of non-trivial tasks. While my A64 does complete most single tasks faster than my P4, it is nowhere near as smooth-running once non-trivial tasks start piling up while I am doing (or trying to do) anything serious.

    32. Re:Newsflash... by MuMart · · Score: 2, Insightful
      On chips with built in memory controllers, as you increase the number of cores on a chip the memory bandwidth per core decreases, however as you increase the number of chips in a system, the memory bandwidth per core remains the same and the number of cores increases.

      Even in a multi-memory controller system the same physical memory is shared, so there has to be some performance hit when running more than one cpu, so I doubt the actual memory bandwidth per chip will be the same. Or is there some architectural trick that removes that bottleneck in practice?

    33. Re:Newsflash... by beezly · · Score: 2, Informative

      That's NUMA - if your OS is NUMA aware it should try to place processes on the same processor as the memory that contains their data.

      But yes, you're right, processes accessing memory on a different processor will suffer a latency (and to some extent bandwidth) hit. A well designed OS will help to mitigate it to some of the extent, but it's one of the reason that CPUs don't scale linearly.

    34. Re:Newsflash... by MightyMartian · · Score: 1

      I've had little experience running SMP desktop systems (did briefly have an old dual 233mhz Pentium machine running Linux but it had problems with my PCI wireless card), but I have been working with them in the server environment for about seven years, and in applications like web servers and database engines, they most definitely make a substantial difference.

      --
      The world's burning. Moped Jesus spotted on I50. Details at 11.
    35. Re:Newsflash... by UltimateRobotLover · · Score: 1

      Windows does it very well without manual affinity adjustment. However, for tuning processes, you can use affinity to take advantage of CPU cache, or to ensure that an application can never take over the entire system.

    36. Re:Newsflash... by fitten · · Score: 1

      While that's true, these dual core chips (especially Intel's lame single-memory bus design) really seem targeted towards the desktop market where the impact is greatest, yet cost differential is realativly small (in relation to the total price of a system including software.)

      I hate to break it to you, but AMD's dual core processors only have one memory bus coming out of them, too. It isn't until you get to Opterons where you can have memory hanging off each dual-core CPU.

    37. Re:Newsflash... by Anonymous Coward · · Score: 0

      Get over it. It was a typo.

    38. Re:Newsflash... by quarkscat · · Score: 1

      Spot on target!

      Some people really just don't "get it". As a relatively early (1993) adopter of Wintel SMP, I can tell you that a multithreaded OS combined with some multithreaded apps can be significantly faster than single processor systems. (I used SMP to improve throughput in Photoshop, and rendering speed in 3D-Studio.)

      Since 1993, I have been doing BYO with computers, and have found that, except for gaming, SMP (combined with SCSI disks) has proved to be a pleasant experience. Multitasking is limited mostly by the amount of memory in the system. SMP seems to provide for a "smoother" user experience (not having run benchmarks for quite some time).

      CPU speed alone is no way to judge the value of a system -- something Intel has been far too quick to tout for many years. Throughput is what really counts. Differing CPU architectures (MIPS/SPARC/INTEL) and system design play a far larger role than the clockspeed. I have seen moderately large databases choke on a single CPU Intel box that would fly on a significantly slower clocked 64-bit SMP SPARC or MIPS computer.

      IMHO, I am quite glad that Dell has decided to stay out of the AMD business, since the HP and SUN Opteron systems already work quite well. Dell is the "Wal-Mart" of computer hardware vendors, and I don't think they could do the Opteron (or dual core Opteron) justice, except lowball pricing.

    39. Re:Newsflash... by DigiShaman · · Score: 1

      Being that a Hyper Threaded P4 shares the cache among both threads, I suppose manually setting the affinity wouldn't make a difference. Am I correct?

      --
      Life is not for the lazy.
    40. Re:Newsflash... by dascandy · · Score: 4, Informative

      Actually, the opposite. Two processes that communicate quite heavily SHOULD be run together on two processors, especially since they will share memory and thus cache lines, plus they can spend time spinning on a lock instead of swapping threads. Given short locks and equal speed, they can work a whole lot more efficient on a dual-core than on a single-core.

      FYI, it's called Gang Scheduling and has been described for quite some time.

    41. Re:Newsflash... by grumpygrodyguy · · Score: 1

      yeah it is a bit dull to anyone who has been following dual core or even just SMP.

      Cue loads of stupid people saying this technology is crap because most apps aren't multithreaded and the clockspeeds are lower than their current cPU.


      Please excuse my ignorance, but I've never really 'got' SMP. How exactly does it work in a typical desktop environment? How do jobs get scheduled across the two seperate cores/CPUs in such a way that it maximizes the available resources? Doesn't there need to be some kind of 'advanced scheduling' to sort through the hundreds of threads in a typical PC desktop?

      How does the dual-core know I want core A to handle anti-virus, anti-spyware, temp monitoring programs etc...and that I want core B to assume the 'important' applications like games etc?

      Does core A just 'fill up' to capacity and then core B 'takes over' the remaining work? Do different programs and sub-threads have differing priorities? If so, what determines these priorities?

      *genuinely wants to know*

      --
      The government has a defect: it's potentially democratic. Corporations have no defect: they're pure tyrannies. -Chomsky
    42. Re:Newsflash... by Anonymous Coward · · Score: 0

      parallizable? I look forward to seeing that in the dictionary one day.

    43. Re:Newsflash... by UltimateRobotLover · · Score: 1
      I think you're right... You might even lose the advantage that Hyper Threading gives you!

      A fairly decent explanation is here.

    44. Re:Newsflash... by webhat · · Score: 1

      I know, as I wrote a linux implementation and send it to the guy who wrote the CPU affinity code for the linux kernel.

      --
      'I am become Shiva, destroyer of worlds'
    45. Re:Newsflash... by StormKrow · · Score: 1

      who cares?

      It's still faster thhan 1Ghz CPU working on the problem alone.

      So it's not twice as fast? Big deal. It's fairly common knowledge that youre processing power isn't the sum of your processor(s) speed. It's still faster than one CPU alone, and that's what power users want. (Throw in dual CPU dual core boards, and the super geeks will be lined up around the block.)

      --
      Who cares about the ozone layer?...thanks to CFC's I can write my name......IN CHEESE!!!
  2. Ideal for developers by cerberusss · · Score: 0, Redundant

    It's a shame they just upgraded our PCs. I'd *love* to get ahold of one of these. We're developers and running Eclipse and JBoss locally (we have servers, but local installations can be useful) and a dualcore could speed things up.

    --
    8 of 13 people found this answer helpful. Did you?
  3. A matter of time. by Renraku · · Score: 5, Insightful

    How long before applications start figuring that they should have an entire core dedicated to them?

    Windows, for example. What if the next version of Windows requires a dual-core processor to be usable? You know..Windows gets one core to idle at 80% of its capacity..and spills over into the other core when loading a text file.

    If things stayed the way they were now, and the entire other core could be kept separate from the OS and used for gaming/other applications, it would be a great idea.

    But guess what.

    --
    Job? I don't have time to get a job! Who will sit around and bitch about being broke and unemployed then?
    1. Re:A matter of time. by gbjbaanb · · Score: 1

      There's no way you can dedicate a CPU to a particular application.. not in any form of pre-emptive OS.

      However, you can constrain an application to a particular CPU (in windows at least) - task manager, set affinity. That's a great way of preventing an application from using your other CPU. If you want a CPU to run a game only, you would have to go through the entire process list and set the other processes to CPU 1, (or write an app to do that), and then set your game process to CPU 2.

      I think you'll get more performance just by letting the OS handle things though.

    2. Re:A matter of time. by samael · · Score: 1

      Windows is currently idling at between 4 and 7% of CPU on my PC. Admittedly I'm running a 2.66GHz machine, but then again, according to the processes list that CPU tiem is mostly being taken up by a couple of apps sitting in the background (and 1% by me typing in IE).

      Why on earth you would want to allocate an entire CPU to that, I have no idea.

      Now you might want to allocate a whole CPU to Doom3 or HL2, but I suspect they'll pretty much get that anyway, as applications are assigned to the quietest CPU, as far as I understand it.

    3. Re:A matter of time. by kromozone · · Score: 1

      I had to stop reading when you mentioned you were using IE.

    4. Re:A matter of time. by samael · · Score: 1

      I am at work. Installing my own software is a sackable offence. I have no choice what browser I use on the desktop.

      And the browser I use makes _no difference_ to the truth of what I was saying.

    5. Re:A matter of time. by jtshaw · · Score: 3, Insightful

      That would be a total waste of CPU time.

      Very few applications, and OS's in particular, are idle most of the time. I don't know the exact profiling characteristics of Windows, but I do know that in linux the kernel rarely, if ever, takes up 100% of a CPU's, and never does for a prolonged period of time.

      If you locked one CPU and made that for OS tasks only you'd be wasting a lot of clock cycles that another application could happily use. Same would go for locking just about any application to a cpu.

    6. Re:A matter of time. by kromozone · · Score: 0

      /sarcasm

    7. Re:A matter of time. by mattdm · · Score: 1

      There's no way you can dedicate a CPU to a particular application.. not in any form of pre-emptive OS.

      What'd'ya mean "any form of pre-emptive OS"? Just because Microsoft doesn't do it doesn't mean it's not possible. You can certainly do this on Irix, for example. And I haven't looked at the Linux processor set tools, but I assume it's similar.

    8. Re:A matter of time. by darrylo · · Score: 1
      There's no way you can dedicate a CPU to a particular application.. not in any form of pre-emptive OS.
      You haven't seen version 31415926 of Clippy, have you?
    9. Re:A matter of time. by 4of12 · · Score: 1

      If you locked one CPU and made that for OS tasks only you'd be wasting a lot of clock cycles that another application could happily use.

      Probably wasted in most situations.

      But I wonder whether an OS sitting on its hands ready to go in a multiple core machine might be useful in soft realtime applications that need a little improvement in latency.

      --
      "Provided by the management for your protection."
    10. Re:A matter of time. by wisdom_brewing · · Score: 1

      not only would it be a waste of time to allocate an entire cpu to a game, it could potentially slow the game down, especially if the game requires more than the computing power provided by one of the processors. windows appears to be inefficient with apparently high idling cpu use, but thats just for show, it keeps some in reserve for each application etc.

      just control priorities and youve got no problems

    11. Re:A matter of time. by bjb · · Score: 1
      I think you'll get more performance just by letting the OS handle things though.

      For the most part, yes, this is true. However, one thing about letting the OS handle the affinity is that it may move your thread around to other processors. I've found a VERY slight speed improvement running some jobs by tying them to a single CPU; the idea is that the cache of the processor may still be relevant for your thread.

      Of course, your L1/L2/L3 cache has to be large enough and there shouldn't be a high contention for CPU resources that would effectively flush the cache clean of your data.

      --
      Never hit your grandmother with a shovel, for it leaves a bad impression on her mind...
    12. Re:A matter of time. by pnatural · · Score: 1

      That would be a total waste of CPU time.

      Yes, yes. This is Windows we're talking about.

