unless you have a truly awful connection, the biggest delays aren't on your side of the broker's server, they are on the broker's side of the server. and as a retail customer, there is sweet FA you can do about that. the only way to collect truly-timed data is to BE the exchange - like Oanda. they're going to make a mint selling that data...
http://www.theregister.co.uk/2004/08/16/amd_quad-core_opteron/ :eek: :eek: :eek: :eek: I wonder if they will be pin compatible with the old mobos? As AMD gets more and more chips on die, the more the advantage of the HyperTransport over Intel Xeons. nitro
You've got to be kidding me right? Your the epitomy of Elite "Nitro", how soon we forget http://www.elitetrader.com/vb/showthread.php?s=&postid=196744#post196744 You bought the "flop" chips and forgot everything I said last year. Second time this has happened.
You were right!!!!!!!!!!!!!!!!!!!! :eek: :eek: :eek: But I have to tell you that I am not unhappy with my Xeon machine, just that this Opteron machine is out of this world in 64-bit mode. FWIW, it is only way better than the Xeons in 64-bit mode, othewise in 32-bit mode it is just an expensive though still very fast machine, giving equivalent thoroughput to say a dual 3.6 Ghz Xeon machine. Sadly, at least in MSFT land on AMD, the 64-bit OS world is still young. BTW, I now have a use for wavelets and genetic algorithms, but it took me a long time to even come close to having a possible application for them. I have no idea if it will give me anything or not even now. Honestly, have _you_ made money trading _directly_ as the resut of the fact that you are using wavelet analysis, or even GAs? And I don't mean a given trade, I mean statistically significant gains? What I am looking to do with them though is not in the way most people use them in finace (I think) but I need them to do Time-Frequency analysis in realtime. It is interesting, but whenever I get too far afield from just plain old market style analysis, I have not made any money trading. That may be more a statement about me than the tools though... Have you won the Fields Medal yet? nitro
I'm going to keep this an Opteron story only, have a new found distrust for sharing information with people after one of my ideas/suggestions was marketed and sold by someone about 6 months ago. The history of the Opteron is the guys who made the Alpha chip left disenchanted by Intel when they bought the division and came to work for AMD. Intel really only wanted to bury the name and collect the customers but did use one peice of intellectual property called EPIC, which they modeled their Itanic (sic) proccessors after. They DISCARDED the rest thinking they were bigger than the market and would make the market swing their way ecomomically , WRONG. The market was over saturated and the MARKET dictated a cheaper but more powerful line of cpu's and real innovation. So the geniuses that made the Opteron brought with them what was on the drawing board for Alpha, this is from a Jan 1999 article. " Now that Intel and Samsung manufacture Alpha processors, the timeline for production of 21264 systems has accelerated. The 21264 will appear in three distinct technologies: 0.35, 0.25, and 0.18 microns. Performance increases and cost reductions will accompany each new implementation. Alpha's next generations, 21364 and 21464 chips, are already on Compaq's drawing boards. Plans beyond the 21464 are as open as the Alpha architecture. Future Alpha chips might include copper technologies, fully depleted Silicon On Insulator (SOI) multiple CPUs per chip, and a technology similar to Explicitly Parallel Instruction Computing (EPIC). Compaq and Samsung plan to keep Alpha machines at least twice as fast as and less expensive than their IA64 counterparts. (For a performance comparison of several generations of Alpha and Intel processors, see the sidebar "The Performance Curve.") Alpha processors will be viable into the next century. They will offer benefits for users who need top performance or want to migrate to a mature 64-bit platform today." You see the dual core design was part of the Opteron design from the start, and Hypertransport was the second technology after the core design to make this all work. You will only know how much better your machine is when you put a load on, I mean a realworld load, if you even can tax it that much for a long period of time. This is what the Opteron was designed for, HUGE loads with multicores. And to tell the future, well I'll just leave this screenshot. Don't ask b/c I won't tell, but will say you have no idea what is coming ..... and fast.
Yes, I remember the alpha chips well. So much promise, but it seemed to fizzle out if I remember correctly, after Compaq bought out Digital. I am in the process of researching porting all of my sofware to a platform independent GUI and network libraries that will run under several OS's. That will facilitate me going wherever the technological edge is without having to wait for MSFT to port to 64-bits. Not sure what your .gif referes to, but it looks like a three core opteron if that is even possible. nitro
Windows also provides a second clock, accurate to 1 to 10 microseconds via the QueryPerformanceFrequency and QueryPerformanceCounter functions: http://www.geisswerks.com/ryan/FAQS/timing.html The 10ms system time resolution (via GetLocalTime, etc) seems to only apply to certain versions of windows. I see 1ms resolution on WinXP.