[citation][nom]aidynphoenix[/nom]I have to make a few points clear here.. depending on the architecture of the processors being compared, one processor might perform better than another depending on what is being run.. so in "Some" situations AMD was the better buy. however the "Majority" of situations the Intel chips would outperform the AMD chips which justified the differences in the price. the prices of the processors have been perfectly balanced, and priced exactly where they should be. AMD was actually the one trying to Stretch the Truth to make themselves appear faster. for example their naming scheme with their processors. it seems shady and misleading to put "AMD 64 5000+" on a processor clocked at 2.6ghz. or 6000+ on a 3ghz.AMD just released "bulldozer" a tweaked 4 core processor and is calling it a 8 core because they added hyper threading.but yes i agree, in "some" situations, yea bulldozer might out perform some of the higher priced Intel processors in Winrar. but that's not what the "Majority" of users will be doing with their processor. so i wouldn't call it a better product.I don't want to come off as a Intel fan, because i'm not.. every build up to my last has been a AMD based system. but i'm not a fan boy either way. i buy what is best for my needs.. i wanted a computer i could use for the next 4 years, so i went Intel with this last build.. my first ever Intel build. i'm not excited or disappointed, i got what i paid for.[/citation]
Let's make something even more clear: You don't know a god damn thing about the architecture.
It's an 8 core and unless you want to change the DEFINITION of a core, there's nothing to argue.
I guess I have to simplify everything to preschool level for you children:
Ferrite rings, vibration....Eight of them. CMT vs. SMT. Let me explain a little more, or maybe you've already lost the ability to concentrate your little amphetamine-damaged mind:
Continuous Multi-Threading (CMT). Intel's "solution". (2600k example) 4 ferrite rings executing data, some features of the pipeline (like instructions) are doubled in the silicon so two threads can be 'placed' for execution. The additional cores are virtual.
The problems: The time it takes to switch between the two threads is highly significant and the switching process often leads to cache-thrashing. Hyper-threading was introduced way back when, before software engineers could really get a grasp of multi-threading. That's also about when HT should have been killed off.
Some software is still designed/optimized to take advantage of HT, ALWAYS at the expense of physical core execution (C4D/ (cinebench is always outdated and completely irrelevent)). More advanced software companies understand the logical fallacy of HT (autodesk).
Since intel is now placing more emphasis on physical cores and has sort of stopped deluding children with HT, things are changing. Even intel has admitted that optimizing specifically for HT (things like changing a section from one instruction to another so you can 'place' the thread for the virtual core) leads to performance degradation, even on their own physical cores.
Continuous Multi-Threading:
8 ferrite rings for executing threads, packaged as 4 'modules'. Why? Certain aspects of the pipeline are virtualized so you can 'place' an additional thread, yet it has two PHYSICAL cores for executing the threads.
Simple logic dictates that the CMT design is superior, mainly because it directly addresses the various shortcomings of SMT.
Just the name explains it all : "Simultanious" vs. "Continuous".
It took a good deal of thought and engineering to come up with such a future-based design. It also takes a bit of extra power to run 8 physical cores. Intel is quietly concerned, you little punks just don't know what you've bought or what's being sold.
AMD was very clear from the beginning: Essentially the same IPC or slightly higher thanks to higher clockspeeds (one of the least important aspects...) with more physical cores.
Only deluded intel children or kids that shouldn't be running AMD in the first place expected anything different.
A design for servers and workstations. This "enthusiast" intel-based (highly ignorant) crowd are peanuts compared to the server/workstation market.
Why the F are you kids looking at an 8 core workstation-optimized system anyways since all you do is play games and post on shitty review sites, trying to explain why you bought the same fucking processor that intel has been making for over four years?
So, kids, before you start posting, make sure you understand the competitions architectures, let alone your OWN.
What a bunch of idiots, though no higher density than anadtech.
Let's make something even more clear: You don't know a god damn thing about the architecture.
It's an 8 core and unless you want to change the DEFINITION of a core, there's nothing to argue.
I guess I have to simplify everything to preschool level for you children:
Ferrite rings, vibration....Eight of them. CMT vs. SMT. Let me explain a little more, or maybe you've already lost the ability to concentrate your little amphetamine-damaged mind:
Continuous Multi-Threading (CMT). Intel's "solution". (2600k example) 4 ferrite rings executing data, some features of the pipeline (like instructions) are doubled in the silicon so two threads can be 'placed' for execution. The additional cores are virtual.
The problems: The time it takes to switch between the two threads is highly significant and the switching process often leads to cache-thrashing. Hyper-threading was introduced way back when, before software engineers could really get a grasp of multi-threading. That's also about when HT should have been killed off.
Some software is still designed/optimized to take advantage of HT, ALWAYS at the expense of physical core execution (C4D/ (cinebench is always outdated and completely irrelevent)). More advanced software companies understand the logical fallacy of HT (autodesk).
Since intel is now placing more emphasis on physical cores and has sort of stopped deluding children with HT, things are changing. Even intel has admitted that optimizing specifically for HT (things like changing a section from one instruction to another so you can 'place' the thread for the virtual core) leads to performance degradation, even on their own physical cores.
Continuous Multi-Threading:
8 ferrite rings for executing threads, packaged as 4 'modules'. Why? Certain aspects of the pipeline are virtualized so you can 'place' an additional thread, yet it has two PHYSICAL cores for executing the threads.
Simple logic dictates that the CMT design is superior, mainly because it directly addresses the various shortcomings of SMT.
Just the name explains it all : "Simultanious" vs. "Continuous".
It took a good deal of thought and engineering to come up with such a future-based design. It also takes a bit of extra power to run 8 physical cores. Intel is quietly concerned, you little punks just don't know what you've bought or what's being sold.
AMD was very clear from the beginning: Essentially the same IPC or slightly higher thanks to higher clockspeeds (one of the least important aspects...) with more physical cores.
Only deluded intel children or kids that shouldn't be running AMD in the first place expected anything different.
A design for servers and workstations. This "enthusiast" intel-based (highly ignorant) crowd are peanuts compared to the server/workstation market.
Why the F are you kids looking at an 8 core workstation-optimized system anyways since all you do is play games and post on shitty review sites, trying to explain why you bought the same fucking processor that intel has been making for over four years?
So, kids, before you start posting, make sure you understand the competitions architectures, let alone your OWN.
What a bunch of idiots, though no higher density than anadtech.
Facebook
Twitter
Instagram