  4. Well... by X0563511 · · Score: 4, Interesting

    I'm still bumming around with a sub-gigahertz chip, specifically an Athlon T-Bird. I've been out of the loop for too long, can anyone tell me the benifits of using a dual core system (and while we are at it, a 64-bit chip)? Any problems to look out for if I decide to jump on the wagon in my next upgrade?

    --
    For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
    1. Re:Well... by FidelCatsro · · Score: 2, Informative

      For the average user , if i were to be totaly honest. Right now there is hardly any real use for either. Duel cores would probably help the system apear faster if the average user is switching around alot of programs ubt for the price you would pay then it is not worth your time.

      64-bit well um if the average user um well runs a massive database setup but it will be more usefull soon in the x86 world (athlon 64 procesors though are excelent because of the onboard memory controler and architecture).

      For the power user , well i don't think we need to explain this to the power users, In a server enviroment duel core and 64-bit computing is wonderfull for many many reasons also in workstations for many many enginering jobs.

      --
      The only things certain in war are Propaganda and Death. You can never be sure which is which though
    2. Re:Well... by shyampandit · · Score: 2, Informative

      Mainly the difference would be found when running many apps at once. For example if you are ripping songs and playing a game simulataneously then it would be faster than a single core machine. I run many programs like proxy servers, mail servers etc. for the home LAN and also use it for games. So in this situation dual core will help me run the game lag free.

      Although for the speed boost to materialize in games they will have to be coded to use both cores, so one dosent just idle away.. When more programs get SMP aware then dual core would be great!

      The only problem righ now is, single core procs like amd64's can beat the crap out of dual core when tested mainly in single app environments because they have a higher clock speed.

    3. Re:Well... by tomstdenis · · Score: 4, Informative

      AMD64 carries more than just "bigger registers". It has more of them and the actual core is an overall improved K7 process with

      - Slightly longer scheduling buffers
      - 128-bit L1 cache bus
      - Larger instruction window (means it can feed the alus better when constants/etc are found)
      - more registers [and they're bigger]

      They also run cooler and takes less power than their k7 brothers.

      Tom

      --
      Someday, I'll have a real sig.
    4. Re:Well... by Anonymous Coward · · Score: 0

      Odd thing to say since the grandparent post only mentioned the memory controller and the Architecture .
      Those things are im fairly sure part of the Architecture , you don't work for AMD by any chance do you .

    5. Re:Well... by JollyFinn · · Score: 4, Informative

      The 64bit is for anyone with more than 2Gb of RAM + x86-64 gives you more registers besides being 64bit so it speeds up the recompiled code.

      Dual core means simply you have TWO processors running. Rember old reviews on SMP dual celeron A and other such reviews. It gives little for games, lots for certain multithreaded applications. As you have two processors running and doing things. And multitasking applications, like being able to run interactive application (doom 3), while system is doing some multihour compilation on background.
      Anyway, it mainly keeps system more responsive when you have some thread or application takes CPU.
      Also with lesser degree helps in some other similar situation, where CPU is tied up with something EXACLY same moment you would wan't it to deal with UI stuff.

      --
      Emacs is good operating system, but it has one flaw: Its text editor could be better.
    6. Re:Well... by tomstdenis · · Score: 4, Informative

      .... Me work for AMD? Ha!

      No, I'm just a happy loyal user. I have both a Prescott P4 3.2Ghz and an AMD64 Newcastle 2.2Ghz...

      For what I do [building software] the AMD64 smokes the P4 ... and does it without getting to 50C or so...

      The AMD approach is just common sense. Be more efficient at what you do and gradually do it faster. Intel went the market route and said "slow clockrate is for pansies!".

      So you end up with a cpu that has a higher clock rate but it doesn't win because the efficency is too low.

      AES on my AMD64 ranges around 260 [or so] cycles/block. On the P4 with Intels compiler I get around 410 cycles/block. If you scale 3.2Ghz to 2.2 Ghz that's still effectively 281 cycles [at 2.2Ghz]. Doesn't seem like much but keep in mind to get this speed they had to draw more power and run at a higher clock rate.

      I did a benchmark a week ago where I built LibTomCrypt with/without hyperthreading and it took the prescott with hyperthreading at 3.2Ghz to even come close to matching the AMD64 speed. That's only on ~45,000 lines of code.

      Now multiply that by say five or ten to get a larger project.

      I'm not saying the Prescott isn't a neat design. Overall it's efficient enough to be useful. Just the AMD64 eats it's breakfast and spanks it's mother is all I'm saying. ;-)

      Tom

      --
      Someday, I'll have a real sig.
    7. Re:Well... by Karaman · · Score: 2, Informative


      Hi, my older PC was a T-Bird@850Mhz with 256 RAM, 160GB HDD- PATA133 (CPU was working at 50 deg C)

      Now I have Athlon64 3000+ (233x8 = 2000MHz) (s939) with 1GB RAM, 200GB HDD- SATA150 (CPU does not go beyond 37 deg C)

      The difference is that with the older pc I compiled and installed LinuxFromScratch in 4 days (well I drank a lot of caffeine products),
      while when I switched to the A64 PC I did the job IN ONLY 4 Hours!

      Unfortunately I was unable to compile a stable x86_64 toolchain to complile a x86_64 Linux and now I have i386/i686 GNU Linux but hell it is fast and I love it!

      --
      sex is better than war!
    8. Re:Well... by Teemu+Alviola · · Score: 0

      I guess the older T-Bird would have been a lot faster with more memory.

    9. Re:Well... by rikkus-x · · Score: 1

      The parent tried to show the difference in speed between a 32-bit and a 64-bit CPU by comparing an Athlon 'Thunderbird' with an Athlon 64, where they had different clock speeds (850Mhz vs. 2000Mhz), a significantly different amount of (possibly different speed) RAM, and different hard disks.

      This was then modded Informative.

      WTF?

      Rik

    10. Re:Well... by Karaman · · Score: 1

      Well I didnt only change the clock speed of CPU, I changed to DDR400 + AMD Hyper Transport. I also pointed that for compilation you dont need too fast hdd transport, you need latter when you have a BIG DB and not enough memory to cache it.

      --
      sex is better than war!
    11. Re:Well... by Karaman · · Score: 1

      No it did not go as much faster! About the same as with 256! After all I had 100% free of 1GB swap every 100 ms (sorry, cannot attach CPU/mem log) But when using processes that needed more ram, the compile process actually ran ot ram and used swap :)

      --
      sex is better than war!
    12. Re:Well... by walt-sjc · · Score: 1

      Dual core means simply you have TWO processors running.

      AFAIK (and I may be wrong, someone please correct) the intel dual core shares one memory bus, where in most true dual CPU systems they don't. There may also be other bus sharing issues...

      On the other hand, SMP may be more efficient with both cores on the same chip.

      It would be very interesting to benchmark a dual Intel CPU machine against a single CPU dual-core machine running at the same frequency, etc.

    13. Re:Well... by fitten · · Score: 1

      I prefer my cores to not fight each other (duel) but to cooperate instead (dual). I tend to get more work done that way.

    14. Re:Well... by Anonymous Coward · · Score: 0

      ......... Ive heard funnier, Chicken road crossing jokes for example

    15. Re:Well... by fitten · · Score: 1

      In the x86 world, up until the Opterons, all the CPUs shared the same bus (Athlons, Pentiums, Pentium Pros, Pentium II Xeons, Pentium III Xeons, and Pentium 4 Xeons). There is nothing "true" or "untrue" about this architecture. It just is an old architecture but it is inexpensive (comparatively, which is why it is done this way. Unfortunately, it is also far from optimal. It works and typical speedups are in the 1.5X to 1.7X range in good cases. More "exotic" methods can give more efficient architectures (crossbars, NUMA, etc.).

      Dual CPU system is simply a system with two CPUs in it. It either has two CPUs or it doesn't. There is no "true" or "false" dual CPU system.

      Similarly with dual-core. Some designs may be better than others, but dual-core means that there are two cores on one piece of contiguous silicon. If this is present, it is dual-core. If it isn't, then it isn't dual-core. Efficiency of design is not a part of that definition.

    16. Re:Well... by kb7oeb · · Score: 1

      You might give the 64bit version of gentoo a try if you like compiling from source. They have it set up so you can still run 32bit binaries easily like quake3 or realplayer. I've been running it for about 6 months now with few problems.

    17. Re:Well... by zaq121 · · Score: 1

      He answered a question (click the parent button on his post), the parent specifically said he had a sub ghz tbird machine, what would he benifit by upgrading. The answer was 'a lot'.

    18. Re:Well... by evilviper · · Score: 1
      They also run cooler and takes less power than their k7 brothers.

      You should know that's not technically true. K8 processors seem cooler because of CnQ motherboards. The K7s are actually lower power (not by a big margin, but still) but unfortunately, AMD went with a stupid power management scheme, where the motherboard has to have the northbridge disconnect (S2K issue) or the CPU won't idle down at all. Unfortunately, practically no AMD motherboards do this, and the software approach (look up fvcool) seems to only work for about 50% of AMD motherboards in my experience.

      That is the biggest mistake AMD ever made, IMHO, which caused their CPUs to be ridiculed for being massively hot. Their only saving grace was the fact that Intel made a much stupider mistake with the Pentim-4 being so massively power-hungry, that even with proper idling of the CPU that AMD doesn't have, the P4 still blew by the power requirements of the K7s.

      There's more info in a previous journal entry of mine.
      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    19. Re:Well... by tomstdenis · · Score: 1

      "You should know that's not technically true. K8 processors seem cooler because of CnQ motherboards."

      You do realize that CnQ is just marketing for what the mobile K7's already had right? that is, clock modding.

      When you are at full load the cpu won't be clocking down or idling. So CnQ doesn't do jack shit.

      The reason the K8 is cooler is

      1. More surface area on the chip for heat dissipation.

      2. Different fabs process (SOI), also lower voltage

      I don't know how much both of those contribute... but given that I've used quite a few K7s in my time I know for a fact they suck back the power like there is no tommorow.

      Tom

      --
      Someday, I'll have a real sig.
    20. Re:Well... by evilviper · · Score: 1
      You do realize that CnQ is just marketing for what the mobile K7's already had right? that is, clock modding.

      The mobile processors had some of those features, but CnQ has more. Bus, multiplier, voltage, fan speed, etc. But we are talking about desktop processors anyhow, and I haven't seen any with PowerNow support, have you?

      1. More surface area on the chip for heat dissipation.
      That doesn't really hold up. Once you put a heatsink on a K7, it has plenty of surface area too... The K8 just has an extra layer of metal to protect the core, but that technically makes it just a bit hotter.

      also lower voltage

      That's nice and all, but the lower voltage doesn't mean it runs cooler when the chip is still rated at ~89watts. The Max Power Dissipation for K8s is just simply higher than K7s. With the exception of a couple lower-clocked (eg 1GHz ~30watt) K8 models.

      I've used quite a few K7s in my time I know for a fact they suck back the power like there is no tommorow.

