All *nix systems that do a significant amount of graphics will instantly benefit from dual core eg Linux desktops. A cursory glance at top will show the X11 server consumes a significant amount of CPU time. The X protocol is asynchronous. Hence any graphics application can go about it's thing while the X server is doing it's own thing rendering graphics for the application. Anybody running TWS (on WIndows, Linux Mac) with a lot of high volume symbols on a page will definately benefit if they are running a separate charting application regardless of the 'threadedness' of the charting app. There are many examples.
there are also situations that hyperthreading will considerably speed up things. This is what I referred to earlier: Without testing you do not know what will make things better. But the public always gets what it wants: Something very generic which is produced at the cheapest possible price. And somthing that because of the cost cutting does not offer a lot of choice in tailoring it to specific needs. Holmes
LMAO nononsense. Amusing to choose that name and then spend thousands of posts spouting nonsense. You're either completely thick or it just makes you feel good.
Most of today CPUs are faster enough for general trading applications. The fact of the real matter is lay under the speed of memory, which is very slow, and keep most of today's cpus awaiting about roughly 1/3 of time. You want real performance boost from a new system, just look for the new memory system. nothing else.
Notice that the new core duos have a faster clock speed so they should be able to move data in and out of (fast) memory faster. Damn, they're so cool it might be time to look at a new system.
sure, facts: Simultaneous Multithreading: Maximizing On-Chip Parallelism (1995) (Make Corrections) (247 citations) Dean Tullsen, Susan J. Eggers, Henry M. Levy http://citeseer.ist.psu.edu/cache/p...ilpzSztullsen95.pdf/tullsen95simultaneous.pdf this is your pentium 4 hyperthreading.. hyperthreading is a marketing term... but the basics are laid out here. in 1995 The Case for a Single-Chip Multiprocessor http://ogun.stanford.edu/~kunle/publications/hydra_ASPLOS_VII.pdf this is the case for duo core, quad core, in 1996 Piranha: A Scalable Architecture Based on Single-Chip Multiprocessing (2000) http://citeseer.ist.psu.edu/cache/p...zSzDatabasezSzisca00.pdf/barroso00piranha.pdf this is your 4x4, or Core 2 Duo... of course the commerical versions are more advanced... and piranha got cancelled i believe (not sure). this paper talks about what they attempted to do, in 2000 and this is the very promising "future" of CPU design Transactional Memory Coherence and Consistency http://tcc.stanford.edu/publications/tcc_isca2004.pdf database transaction style memory accesses
You may speak for yourself but you are most definitely not speaking for me. I have a 20 year + IT career behind me and I do know how to tune a PC, thank you. (Once I finish tuning the setup is at least twice, if not three times as fast as the standard XP installation). Hope you do not make the same generalisations when you try to trade: the devil is in the details. Holmes