witnessed in the Winrar + Witcher tests
I still feel that this is purely L3 cache-size effects, which can be controlled by the dictionary size chosen for WinRAR.From their results i5s are adequate for Integer + Intger tasks but not for intensive Integer + FP.
The Celeron and i3 are to be discarded even in INT + INT as they lack the necessary grunt, this can be witnessed in the Winrar + Witcher tests where they lose throughput on much bigger %ages than all other CPUs, and they have less throughput to begin with.
Can I get some context on this transconductance? I've seen it a few times before but I don't get the joke
Where do they say that? Why would that even make sense?Their methodology is still adequate, actualy they must run winrar at 100% the time it take for CB to make the test, then they run CB several times consequently if necessary to check the compression time.
I can't imagine why anybody wouldn't appreciate that benchmark...
Basically, there is a certain forum member who constantly claims that the "transconductance" -- basically a way to express the performance of a bipolar transistor -- of transistors from one manufacturer is superior than those from another.
Yes, so fair to compare a CPU with just two floating point units to one with four.And did they compare the FX-4300 against the i5-4670K? Because that is what would be fair, 4T versus 4T CPU.
For one transconductance is a caracteristics that relate an output current to an input voltage and has nothing to do with the transistor being bipolar or being a fet.
Indeed fets are commonly caracterised by their transconductance while bipolar are more often caracterised by the output current/input current, that is, by their current gain and not by their transconductance wich is way higher than the fet s..
Anyway the more you talk about those issues the more you are showing your incompetence in that matter, so it s no surprise that you are trying to ridicule what you know being out of your reach, the same as resorting to insults when there s no more valuable arguments to oppose, but i guess to each his own with his own capabilities and skills..
What are you talking about? I thought that was a very common workload? I always fire up Winrar when I game.
Where do they say that? Why would that even make sense?
They only say that they run CB for single and multi,for winrar they explicitly state that they compress the folder several times.
Of course the less cores you have the more % of the CPU is being used by a task like winrar.
Yes, so fair to compare a CPU with just two floating point units to one with four.
If you're doing pure integer testing then I suppose.
It make sense to keep the CPU fully loaded with winrar to check the CB runtime, otherwise the test wouldnt make sense..
Likewise CB should be running all the time needed to compress the folder to see the influence on the compression time, otherwise it wouldnt make sense as well.
Or would your do otherwise..?..
Given the price difference and the Cinebench obsession it's a lot more fair.Yes, so fair to compare a CPU with just 4 integer threads to one with 8.
That's not an efficient scenario for the FX design. It was designed to have both integer cores in a module. How about comparing transistors per core?Then compare an FX-8xxx with an i5, while limiting the number of CPU logical cores on the FX to 4
Xvid isn't even close for overall usage. x264 is to go-to codec for nearly every cell phone and tablet besides being incredibly common on desktops.
From what I understand, Abwx implies that due to their OoOE design, cores end up executing instructions from a thread even when their time slice is done and the CPU is currently servicing another thread. I find that hard to believe, but I'm already near the limit of my CPU&OS understanding, in which a thread requires execution context in order to run.
I would really appreciate it if someone with better understanding of CPU internals could shed some light here.
Edit: Thinking about it more, I feel strongly that there is at least SOME serialization going on in the core, when switching thread contexts, because otherwise, a few micro-ops from one (prior) thread, would get access to the page(s) of memory of the (newly-scheduled) thread, and that could present a security issue from the standpoint of the processor architecture.
Likewise, there would be two sets of registers for the memory access descriptors in the processor, one bank for each hyperthread. (This is my speculation.) Unless the AGUs and load/store units only deal with physical addresses somehow.
You should only compare NATIVE cores, not the SMT or CMT cores / threads at all. CMT on 15h designs produces 80-93% yield of a native core while SMT peaks at < 27%. I.E. on AMD 15h parts the non-BSC of the module should be disabled through down coring and on Intel the Hyperthreading should be disabled.
~33% area wiseVishera: 1.2B
Haswell GT2 4C: 1.4B
Haswell likely has quite a few more transistors per core, although it would be nice to know how much is eaten by iGPU.
That is not fair comparison, BD single Core is a lot smaller than Intel Core. Comparing one Module vs one Intel Core + HT is the only technically sound action. At 32nm both AMD Module and Intel Core (+HT) almost has the same size.
That way both could use 100% of their resources and that will give you Throughput for each of them.
No. XviD is incredibly common still a decade later as an all rounder which x264 can't touch. If you are not on a stack of trackers don't comment.
Most of the pirated content available through torrents where I live is x264. xVid is present as the second choice, but mainly for low resolution / low bitrate content.Well, yeah -- it's fairly popular for pirates. People who actually pay for their content legally, not so much. x264 is the #1 most common codec used worldwide for HD content -- and it has been that way for nearly a decade. You see, facts are not based on your opinion. That's why they are facts.
Considering xVid can't even be used in many countries due to legal reasons / patent problems -- you really don't know what you're talking about. H.264 is way more common worldwide.