      Like I said, it's because the motherboards you've used didn't have proper power management for your AMD CPUs. On any of the K7s you have, install fvcool and run it ("-e -i") and if you're lucky, you'll see the CPU drop in tempurature, and your whole system may even drop to half-power.

      This brings up another good point though. Ever since the introduction of DDR RAM, Northbridges on AMD motherboards have been running incredibly damn hot... The exact same components in a different motherboard can make a difference of 40+watts. Since it's exactly the same components, you know that means the northbridge is using up 40watts more power, and it's putting off the heat to prove it! With the integrated memory controller on K8s, that might have made a significant difference in heat as well.

      If you want something that really draws power, get a Pentium 4 EE.
      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    21. Re:Well... by tomstdenis · · Score: 1

      Well I own both a Barton 3200+ and a NewCastle 3200+ and the Barton [even with athcool which does the same thing as fvcool] it still runs fairly hot.

      That's totally beside the point ... cuz at FULL LOAD there is NO IDLE TIME.

      When both [2.2Ghz] cpus are at full tilt the K8 is cooler by at least 10C. Both have Thermaltake HSFs and the K8 has the "SilentBoost" variant [which is very quiet].

      Tom

      --
      Someday, I'll have a real sig.
    22. Re:Well... by evilviper · · Score: 1
      When both [2.2Ghz] cpus are at full tilt the K8 is cooler by at least 10C.

      Ah ha. I'm glad we've gotten down to the actual technical details.

      There's two possibilities... Either your two motherboards could have tempurature sensors that are 10C different from each other. Or, perhaps you were just lucky, and got the 1.4V Newcastle Athlon-64.

      Athlon XP-3200+ (6-A-0 2.2GHz Barton) 1.65V 76.8W
      Athlon 64-3200+ (S754 - 2.2GHz - 512KB) 1.5V 89W
      Athlon 64-3200+ (2.2GHz - 512KB) 1.4V 72W
      Sempron 3300+ (2.0GHz - 128KB) 1.4V 62W
      (Both Athlon 64-3200s are listed as Newcastles)

      Yours seems to just be a case of you having an old K7 and a newer K8. If you had a new K7 (Sempron) at the same speed, it would be even lower power than your K8.

      I see the same thing very often with Pentium/Celeron and Athlon/Duron processors. Because the "economy" line of processors is made months later than the "performance" version, the Celerons/Durons are just about always lower power.

      I'm glad you brought it up though, because it's been a while since I looked-up the specs. It seems AMD has been improving their K8 processors significantly, and moving their power requirements down, closer to matching their K7 counterparts.
      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    23. Re:Well... by tomstdenis · · Score: 1

      it's a 1.5V NewCastle. As for "old K7" ... Barton cores are not new. Sempron [754-pin and up] cores are based on the K8 design not K7. They just have the x86_64/extra cache disabled. ...

      You really ought to know what you are talking.

      The point is even with the bus disconnect the K7s still idled warmer than a K8 idles. More to the point at full load the K7s are hotter than the K8s.

      Maybe the K8 takes more power [I personally don't think so] but it certainly doesn't get as hot. Either that means it dissipates heat quicker or it just makes less heat [more efficient] who knows.

      Tom

      --
      Someday, I'll have a real sig.
    24. Re:Well... by evilviper · · Score: 1
      You really ought to know what you are talking.

      A simple mistake... The Sempron in question was listed along with other Socket-A Semprons, so I didn't even suspect it might be a K8 version...

      The point is even with the bus disconnect the K7s still idled warmer than a K8 idles.

      A little bit warmer is to be expected, as the two are close in power rating, and lowering bus speeds, voltages, etc. like CnQ does is obviously going to save power over simple processor idle states.

      More to the point at full load the K7s are hotter than the K8s.

      Your single anecdotal case isn't quite enough evidence to make such a blanket statement. Especially since you're just going off the motheboard sensor readings, which can vary greatly.

      Maybe the K8 takes more power [I personally don't think so]

      You don't have to believe me, just look up AMD's specs on your processors for yourself.

      or it just makes less heat [more efficient] who knows.

      No, that's not a possibility. If it's taking in 89watts of power, it's outputting 89watts of heat... That's just how electricity works.
      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    25. Re:Well... by tomstdenis · · Score: 1

      The Socket-A semprons are K7, the 754-pin are K8 based.

      The power consumption is based on transistor count, frequency and switching probability. They don't say how they calculate TDP. It might be if ALL transistors switched at clock rate. It might be if 50%, it might be if 65%.

      As for my "one anecdotal case" ... I've been using AMD processors since the K6-2 started shipping. I've used quite a few K7s and I have one K8. you don't need a mobo sensor to tell the K7 is hotter... just put your damn hand on the heat sink. At full load you can really feel the heat from several inches away on a K7.

      The point is I actually own a K8 and I know for a fact it runs cooler at idle/load than my previous K7 boxes.

      Like I said I chalk it up to more efficient heat dissipation (heat spreader, larger HSF, etc...) and more efficient transistors (lower voltage, less leak, etc...).

      Tom

      --
      Someday, I'll have a real sig.
  5. Uber amounts of RAM by Christopher_Hansen · · Score: 1

    allows for more than 4gb of ram. That is all the advice I can give you on that subject. The article should answer your questions about dual core.

    1. Re:Uber amounts of RAM by Stocktonian · · Score: 1

      allows for more than 4gb of ram.
      So do many 32 bit cpus with the right OS. What a 32 bit processor can't do as well is handle 64 bit intergers etc...

      --
      XePhi Computers sell really cheap Linux CDs! http://www.xephi.co.uk
    2. Re:Uber amounts of RAM by norton_I · · Score: 1

      The real issue is "allows for more than 4 GB of address space". PAE allow for physically addressing up to 64 GB of memory (albiet with a performance penalty), but an application can only address 4 GB at a time. This may be a limiting factor for database and scientific applications, and also is inconvenient if you wnat to mmap large files (such as block devices or virtual drives) for other purposes.

      Furthermore, in massively multi-threaded applications, you can run out of virtual address space for the program stacks. The spacing of the stack base address must be greater than the largest stack any thread will ever have, which if you have thousands of threads can use up a lot of virtual address space. I leave it as an excercise whether this sort of application architecture is ever a good idea.

      A 64 bit OS running on a 64 bit CPU allows more efficient access to all of physical memory + gives applications the ability to address much more virtual memory (which can be useful even if physical memory is 4GB).

  6. Well? by Anonymous Coward · · Score: 5, Funny

    Does this mean my Windows XP machine wont pause when I put in a floppy or Cdrom? Wow, sign me up.

    1. Re:Well? by EpsCylonB · · Score: 4, Funny

      no but it will make your internet faster.

    2. Re:Well? by Anonymous Coward · · Score: 0

      You can say that again. Makes you wonder what
      crack some developers were smoking to let that
      behavior in to the wild.

      Does anyone else see a lengthy delay after loging
      in before etheret/internet connections are properly
      established? Or is it my crappy netgear wireless
      hardware.

      - Moomin

    3. Re:Well? by Anonymous Coward · · Score: 1, Informative

      Sad when its funny because its true. Windows pauses horrible when waiting for resources. Outlook/office when searching for printers, network shares loading for first time, removable devices, file explorer scanning a directory first time, etc.

      Systems should never pause this bad, horrible design.

    4. Re:Well? by Anonymous Coward · · Score: 0

      Or better, if you are still using Outlook/Exchange, then you'll be able to listen to MP3s without studdering due to Outlook taking 100% of the CPU every 15 seconds!
      Maybe.

      I've been buying a bunch of dual core Power5 and UltraSPARCiv servers the last few months. The Power5 arch seems to do a much better job in the dual core area. No experience yet on Intel or AMD dual cores. spec.org usually has SPECInt_Rate2000 info for most servers. They don't have some of the newer CPU speed tho.

    5. Re:Well? by grouse · · Score: 1

      Not me. But I do frequently get pauses when I put in a CD :(

    6. Re:Well? by kryptkpr · · Score: 1

      I've seen what you describe happen, and it is usually due to shitty network card drivers.

      My personal favorite is the "Saving your settings" wait upon shutdown. After you've installed a few apps and run the system for a while, this can take a few MINUTES.

      The CD-ROM thing the GP mentioned is also hella annoying.. Floppies aren't as bad as they used to be under the 9x windows though (want to format a floppy? can't do ANYTHING else!)

      --
      DJ kRYPT's Free MP3s!
    7. Re:Well? by MattHaffner · · Score: 2, Funny

      Internets, you mean. Multi-core and all...

    8. Re:Well? by Duhavid · · Score: 1

      What I find odd is this: I lock my machine when I leave work for the day. When I come in the next morning, XP spends about 30 seconds repainting *everything*. Why?

      --
      emt 377 emt 4
    9. Re:Well? by kabocox · · Score: 2, Funny

      Does this mean my Windows XP machine wont pause when I put in a floppy or Cdrom? Wow, sign me up.

      Nope, that feature isn't scheduled until after we have 16 cores on chip, 32 Gb of RAM, 10 Terabytes of HD storage, and optical media is at 1 Terabyte per disc. They said it was a wierd hardware limit and it would require at least that much processing power for Windows XP to read a floppy or CDROM and do anything else. You don't even want to know what it will take for Longhorn to do that.

  7. Something missing by FidelCatsro · · Score: 4, Interesting

    What this test really was missing was a direct comparison to SMP systems which really for me makes the results entierly boring and expected .
    If he had shoved in a duel opteron set-up and a duel xeon set-up then it may have been a little more intresting , though as it stands its like stating the obvious.

    --
    The only things certain in war are Propaganda and Death. You can never be sure which is which though
    1. Re:Something missing by shotfeel · · Score: 1

      IMO what's missing is actually testing single vs. dual cores -all else being equal.

      I'm more familiar with the tools available on Macs, but given that there are simple utilities that allow you to turn one CPU off on a dual-CPU system, I assume the same is true on the Intel side and that they would also allow one to turn off one core on a dual -core system.

      Which always makes me wonder why benchmarks that are supposedly testing one vs. two processors (or cores) don't use these tools so they can actually test single vs. dual without a dozen confounding variables.

    2. Re:Something missing by PsychicX · · Score: 0

      I saw one benchmark comparing dual Opterons to dual core Opterons...the dual core was faster than conventional SMP, but only by a little bit.

  8. Fundamental question about dual core by Anonymous Coward · · Score: 0

    How do dual-core single cpu desktops stack up against dual-cpu desktop with similar mhz?

    Anyone have benchmarks?

    1. Re:Fundamental question about dual core by Anonymous Coward · · Score: 0

      I dont think mhz are an issue unless your comparing oranges to oranges .

      a 2* opteron setup @2.2ghz(perhaps 2.4) vs 2* xeon 3.2ghz vs the duel core at 3.2ghz and where its possible to compare a duel 2ghz(perhaps 2.5) PPC g5 setup.
      The oranges to oranges and an apple(opteron) thrown in there to make the test a little more fruity and just incase the test is too sweet i throw in a duel g5 setup to add some salsa.
      Now you have an intresting test .

    2. Re:Fundamental question about dual core by rodac · · Score: 2, Insightful

      In general and assuming a non broken cache architecture, a 2CPU/core solution will feel faster than a single cpu solution with twice the cpu frequency.
      The total number of cpy cycles are the same, but the average queue-length for a process waiting for the CPU is half, i.e. the latency before your process is scheduled is lower making it "feel" faster.

    3. Re:Fundamental question about dual core by Anonymous Coward · · Score: 0

      Ah, BUT, 1. the OS has to spend time deciding which CPU to give a task to (and in the case of dual CPU there's a chance of NUMA-style RAM footprint spillage between them) and 2. if the processes are not equally CPU-intensive the speedup you see with dual-core is limited to the simpler process.

      A single core with 2X the MHz will rip through everything faster no matter what percent of the time your dominant threads take.

    4. Re:Fundamental question about dual core by Anonymous Coward · · Score: 0

      Throw in some garlic too. BAM!

  9. it's nice to see by Adult+film+producer · · Score: 0, Flamebait

    that slashdot is on the intel payroll too. people talk about the giant marketing machine intel operates, i just wish all of their lackeys would be a little more open about their financial relationship with the cpu juggernaut.

    oh btw, amd made a real dual cpu release yesterday... You can actually buy the chip, unlike intel's paper release just a few days ago (ya, that was a really desperate marketing effort.) But at least slashdot gives intel the last laugh.

    1. Re:it's nice to see by Anonymous Coward · · Score: 0

      Fanboy type arguments , the world has room for many difrent procesors and no need to go around being a dick about your oposition.

  10. dual-core SMP? by Anonymous Coward · · Score: 0

    Does (or when will) there be support for an SMP setup using these beasts? I'd just love to see how a 4-core/4SMT system performs on rendering.

  11. Anandtech by iamthemoog · · Score: 5, Informative

    Has the new dual core opteron up against a quad Xeon with 8MB cache, amongst many others.

    Well worth a read:

    http://www.anandtech.com/cpuchipsets/showdoc.aspx? i=2397

    --
    No Norm, those are your safety glasses; I'll wear my own thanks...
    1. Re:Anandtech by astro-g · · Score: 2, Informative

      I love how he says the only difference between the 8xx operons and the 1xx opterons is the ammount of validation testing each chip gets.

      Umm, what about the number of available HT channels?
      There is a reason you cant use the 1xx chips in the 8 was motherboard.

    2. Re:Anandtech by rob_squared · · Score: 0

      8MB OF CACHE?!
      I'm sorry, but I needed caps lock for that one. Why not just stick a DIMM in there and call it cache, that'll slow the sucker down nicely.

      --
      I don't get it.
  12. Java benchmark by Anonymous Coward · · Score: 0

    Does anyone has java benchmark numbers on dual cores?

  13. One thought I had... by Kjella · · Score: 4, Interesting

    ...and I'm not quite sure if it's a good one, but for desktops:

    The foreground program has a dedicated core. If you switch programs, put the old on the "other" core. The new moved from the "other" core. Essentially, your current program has full responsiveness (assuming you don't do things that lock up the application itself), no context switches, no other programs that can run some weird blocking call (on a single core machine, it certainly looks that way at least, especially CD-ROM operations).

    Granted you could end up with your fg processor being idle most of the time. But the way many people work with the computer, the foreground program is the ONLY time-critical application.

    Kjella

    --
    Live today, because you never know what tomorrow brings
    1. Re:One thought I had... by Anonymous Coward · · Score: 0

      First of all you'd kill the cache every time you swaped a process between the two cores. Second it doesn't work like that because global system resources still have to be shared between the two cores, and that means they have to be locked during access. Just like on a real SMP system, the operating system uses spinlocks to stop two threads on different CPU's from running the same critical sections at the same time, and it is still perfectly possible for a process on one CPU to block waiting for a mutex or semaphore. You'd probably waste a good 80% of the available processing time of the "dedicated" CPU because of this.

    2. Re:One thought I had... by shotfeel · · Score: 1

      I think you hit on one of the biggest problems with that. What if the foreground app doesn't need any real horsepower, but others do? Espceially if you have real-time processes (video/audio capture...) going that might get CPU starved.

      I think a much better way to handle making the foreground app more responsive would simply be to raise its priority level. That way it only hogs the CPU if it really has something to do.

  14. Useful benchmarks by Jumpin'+Jon · · Score: 1

    The benchmarks they've done would seem to give a good real-world example of the benefits of multi-core. As many have noted, "proper" multithreaded application of course see the biggest benefit, but they're relatively rare on most Desktop systems. Having additional CPU power available to simply dedicate to multiple applications does have benefits.

    Their tests, doing such things as Anti-virus scans while performing other cpu-intensive tasks (e.g. Doom 3 ;), really do give an insight into the kind of benefits regular users will see.

    Has anyone seen the various chip manufacturers roadmaps, with regard to the demise of their current ranges? My point is, I was amazed at how fast the AMD64 range replaced the Athlon XP. I wonder how long - realistically - it'll be before multi-core is "the norm"... less than a year?

    JJ

    1. Re:Useful benchmarks by shotfeel · · Score: 1

      wonder how long - realistically - it'll be before multi-core is "the norm"... less than a year?

      Interesting question. A big factor is one of the questions I have. Does dual-core have a price/performance advantage over dual-processor? Some factors:

      How do the yields compare? I may be wrong, but I'm guessing that the clock rate on dual-cores is lower because you just can't get a good yield at a higher clock rate. You have to go with the clock rate of the lowest performer in the dual-core.

      How does performance compare? I can see some performance advantage to the two cores being "closer", but if they're so close they have to share the FSB, what does that do to performance?

      Other factors?

  15. Will my Spyware support these? by Anonymous Coward · · Score: 2, Funny

    Or do I have to wait for Service Pack 3?

    Yours,

    Gator Fan.

    1. Re:Will my Spyware support these? by kesuki · · Score: 1

      Actually multi-core systems will make it harder for a 'user' to notice they have spyware... since a 'slowing down' of the system can often times be quite noticable when a spyware has hijacked you.
      With dual cores, spyware could potentially take up the better part of a core's processing power, and the luser would never be the wiser...

  16. Real comparison? by Barny · · Score: 1

    For those wanting a real idea of how good these things are (compared to other SMP server setups) http://www.theinquirer.net/?article=22711 has a good set of links.

    --
    ...
    /me sighs
  17. Sluuuurp..... by Diakoneo · · Score: 3, Interesting

    That last page raised my eyebrows. 291 Watts under load, that's some serious power draw compared to what I'm used to. And that had to be kicking out some serious heat, too.
    Anybody know what is the draw for a 4x Xeon system? I'd be interested in seeing how they compare.
    I wonder at what point the facilities people will want to use the server farm to heat the building, too. A weird convergence, the PC world is becoming more like the old mainframe world.

    --
    "Well..here I am..." - Jubal Early
    1. Re:Sluuuurp..... by Anonymous Coward · · Score: 0

      The Intel systems consistently draw more under load than the AMD ones. If you want to confirm it's in the Tech-Report review from yesterday I believe.

    2. Re:Sluuuurp..... by the+eric+conspiracy · · Score: 1

      I've crossed Intel designs off my list for that reason. If power is your concern, the Opteron dual cores will be available in low voltage designs that draw as little as 30 watts.

      Or you could wait for a dual core Pentium M.

  18. How 'bout some Adobe CS benchmarks? by Aqua+OS+X · · Score: 4, Insightful

    Who the hell runs benchmarks with FireFox and iTunes.

    if you ask me, the people that desperately need the ability to multitask are folks in the creative industry. Every 5 minutes bounce back and forth between massive applications rendering huge files.

    Nothing sucks more then opening a 400dpi photoshop document and not having InDesign respond since your single core CPU is being bogarted.

    SMP is probably the only reason I still find my crusty old Dual 450 g4 useful. It does things slowly, but it doesn't "feel" slow. If something is taking its sweet ass time, I can usually do something else without waiting years for windows and menus to draw.

    --
    "Things are more moderner than before- bigger, and yet smaller- it's computers-- San Dimas High School football RULES!"
    1. Re:How 'bout some Adobe CS benchmarks? by samael · · Score: 1

      Surely those kinds of apps are more memory bound than CPU bound?

    2. Re:How 'bout some Adobe CS benchmarks? by Jameth · · Score: 3, Funny

      You're dead on accurate with that one. I want a benchmark that will tell me what kind of performance I can expect if I have a logo I am editing in Illustrator that I open in Photoshop to clean up a bit and then insert into a document in InDesign while I'm trying to make it look similar in the webpage I'm putting together in TextPad, viewing both final documents through Acrobat, IE, FireFox, and Safari, all at the same time. (While listening to music.)

      And no, I'm not being sarcastic. Although I rarely do all of that at once, it has been known to happen. And don't even get me started about what happens when I have something compiling behind all of that. I'm just thankful, in a way, that since I don't do 3D work I'm not tossing Maya into that mix.

    3. Re:How 'bout some Adobe CS benchmarks? by LookSharp · · Score: 1

      Who the hell runs benchmarks with FireFox and iTunes.

      Someone who's paid by Intel, in money and/or product, to write a review whose conclusions match Intels marketing materials?

      That's not a guess, it happens pretty much every day. Some people are naive enough to believe that a website like this can make tons on cash on google ads and a couple of banners, but not folks that have seen how trade press reviews work.

    4. Re:How 'bout some Adobe CS benchmarks? by archen · · Score: 2, Informative

      Firefox might not be such a bad benchmark. Go a bunch of Japanese pages with status bar scrollers (in japanese) then open up about 20+ tabs. I do this every day, and the CPU usage on Linux can go up to a sustained 70-90% or more. If gcc is working in the background, the system can really lack responsiveness (and I'm on an AthonXP 3000 with 1Gb of RAM). Out of all the apps I use, Firefox and Mame take the cake in CPU usage.

    5. Re:How 'bout some Adobe CS benchmarks? by Anonymous Coward · · Score: 0

      in that case, why did amd win something like 20 of 22 benchmarks this time around?

      intel releases dual-core: pounds all single-core solutions

      amd releases dual-core: pounds all single-core solutions and is marginally faster than intel's dual-core

      maybe the lesson isn't that Anand is in anyone's pocket, but that DUAL-CORES ARE GOOD!

      face it, you're a dope for believing that inquirer article

    6. Re:How 'bout some Adobe CS benchmarks? by shotfeel · · Score: 1

      Who the hell runs benchmarks with FireFox and iTunes.

      Someone who wants to see how they handle the types of apps most people actually use?

      Nothing sucks more then opening a 400dpi photoshop document and not having InDesign respond since your single core CPU is being bogarted.

      Or your music drops out because you just clicked the link to a web page full of Flash animations and tables.

      Or a 'high priority task" like inserting a disk blocks the CPU.

      Even for home use, I prefer what on paper would be a slower dual-processor system to a slightly faster single processor. But then I tend to have a lot of stuff going on all the time and usually have one or more "real time" tasks going on (from listening to music to capturing video)..

    7. Re:How 'bout some Adobe CS benchmarks? by PsychicX · · Score: 0

      Yeah...but that's because Firefox is a memory and CPU black hole....have it render a large animated GIF one time, just see what happens.

    8. Re:How 'bout some Adobe CS benchmarks? by Anonymous Coward · · Score: 0

      We're talking about "PC Perspective" here, not Anandtech (who, along with HardOCP, I believe to be pretty ethical about their reviews).

      Secondly, just because AMD won most of the benches does not mean the review is any more meaningful. Look at everyone's "previews" of the Intel dual-core part, and note how the focus was on performance of desktop apps instead of games. I think the Inquirer "editorial rant" was pulled off poorly, but wasn't inaccurate.

  19. shared FSB (intel) or not (AMD); other benchmarks by coats · · Score: 4, Informative
    After reading the article, I realised that the frontside bus was shared. I didn't expect that. It seems to be a transitory solution in order to have the "first dual-core" CPUs on the market. When AMD releases theirs I expect them to have a superior solution.
    AMD64 has had the circuitry for dual-core on-chip memory controllers from the very first -- they just didn't have the second CPU core. For a good discussion of the differences, see http://www.linuxhardware.org/features/05/04/21/174 7217.shtml at LinuxHardware.

    For benchmarks relating to serious DB and web use, see this review by Anand Shempi: http://www.anandtech.com/cpuchipsets/showdoc.aspx? i=2397 or these two at FiringSquad: http://www.firingsquad.com/hardware/amd_dual-core_ opteron_875/ and http://www.firingsquad.com/hardware/colfax_dual_op teron/

    --
    "My opinions are my own, and I've got *lots* of them!"
  20. Re:shared FSB (intel) or not (AMD); other benchmar by dchallender · · Score: 3, Informative

    Extremetech have a review of AMD opteron dual core http://www.extremetech.com/article2/0,1558,1788685 ,00.asp?kc=ETRSS02129TX1K0000532

  21. Comment removed by account_deleted · · Score: 2, Informative

    Comment removed based on user account deletion

  22. It's bad news, actually... by pmadden · · Score: 5, Insightful

    I'll probably get flamed for this....

    Increased performance in CPUs has normally come from faster clock rates and more complex circuitry. As we all know, Intel (and the others) have bailed out on faster clocks. If you add more complex circuitry, the logic delay increases--to keep the clock rate up, you have to burn power.

    What does this mean? The old-fashioned ways of getting more performance are dead--if you try it, the chip will burn up. It's easier to build two 1X MIP cores than one 2X MIP core. Like it or not, dual cores are the only solution; with transistor scaling, we'll have to go to 4, 8, and 16 cores in the next few years. IBM went dual-core with the PowerPC in 2001. Intel, AMD, and Sun are just following suit.

    Not bummed out yet? Massive parallelism works well for people doing scientific computing, but for the average joe, it's useless. I don't care how fast a processor is--I usually have one task that will crush it--but rarely do I have two time-critical things to worry about at the same time. In the article referenced, they had to work hard to find things that would test the dual-core features. Parallel computing and multiple cores sounds great. History buffs will know about Thinking Machines, Meiko, Kendell Square, MasPar, NCUBE, Sequent, Transputer, Parsytec, Cray, and so on.... Not a happy ending.

    So.... we can't get more single processor performance without bursting into flames. And parallel machines are only useful to a small market. IMO, it's gonna get grim. (And before anyone says new paradigm of computing to take advantage of the parallel resource, put down the crack pipe and think about it--we've been waiting for that paradigm for about 40 years. Remember occam? I thought not.)

    1. Re:It's bad news, actually... by springbox · · Score: 1

      Maybe it's not that bad after all. With dual core systems becoming mainstream, maybe more application developers will try to stretch the design of their software so that more parallelism is introduced to get more performance out of systems with these setups. Remember that the majority of normal consumers have single processor systems at the moment.

    2. Re:It's bad news, actually... by Alioth · · Score: 1

      Grim? Hardly. Most users don't even need a 1.5GHz processor. We've got plenty of older boxes around here (733 MHz Compaq P3 systems) which are running the latest office software quite happily.

      It's only going to get grim for gamers - well, not even then. I'm sure Carmack and company will figure out a way of taking advantage of multiple cores and multiple GPUs in their future generations of games.

    3. Re:It's bad news, actually... by eddy · · Score: 3, Insightful

      Yeah, actually I think this might become a big boon for gamers (such as myself). The stuff that makes games interesting to me is AI (which is a very wide field certainly, but I think of such things as finally being able to use reasoning-engines (F.E.A.R is the first game I know of that use one), better pathfinding (AI can now use focussed D* instead of cheating with A*, etc) all of which will finally get to some love and tender care.

      With only one CPU, AI was always the ugly step child. "Yeah, sure.. we can give that 10% raster..." (okay, so I'm dating myself, anyho...). Now there will finally be CPU-power available that CAN NOT EASILY BE USED FOR TRANSFORM AND RENDERING (note "easily", not "possibly").

      In short; ubiquitous dual-core CPUs will revolutionize gaming.

      --
      Belief is the currency of delusion.
    4. Re:It's bad news, actually... by Taladar · · Score: 1

      Maybe that means designers of Wordprocessors and things like that (average Joe software) will have to be content with 2 GHz then and only designers of games, render- and similar programs where parallelization is possible can use more than that. Wouldn't be the worst development IMO.

    5. Re:It's bad news, actually... by pmadden · · Score: 1

      Should clarify the grim thing--it's going to get grim for IBM/Intel/AMD/... It'll be hard to sell a CPU with 4 cores if people never fully utilize the first two. And when these guys can't sell CPUs, lots of things will get interesting. No more push on Moore's law, for one.

      Games is really the only consumer app where multi-core works -- and even here, it's tricky. You don't want a general purpose CPU--you want something tuned for 3D graphics. IBM has some good news; they make all the video game console chips at the moment, and the CELL is very cool. If we want to continue Moore's law, we need more games. Help us, John Carmack. You're our only hope.

    6. Re:It's bad news, actually... by n0mad6 · · Score: 1

      Not bummed out yet? Massive parallelism works well for people doing scientific computing, but for the average joe, it's useless. I don't care how fast a processor is

      Speaking as a physicist who used 5 CPU years last week alone, just because it may not be "useful for the average joe" now, doesn't mean its not something worthwhile-- people like me will really be able to use dual-core chips (as the parent states) and will buy these things now.

    7. Re:It's bad news, actually... by SilentTristero · · Score: 1

      Not bad news at all for the video editing/effects and audio markets (After Effects, Avid, Discreet on Linux etc.). Most apps in that space are already multiprocessor ready, and hi def especially can soak up all the cycles you can throw at it.

    8. Re:It's bad news, actually... by Junior+J.+Junior+III · · Score: 2, Insightful

      If they can't ramp up hardware any more, the next revolution in computing will not be faster hardware, it will be cleaner, more efficient code. Personally, I think that there's a lot of potential left, if not with silicon, then with diamond wafer chips, or optical computing.

      --
      You see? You see? Your stupid minds! Stupid! Stupid!
    9. Re:It's bad news, actually... by Anonymous Coward · · Score: 0

      Should clarify the grim thing--it's going to get grim for IBM/Intel/AMD/... It'll be hard to sell a CPU with 4 cores if people never fully utilize the first two.

      It should have also been hard for them to sell 3+ GHZ CPUs when people don't even fully utilize 1 (spyware notwithstanding):-)

    10. Re:It's bad news, actually... by wgaryhas · · Score: 0

      That is actually more since 5 CPU years in a week is 260.7 cpu hours per hour, while if you were able to sustain that use for a week you would use 15.34 cpu years

      --
      "For every complex problem, there is a solution that is simple, neat, and wrong." - H.L. Mencken
    11. Re:It's bad news, actually... by ciroknight · · Score: 3, Interesting

      Well you're right about what you were saying, those words would arbit a good deal of flames. But everything has its place and there's a place for everything. Lemme explain.

      Clockspeed is the easiest race, if you want to think of the CPU industry as a continuous race. All you have to do to crank out a faster CPU is continually shrink the die (because smaller gates flip faster), and make sure that everything is arranged neatly on the chip. When you hit thermal walls like we are now, it's simply time to reduce the voltage, and shrink the die again.

      The only problem is, Intel's flagship for doing this now, happens to be one with a lot of baggage. The Netburst core design pretty much dictates there is to be at least two of everything, and both of them should be running all the time, especially if Hyperthreading is on. This effectively doubles your transistor count (though in reality it is less than that; there's only a single copy of bus administration, micro-op decode, etc). Keeping them on all of the time also helps jump the heat production.

      But here's a truth; their CPU clock game could still be running if they would like it to. The Pentium-M is still running extremely cool. Shrink it to a 90 micron core, use SOI, strained silicon, more of their substrata magic, and a healthy dose of SpeedStep, and you could see a Pentium-M hitting 3.5GHz clockspeeds that would put both the Athlon 64 and the Pentium 4 to shame. Sadly, to build this processor is to admit defeat with the Netburst core, and Intel's being very stubborn.

      On the other hand, I believe AMD's got some magic they haven't used yet up their sleeve. Though honestly I couldn't tell you what it is. There has to be a reason they aren't playing up the Turion more other than the fact it isn't scaling down as far as the Pentium-M can. I'm also surprised they're being so slow about ramping their clockspeeds, but this is probably just so their thermal profiles look superior to Intel's. A 3GHz Opteron could easily decimate a dual Xeon setup, but at the same time would probably produce just as much heat, and I think AMD would see that as a defeat.

      --
      "Victory means exit strategy, and it's important for the President to explain to us what the exit strategy is." G.W.Bush
    12. Re:It's bad news, actually... by sammck · · Score: 1
      If you add more complex circuitry, the logic delay increases--to keep the clock rate up, you have to burn power.
      That's only true if the complexity you add is serial. Circuit complexity that adds parallelism does not increase logic delay. By adding gates you can reduce logic delay -- think of flattened implementations of multipliers, for example.
      --
      sjm
    13. Re:It's bad news, actually... by shotfeel · · Score: 2, Insightful

      Interesting issues, but I disagree with you on some points.

      First, there's really nothing new about multiple cores IMO. Instead of "core" try the term "execution unit". CPUs used to have a single execution unit. Then something like an Integer Unit was added. Remember the days of math coprocessors? They soon got moved into the CPU by adding more, and more specialized, Integer Units and a Floating Point Execution Unit. Now you have CPUs that have multiple execution units, to the point of having multiple Integer and Floating Point Units, and even SIMD units.

      So then, isn't multi-core simply an easy way of doubling the number of execution units in the CPU and using the OS to handle some of the scheduling?

      I'm also not sure I agree with you on parallelism. Seems to me we've had to do a lot of work to make very parallel processes linear. For example, I want to add the numbers in column A, to the numbers in column B, and multiply them by x. We start with the first pair of numbers, add and multiply. Get the next pair, add and multiply.... Most of the stuff your computer does could be done in parallel, problem is its very hard to write general purpose software or create a general purpose CPU that can really take advantage of that parallelism.

      And whenever I see someone say something like, "Its good for high performance scientific computing", that means its good for general purpose computing too. Remember most scientific computing involves repeatedly doing simple operations on arrays of data (both large and small) -almost exactly like the type of stuff your computer does for every frame it displays on your monitor. In fact, some scientific apps have even been written to use the GPU on some video cards as a digital signal processor, because that's essentially what GPUs are/do.

    14. Re:It's bad news, actually... by PsychicX · · Score: 1

      On the other hand, I believe AMD's got some magic they haven't used yet up their sleeve. Though honestly I couldn't tell you what it is. I'm also surprised they're being so slow about ramping their clockspeeds, but this is probably just so their thermal profiles look superior to Intel's. I think they have fab problems, actually. They've hired a third party company and are now licensing technology as well in order to get to strained SOI. Even the dual Athlon64/Opteron architecture has existed since before AMD has had the ability to manufacture them (90nm was necessary). If they have something up their sleeve, we won't see it until that third party company (can't remember name) and Fab 36 go into high gear.

    15. Re:It's bad news, actually... by loose_cannon_gamer · · Score: 2, Insightful
      Not bummed out yet? Massive parallelism works well for people doing scientific computing, but for the average joe, it's useless. I don't care how fast a processor is--I usually have one task that will crush it--but rarely do I have two time-critical things to worry about at the same time. In the article referenced, they had to work hard to find things that would test the dual-core features.

      I had a small epiphany reading a review of one of these dual cored setups that released this week. One of the reviewers (I'd link, but I don't know where) mentioned that in a benchmarking discussion with a coworker, someone mentioned that the only reason we don't do benchmarks like 'run a mega intense first-person-shooter while doing some large encoding job in the background' is because until now we couldn't...

      The point was having dual cores will let you use the computer in ways you haven't been able to before. If you continue to use a dual core machine the same way you've always used a single core machine (and we've all been well trained on what processes you can run concurrently on a multitasking operating system and which ones you'd just better not), you're not going to see a benefit.

      The real benefits will come when people start to realize what they can do with a dual core system. The OS has been able to effectively multitask for a long time now, but most people don't -- they have one application going most of the time. But now we may be able to get out of the real parallelism bottleneck -- the user -- in accomplishing work.

      I think this opens up a serious new niche in most application markets -- processes you can start and let run in the background that actually do real work without much user intervention.

      And let's not forget, having dual cores is going to make everyone's SETI scores shoot through the roof. :)

      --
      In Soviet Russia, us are belong to all your base.
    16. Re:It's bad news, actually... by corngrower · · Score: 1
      AMD says it's doing good with it's 90nm yields, but I'm kind of doubting that myself. The pricing of these new dual cores seems to be high enough to want to restrict sales volumes, which indicates to me that they don't yet have huge volumes of these to sell.

      Of course I could be wrong. For the systems that these things will be used in, memory costs will probably dominate. The fact that the dual core opterons require no specialized support chips (unlike intel's solution) means that AMD can grab some of this money that would normally go to the makers of the support chip. Also, being a new chip, you want to maximize your profits, so you sell the first few to those that absolutely have to have them at high prices. Later you reduce prices (and profit margins ) a bit, sell some more, and so on, until it really hits the mass market.

      Still, these chips aren't clocked as fast as they might have been. This, says AMD, is due to thermal power considerations. It may be that they weren't able to clock the chips quite as fast as they really were expecting, meaning the manufacturing process isn't quite where it should be.

    17. Re:It's bad news, actually... by MikeBabcock · · Score: 1

      Actually, multiple processors affects the average Joe too.

      Most people will benefit, albeit only slightly, simply from having the OS get its own core (if it so decides to schedule itself). Your video game can now have a core all to its graphics pipeline, while the kernel has a different core for handling disk and network data, and your mail client checking for new messages won't slow down your video game almost at all.

      Average Joe plays video games with his PC. Average Joe will like multi-core CPUs because of much lower numbers in task-switches which speeds up each individual CPU.

      Intelligent OS schedulers will even start giving more cycles and switching tasks less often if possible to let a given core use its cache more effectively on an specific thread.

      --
      - Michael T. Babcock (Yes, I blog)
    18. Re:It's bad news, actually... by LionKimbro · · Score: 1

      Not bummed out yet? Massive parallelism works well for people doing scientific computing, but for the average joe, it's useless. ...rarely do I have two time-critical things to worry about at the same time.

      In the 2005 Intel keynote speech, distributed computation expert Justin Rattner noted that "without language support, this isn't going to work."

      Pretty much all apps can make use of parallel execution. If you have to interpret a big chunk of data, you can usually break it into segments, and process them in parallel, and then perform a fuse step at the end. Dividing into 2 isn't so exciting, dividing into 4 is pretty amazing.

      Right now, few programs make use of parallel execution. This is why you have to run a bunch of programs at once to see basically anything. Threading and multiple processes is a nice way to take advantage of wait time, but that's different than parsing a big XML file in parallel.

      Also, you're thinking about current use of computers. In the relatively near future, you're going to have CPU's cranking pretty much all the time. You're going to have face recognition running on your computers, you're cameras are going to be building 3D reconstructions of everything they see before them, yadda yadda yadda. In the nearer term, search and indexing loads are probably going to go up. There are going to be more things like the google desktop search bar that want to run and calculate all the time. We are rearchitecting our computers to support this sort of thing.

    19. Re:It's bad news, actually... by pipingguy · · Score: 1


      ...happens to be one with a lot of baggage

      I think I need some time away from the computer; my brain initially interpreted word that as "babbage".

    20. Re:It's bad news, actually... by UncleFluffy · · Score: 1

      That's only true if the complexity you add is serial.

      Sort of. Fanout comes into play very quickly when you start parallelising stuff. Try making a 2-level implementation of a 32 bit adder, for example.

      --

      What would Lemmy do?

  23. Built in coffee pot! by Senor_Programmer · · Score: 1

    If the article is correct, Intel could build in a coffee pot for those long nights of full load modeling.

    "the XE 840 almost hits the 300 watts level under one"

    Another plus! Demand for dual core laptops should give battery technology a new push.

    1. Re:Built in coffee pot! by zakezuke · · Score: 1

      If the article is correct, Intel could build in a coffee pot for those long nights of full load modeling.

      Ah, this would explain why Asus released a barebones solution called S-presso. Just add a couple of dual cores, water cooling, a fine italian pump, and poof the next generation in computers the e-s-presso. As the water travels over one each dual core chip it's super heated quickly and the italian pump takes over. Through the grounds to your demitasse mugs.

      Warning, not drinking enough espresso may result in premature burn out.

      --
      There is no sanctuary. There is no sanctuary. SHUT UP! There is no shut up. There is no shut up.
  24. Dual Core CPU's by antivoid · · Score: 2, Interesting

    I feel a good use for Dual-core systems is to put the OS on one core, including all explorer.exe instances and threads.

    The operating system shoul employ a smart system of monitoring CPU usage per thread and move the high- usage threads to the other core.

    I wonder though, on a slightly different topic - heat dispersion: nobody seems to talk about it - but two cores mean twice as much heat. How the hell do they do away with the heat? It dissapointing but they might be speedstepping/downclocking the cores dynamically at peak load.

    This certain warrants discussion.

    1. Re:Dual Core CPU's by techguy911 · · Score: 0

      What we see on the new heatsink is an enlargement of the copper bottom on the heatsink and more surface area thanks to the more numerous, slightly thinner fins. It is however, a bit shorter. The heatsink did get hot to the touch during our testing, registering temperatures of over 100 degrees Fahrenheit on our infrared thermometer.

      Happy?

    2. Re:Dual Core CPU's by ThaReetLad · · Score: 2, Informative

      Actually several of the articles do mention heat and the answer is this. Total thermal output from the top end dual core opteron is no more than for the top end single core processor (95 watts max). This has been achieved by using more energy efficient (and slightly lower perfomance) transistor designs in certain areas. AMD appears not to be doing any thermal throttling either.

      --
      You can't win Darth. If you mod me down, I shall become more powerful than you could possibly imagine
    3. Re:Dual Core CPU's by pohl · · Score: 1
      I feel a good use for Dual-core systems is to put the OS on one core, including all explorer.exe instances and threads.

      I've read this sentiment several times now since the dual-core craze came about. So long as your kernel and web/file browser tasks get scheduled, why should you care which core they're running on?

      --

      The "cue the foo posts in 3, 2, 1..." posts will commence with no subsequent foo posts in 3, 2, 1...

    4. Re:Dual Core CPU's by fitten · · Score: 1

      This is ancient scheduling rules. Way back (get in your Way-Back machines), multiprocessor machines weren't necessarily SMPs ("S" stands for Symmetric, meaning that any one is equivalent to any other one). Early on, Apple machines had accelerator boards that were similar (just to show some consumer relevance). In both of these cases, one CPU ran a scheduler and farmed out tasks to the other CPUs in the machine. The other CPUs were told when/if to switch tasks and everything. Typically, the "main" CPU also handled all I/O as well. This simplified the kernel somewhat (no need for elaborate locking schemes, for example) because essentially, the OS was single-threaded.

      Among the problems with this design is that there is still a point of contention in the kernel. What if 10 processes all wanted to do I/O at the same time? What if they were all I/O to different devices? Well... you had a bottleneck because only the one "special" CPU could handle any I/O requests, even if the I/O requests had nothing to do with each other (different devices, different data).

      Anyway, even in an SMP system, the scheduler should attempt to schedule threads on the last CPU they ran on (benefits of caching effects, etc.). However, in case of things such as I/O (which must be serviced *now*), it may be better to schedule the task on any available CPU, for example. So, modern SMP OSs tend to try to offer the best of both worlds (especially NUMA ones). Schedule a thread on the last CPU it ran on, but if it can't and the thread really, really needs to run, schedule it on any available CPU. Remember also that SMP is Symmetric. Any thread (even any OS thread) can run on any CPU. That means that the scheduler, interrupts, I/O, whatever can be handled by any CPU in the system, not just one special one which may be a bottleneck.

    5. Re:Dual Core CPU's by Anonymous Coward · · Score: 0

      IRIX can do this. You can lock one CPU to do the OS and give yourself guaranteed CPU time on other CPUs for quasi-real-time OS ability.

    6. Re:Dual Core CPU's by Anonymous Coward · · Score: 0

      It only matters if you want non-OS tasks to be guaranteed 100% CPU time.

  25. OP Misses the point by Craster · · Score: 2, Insightful
    It's worth noting that each scenario consists of only desktop applications, and it'd still be interesting to see some common server benchmarks, such as a database or web server.


    Except that this is a desktop processor, that won't be shipping in server systems. So in actual fact it's worth noting that the entire point is that each scenario consists of only desktop applications.
  26. Re:OP Misses the point - or you do by OneSmartFellow · · Score: 1
    Except that this is a desktop processor, that won't be shipping in server systems

    What, exactly, do you think the difference is between a 'desktop' and a 'server'.

  27. who cares? by lampajoo · · Score: 1

    really... modern software is bloated as all get out. who cares about new fancy shmancy processors when all it would take for our computers to run fast would be some good, old fashioned efficient coding? These new procs are just a way to get money for the chip manufacturers they do NOT translate to increased productivity.

    1. Re:who cares? by loose_cannon_gamer · · Score: 1

      Assuming you get back to 'good old fashioned efficient coding' (and don't get me wrong, I'm a performance oriented coder myself), useful computational work is limited by two things: processing time, and memory space.

      This new technology is a means of decreasing (or therefore increasing the available) processing time, which leads to the The Point:

      Better hardware lets you solve Bigger / Better problems. That's why people care.

      If you already have the hardware to solve every problem that was ever interesting to you, then by all means, ignore advances like these. But every improvement in computational power is an expansion in the all-important-realm of problems which can be solved in efficient, fiscally responsible manner.

      In layman's terms, more power means more interesting and useful potential applications (and employement for the software development community at large).

      --
      In Soviet Russia, us are belong to all your base.
  28. Re:OP Misses the point - or you do by Anonymous Coward · · Score: 0

    Huge differences. We're not talking about a dinky little MP3 server in your basement. No one will be producing motherboards with these 840EE's in them and no one will be shipping tower nor rack machines with one of these 840EE's in them.

  29. why all the speculation and hoopla? by cahiha · · Score: 1

    As far as I can tell, this is basically no different from a dual processor system, except that you are probably going to get a little less performance out of the dual core than out of two separate processors. In return, you are going to save a bit on hardware (sockets, etc.) compared to a true dual processor system.

    All these questions about whether Windows will usurp one of the cores or how to schedule the two cores seem positively bizarre, given that the answer is no different from dual processor machines.

    There could be multi-core designs that can get better performance than multi-processor designs (cell attempts to be one), but this doesn't seem to be one of those.

    1. Re:why all the speculation and hoopla? by Kz · · Score: 1
      As far as I can tell, this is basically no different from a dual processor system, except that you are probably going to get a little less performance out of the dual core than out of two separate processors.


      i think that's why the intel chip is a desktop one. to a single user, it's cheaper to get a multicore system than a real two chips machine; but for servers it's not the real thing.

      and because of the FSB arch of intel; it's not really a way to turn 2-way designs into a 4-way. the bus design makes it really hard (that's why 4-way Xeon MBs are so expensive!)

      OTOH, for switched archs like Opteron and POWER (and maybe PowerPC soon), it's really a cheap way to get twice as many cores. with minor redisigns, 2-,4- or even 8-way MBs could get 4-, 8- or 16-way cores! definitely much better economics for servers, and high-end workstations (there are rumors that MacOS X is ready for dual-core dual-chip machines)
      --
      -Kz-
    2. Re:why all the speculation and hoopla? by fitten · · Score: 1

      and because of the FSB arch of intel; it's not really a way to turn 2-way designs into a 4-way. the bus design makes it really hard (that's why 4-way Xeon MBs are so expensive!)

      Technically, Intel's multiprocessor bus design/architecture is pretty much the oldest and most simple out there. Quads are expensive because they simply can be. If someone feels they must have a quad processor machine (there are few folks who *really* need them and must have them) then vendors can charge a premium for those since they aren't produced in the scales that the other CPUs are produced and because people who feel they must have one will pay such a premium to get one.

  30. Re:Mod parent offtopic by tomstdenis · · Score: 1

    He said the amd64 is only for large databases... that's like wrong.

    Tom

    --
    Someday, I'll have a real sig.
  31. Re:OP Misses the point - or you do by OneSmartFellow · · Score: 1

    The machine I use as a workstation is a Dual Xeon box with SCSI drives. The only difference between that and a 'server' is what I do with it.

  32. Graphics by Gr8Apes · · Score: 1
    It will probably remain away from the common desktop for a long time to come. But it does have its uses.

    Graphics processing (large amounts of data in multi-dimensional arrays) are a perfect scenario for multi-threading. This is essentially lots of independent little problems.

    Another one would be writing real AI. AI running on its own threads could act much more like a real player in a game.

    So, there's 2 processes that could benefit from multi-processing. I bet you're saying but that's specialty graphics/games only. Well, what about an intelligent agent handling processing of your email/IM/IRC/web browing? (Not here yet, but everyone thinks eventually there will be some sort of agent technology to help us filter through the chaff)

    --
    The cesspool just got a check and balance.
    1. Re:Graphics by CastrTroy · · Score: 1

      The problem I found is that most parallel algorithms don't really get much speed up from using 2 processes, or even 4 processes. Most of them work by using f(n) processes for a data set of size n. Simply running 2 threads, or 4 threads to do the work doesn't really result in that much of a change. There's nothing stopping you from running n processes/threads on a fixed number of processors, but it kind of loses the effect and entire point of running all the processes.

      --

      Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
    2. Re:Graphics by Gr8Apes · · Score: 1

      I have no idea what you've been exposed to, but I can assuredly say that you're not looking at the right algorithms if you believe that you won't see a performance increase by dividing a parallizable process across multiple CPUs. I used to run large models on 16 CPU crays. The run time was about 1/14th to 1/15th of what it would have been running on a single CPU. (Real time, actual CPU time increased by about 10%)

      --
      The cesspool just got a check and balance.
    3. Re:Graphics by bluGill · · Score: 1

      Not all data set operations can be split up. If your operation can be split, then great. Clearly when you see operations running at 1/14th the time on a 16 CPU, your data set needs operations that scale well. At somepoint you will hit a wall though. Eventially adding more processors will make no difference at all (Unless you also add data).

      Brute force encryption cracking scales to 2^key size processors (perhaps a little less cause you have to get code and keys to the processors), but after that adding more processors will make not difference because every operation now depends on the previous one.

      Traversing a linked list does not scale because every operation depends on the previous. (though one would question why a linked list was used in a multiple CPU environment)

    4. Re:Graphics by Gr8Apes · · Score: 1

      Your points are all valid, and also indicate that the algorithm must be parallel capable. Traversing a link list obviously is not. Linear algebraic equations represent by multi-dimensional arrays can be, depending on the algorithms used to solve them.

      Some smart people spent a lot of time on coming up with interesting algorithms to solve such sets of equations, and how to compartmentalize the logic so that there are discrete subsets of work that can be done asynchronously, the key to being able to parallelize a task.

      --
      The cesspool just got a check and balance.
    5. Re:Graphics by CastrTroy · · Score: 1

      But This requires 16 processor machines. You can sort data in constant time if you have n^2 processors. These things don't really exist on the desktop, and it will be a long time before they do. Going from 1 to 2 doens't really add that much. Going from 1 to 16 will add a lot. I don't think you'll see a lot of mass marketed software for multi processing until you have desktops with 16 processors.

      --

      Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
    6. Re:Graphics by Gr8Apes · · Score: 1

      I'd disagree. Going from 1 to 2 gives you significant advantages in that you can now run two concurrent threads.

      If the file system as a DB view ever gets going, you'll see immediate benefits with 2 CPUs over 1, even if the 1 CPU is a little faster.

      --
      The cesspool just got a check and balance.
  33. Re:Uber amounts of RAM in 64 bit by Lonewolf666 · · Score: 1

    Actually, on the "normal" versions of Windows 2000 and Windows XP applications will only have 2 GByte of adress space available. So if you want to run One Big App that consumes all the memory, limitations on 32 bit Windows will start at 2 GByte.
    About Linux, I'm not sure...but someone else might help out with that question.

    In practice, 512 MByte are comfortable for typical desktop use. My private PC has 1GByte and I still count that as a luxury.

    --
    C - the footgun of programming languages
  34. Re:OP Misses the point - or you do by Anonymous Coward · · Score: 0

    If you think the only hardware that matters with a server is the CPU and hard drives, I can tell you've never had to actually manage a real server. Does your workstation offer redundent power-supplies? Hot-swapable components? 64bit PCI slots? Local console ports? Can it be remote booted? What chipset and motherboard does it use; server oriented chipsets are generally tuned to server tasks (Which are generally I/O and memory bound, rather than CPU bound) Whats the cooling like? What monitoring and reporting functions does it provide? Is it rack mountable? What memory does it have; is it checked and certified? How much memory can it take? Servers generally take GBs of memory; most desktops either don't have the DIMM slots or don't have the BIOS support for a fully populated 4GB, let alone enough slots to make use PAE on 32bit CPU's. What NICs does it have? Gigabit? Dual Gigibit? Fibre?

    No?

  35. Multimedia programs could use dual-core CPU. by MtViewGuy · · Score: 1

    I think the biggest thing that is driving the need for dual-core CPU's is the fact that multimedia-editing programs are very CPU-intensive tasks and could benefit from the use of a dual-core CPU.

    The Adobe Photoshop CS example you cited is a good one; imagine being able to use both CPU cores to dramatically reduce rendering times for processing high-resolution images in Photoshop CS. Also, video-editing programs such as Adobe Premiere and its competitors could also benefit from a dual-core CPU, given how much CPU time you need to edit videos downloaded from your MiniDV/MicroDV camcorders nowadays.

  36. Absolutely Perfect for multi-user systems by digitalride · · Score: 1

    My company wanted to try dual processors in our multi-user systems, but the motherboards and processors are just too expensive.

    These dual-cores are just another reason we feel that multi-user computing is going to be the efficient choice of the future, instead of everyone having their own high-powered machine. Single user machines will probably not take full advantage of the dual-cores for some time

    These are great for our local multi-user approach, but will also help networked thin client approaches like LTSP. Having one processor or core per user isn't even necessary, just having the second core will do a lot to lower latency when one cpu is busy with a thread.

    We can't wait to start putting them in our http://groovix.com/ systems in June...

    --
    Open Source is Common Sense: http://groovix.com/
  37. re: Dual will cores revolutionize gaming by Stunning+Tard · · Score: 2, Interesting

    There's this interview with Tim Sweeney, the leading developer behind the Unreal 3 engine.
    They're working on a multithreaded engine for unreal 3, exciting stuff.
    Like you said, AI is a logical chunk of processing that should be on a separate thread. Other logical chunks he mentions are physics, animation updates, the renderer's scene traversal loop, sound updates, and content streaming.

    So at least one multi-threaded game engine is in the pipe. This is good because we don't really have a chicken and egg problem.

    So I agree games will improve a lot with multi-core.

    But for other apps I'm not as excited. I don't know what other apps I use regularly could be sped up with a multi-threaded rewrite. Virus scanning? Searching? Media playback? eMule? SETI? Maybe lots of apps can be sped up, but will any of them do it? In the interview I linked Tim says a multi-threaded system takes 2-3 times longer to write and test.
    I don't use most of the apps that have a lot to gain from multi-cores (Media creation apps, server apps). Maybe I'll start doing more things at once. Or maybe run a dual-head system. Maybe.
    It seems games are the only resource pig apps I've ever really run, so they're the only apps that will prompt me to upgrade to multi-core. Maybe once dual cores are common non-game developers will start to exploit them. And maybe some app will suprise me but I'm not holding my breath.

    Until then a single CPU serves my needs fine. Sometimes I come across situations where I close one app to give another a boost. Such as shutting down apps to make the game faster. With a dual core I could probably run everything, but for now I'll settle for shutting extra stuff down.
    In the future when playing a multi-threaded game on a multi-core PC I'll probably still shut down extra apps just to squeeze out the extra fps.

  38. I don't understand what you're saying.. by killa62 · · Score: 1

    Are you Canadian?

  39. Re:OP Misses the point - or you do by OneSmartFellow · · Score: 1
    I can tell you've never had to actually manage a real server

    Actually my main function is monitoring Server Hardware, but that is hadly significant.
    OK, point by point, since you are so curious:

    Does your workstation offer redundent power-supplies? Redundant power is supplied to my power outlet, the PC doesn't need any additional hardware for this.

    Hot-swapable components? Yes, actually, ALL USB Based devices are hot-swappable.

    64bit PCI slots? no, but then I don't see too many 64 bit components yet either.

    Can it be remote booted? Yep, sure can.

    What chipset and motherboard does it use Irrelevant

    ; server oriented chipsets are generally tuned to server tasks (Which are generally I/O and memory bound, rather than CPU bound) Dude, my CPU is idle 99% of the time, it almost always is waiting for either the disk or the network or the keyboard, just like every computer out there.

    Whats the cooling like?fine, thanks for asking

    What monitoring and reporting functions does it provide? Like most modern hardware it has a health chip and has been configured to use SNMP, after all it is on a corporate lan, and that's just plain sensible

    Is it rack mountable?No, but that's just the box, not the computer inside.

    What memory does it have; is it checked and certified? How much memory can it take? Servers generally take GBs of memory; most desktops either don't have the DIMM slots or don't have the BIOS support for a fully populated 4GB, let alone enough slots to make use PAE on 32bit CPU's. Actually it runs ECC RAM, and is currently configured with 2 Gig. That's quite a lot for a workstation, but yes, it can hold 4 Gig.

    What NICs does it have? Gigabit? Dual Gigibit? Fibre? Well, no, it doesn't handle fibre, but that is immaterial, it's just a different kind of card after all. It is on a Gigabit LAN though if that helps.

    So you see, in almost every respect, my 'workstation' is a 'server' except that I don't use it as one.

    P.S. have a look at the specs for the Dell Precision 6700 'workstation' - not the machine I have, but it sure looks like a 'server' according to you.

  40. I also like ot think of this Intel offering as... by Senor_Programmer · · Score: 1

    The SUV of CPU's!

  41. Simple by The-Perl-CD-Bookshel · · Score: 1

    In Windows, from the task manager "processes" tab just right click on the process and select "Set Affinity..." I'm running a dual processor right now and I can force any process to run on either CPU. This is very useful for multi-tasking, not just SMP programs.

    --
    I don't keep a lid on my coffee so when I walk around I look busy -me
  42. Whats better - dual core or multiple CPU by Anonymous Coward · · Score: 0

    How does the performance compare with a dual XEON server vs a dual core (or the AMD equiv) ?

    I do a lot of video encoding and visual effects work and am about to build a dual 3.6GHz XEON server. I wonder if I would be better off/cheaper with a dual core instead?

  43. Memory bandwidth by ntropic · · Score: 1

    I haven't been following this too closely, but do they have one integrated memory controller or two for the dual core devices ? It would seem to be a fairly obvious thing to do to maintain the memory bandwidth per core. And yes, that would probably need a higher pin count package, but the overall system cost would be lower.

    1. Re:Memory bandwidth by Anonymous Coward · · Score: 0

      1 memory controller for both cores on AMD.
      But don't worry. It actually doesn't matter most of the time because memory bandwidth is overrated.

      K8 is almost never ever bandwidth-starved. anandtech did one benchmark awhile back where they used 333 MHz RAM and 400 MHz RAM and there was no difference at all for one test and little difference on another test.

      Or, if you want to conduct a similar experiment, try comparing performance with one channel and 2 channels. The performance is never double--it's less than 5-10%, which means you already have almost double the bandwidth you need with a dual-channel memory controller---perfect for a second core!

      (It also means the MHz myth is no myth at all anymore. Clock speed and/or dual-core is the real secret to higher performance again, much to the chagrin of most of the /. crowd.)

    2. Re:Memory bandwidth by SpinJaunt · · Score: 1

      All AMD K8's (Opterons, Athlons, Semprons [socket 754]) have One memory controller, including dual cores as there is more then enough aggregate bandwidth.

      --
      /. is good for you.
  44. ..infact you misunderstood me by FidelCatsro · · Score: 1

    Actualy , i implied that the AMD64 bit extensions were currently only showing noticble advantages in such operations as things are still not fully optimised for AMD64 in the normal users sphere(faster for some things but not by enough to be needed) , but i did go on to note that AMD64 chips were excelent for other reasons such as the architecture and the onboard memory controler which gives them an edge (compelation times are solid and i wouldnt do without them in x86) ;).
    Not offtopic no as grandparent said though , just a simple missunderstanding

    --
    The only things certain in war are Propaganda and Death. You can never be sure which is which though
  45. Re:OP Misses the point - or you do by Anonymous Coward · · Score: 0

    Why are you answering with totally irelivent information? Your original argument was that these 840EE CPU's could be used in servers because you consider your Dual Xeon workstation server-class hardware. My point is that it is not; server-class hardware is far, far more than than a fast CPU and SCSI disks, as I attempted to point out to you.

    The very fact that you seem to seriously believe that you get "redundent power" through your power cable and that USB is "hot swappable" just highlights this nicely. You prove further total ignorance of server hardware with your comment that you "don't see too many 64 bit components yet either"; maybe not on your home PC or workstation, but you sure as hell see 64bit PCI cards in server class machines.

    I'd consider changing your UID; perhaps "OneIgnorantKid" might suit you better?

  46. Causality? by Just+Some+Guy · · Score: 1
    I don't care how fast a processor is--I usually have one task that will crush it--but rarely do I have two time-critical things to worry about at the same time.

    Maybe that's because you've been trained not to run two intensive applications at the same time. If your subconscious is telling you not to launch a compile job until your DVD transcode is done, then you probably wouldn't get a lot out of an SMP system. Break that habit, though, and you might like what you find.

    --
    Dewey, what part of this looks like authorities should be involved?
  47. some crazy thoughts by Jaspers · · Score: 1

    thinking about it, why would Joe User want to upgrade his PC, if there is not any advantages to him. Think about it, someone is buying today a computer for about 600-700 dollars. the computer happily works for the years to come. The processors have become fast enough to work for everything that a user wants. Joe happily uses his PC to browse the web for the next 10 years. He is using email, internet and maybe excel / word / open office (take your pick). is there anything, some kind of technology that will actually appear and make people want / need to upgrade? I'm not talking about the power users nor the gamers. How is the industry going to make people upgrade their PCs if their PC is working just fine? Dual core means nothing to normal users.

    1. Re:some crazy thoughts by loose_cannon_gamer · · Score: 1

      All these 'who cares' posts are killing me -- does nobody have any imagination?

      New technology is good because it makes possible new applications. There's plenty of good money out there for people who are creative and visionary enough to realize that a whole new class of problems and applications just became reasonable with this hardware improvement...

      I concur that something like web browsing is pretty much bound in performance to internet connection, not processor. All that means is that applications like web browsing are ripe to be replaced with something more interactive, more meaningful, things that you couldn't do when the web was invented because the hardware hadn't been invented yet that you need.

      --
      In Soviet Russia, us are belong to all your base.
    2. Re:some crazy thoughts by What+me+a+Coward · · Score: 1

      Who was it that said we would never need a pc faster than a 286?

      I can't recall for sure (Darn memory hole in my head) but i think it was Bill Gates.

      In any event whoever said it shot himself in the foot on that one. Their are others who have fallen into the same trap making PC predictions like that. You just never know whats comming down the road that will change things in ways you can't even think of right now.

      Making such predictions in an offhanded way like that is a good way to end up eating your hat or pc chip as the case maybe.

      Never say never unless the world is ending in .001 seconds and even then something could still happen to prove you wrong. You may not be around anymore to know it but it could still happen;)

      --
      Coward? Coward! Thems fighten words!!
  48. Can each core throttle speed independently? by Anonymous Coward · · Score: 0

    If a dual-core AMD is fully loaded on one core and half loaded on another core, will it be able to throttle the speed of the second core and keep the other core running full-speed?

  49. 100MHZ will blow up. by Anonymous Coward · · Score: 0

    It's all been said before. 100MHZ, 1GHZ, 2GHZ. Just check google groups.

    http://groups-beta.google.com/group/comp.sys.pow er pc/browse_thread/thread/feaaa145095f4f1b/dfda61bc4 c0674b9?q=cpu+mhz+limit&rnum=1&hl=en#dfda61bc4c067 4b9

    The limit is qualified with "todays materials limits." New materials will change these limits.

  50. Am I the only one that finds this funny? by hedora · · Score: 1
    From the article:
    Windows XP Professional and most variants of Linux support multiple processors and thus multiple threads of execution. In fact, Windows XP Pro is capable of working with a total of four execution threads

    Is XP Pro really limited to 4 cpus? I wish my copy of Linux 2.6.11 Home Edition supported SMP... If only I'd forked over another $75 to Linus! ;)

    Anyway, in response to the parent, I'm pretty sure the reason that most people see an improvement with WinXP and hyperthreading is that hyperthreading subverts Windows's braindead scheduling policy.

    (Unless they've changed this recently) By default, Windows puts CPU hungry jobs in the foreground, so if you're running multiple apps, the only one that responds well is the one updating its progress bar.

    With multiple cores, you need to have at least one cpu intensive job per core before the scheduler can get in the way of interactive tasks.

    I've always wondered what the "Schedule the CPU for my applications" vs. "Schedule the CPU for system services" toggle actually did. The services choice might lower the priority of processes that are CPU intensive. (If that is all that it does, it should increase responsiveness when you multitask, and I doubt it would have much effect on games or multimedia, but I could be wrong.)
  51. The real reason for Dual Core by a4w5vffg · · Score: 1

    A secure DRM solution needs a secure kernel. What could be better than for Windows Secure Kernel to run on one CPU, and non-secure Windows and apps to run on the other.

    Microsoft needs dual-core, not multi-threading to make this happen in a very secure manner.

  52. DMA ;) by KZigurs · · Score: 1

    Enable DMA on your CDROM.

  53. Re:OP Misses the point - or you do by OneSmartFellow · · Score: 1
    I'll have to go demand my money back from the power company, I am sure they told me that the redundant power supply system that was installed at great expense within my building was as close to failure proof as possible. But hey, maybe they lied to me, do you have a better solution than that ?.

    Are you trying to say that USB devices are not hot-swappable, strange, I can plug and un-plug my USB drive and it always seems to work fine