• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

besides benchmarking is there a program that can utalize all of a core I7

alanwest09872

Golden Member
Just as the title states. are there any softwares that can take full advantage of a core I7 940 or 960. Benchmarking is not included in this since benchmarking is designed to use all the cores
 
Certain flavors of object modeling and video transcoding software --- from high-dollar Premiere, Vegas, et.al., to freebies such as Handbrake.




--
 
Just as the title states. are there any softwares that can take full advantage of a core I7 940 or 960. Benchmarking is not included in this since benchmarking is designed to use all the cores

first off you know half your cores are virtual right?

and not all physical?
 
1. all video encoding software.
2. all audio encoding software.
3. all distributing computing projects
4. a bunch of compilers
5. google chrome

Also... as aigomorla said, you only have 4 physical cores... so, a whole bunch of games.
 
first off you know half your cores are virtual right?

and not all physical?

In reality, 4 cores are real and 4 are virtual. Nero Encoding software loves multi cores, WinRaR too. Cyberlink MediaShow Expresso is the only software besides Nero Vision that can almost max all my cores.
 
3ds max is probably the only thing I use that would load 4+ cores to 100% usage. Folding@Home will do it to, so might as well mention that.
 
I have a quick question. How does Windows put the real/virtual cores in order? Say, if you open Task Manager, does it go like this?

Task Manager -- Physical CPU core
CPU 0 & CPU 1 -> Core #1
CPU 2 & CPU 3 -> Core #2
etc.?
 
3ds max is probably the only thing I use that would load 4+ cores to 100% usage. Folding@Home will do it to, so might as well mention that.

How do you get max to use 100% of all your cores? I had backburner and 3dsMax working on a 100m poly scene the other day and it wouldn't max out all the cores...

Do you have HT turned on?
 
How do you get max to use 100% of all your cores? I had backburner and 3dsMax working on a 100m poly scene the other day and it wouldn't max out all the cores...

Do you have HT turned on?


I'm assuming he means while rendering. I use C4D and it will use all cores/threads @ 100% while rendering.
 
I have a quick question. How does Windows put the real/virtual cores in order? Say, if you open Task Manager, does it go like this?

Task Manager -- Physical CPU core
CPU 0 & CPU 1 -> Core #1
CPU 2 & CPU 3 -> Core #2
etc.?

I believe it is mapped like that.
 
In reality, 4 cores are real and 4 are virtual.

No, technically all 8 cores are virtual. HT is 'symmetric' that way. One core is turned into a siamese twin, so to say, with just enough hardware doubled up to run two threads in parallel, but sharing a single execution backend.
So it's 1 physical core -> 2 virtual cores, rather than 1 physical core -> 1 physical and 1 virtual core.
 
As I understand it, the two threads are literally going down the pipeline in parallel, so they get to utilize more of the core's capacity...
 
As I understand it, the two threads are literally going down the pipeline in parallel, so they get to utilize more of the core's capacity...

Exactly.
The first step was to make a core superscalar (Pentium), so that multiple instructions of a thread could be executed in parallel.
The next step was to make a core execute out-of-order (Pentium Pro), basically buffering instructions and reordering them, so that more parallelism could be explored.
The third step is Synchronous Multi Threading (SMT), known as HyperThreading Technology in Intel's implementation (Pentium 4). Basically you feed two streams into the instruction buffer at the same time... Since instructions of different threads are independent of eachother by definition, there will be more parallelism than with a single thread (where you may have a lot of dependencies on instructions, and no matter how you reorder, the extracted parallelism will be limited). This exploits the massively superscalar backend to a much higher degree.
 
are there any softwares that can take full advantage of a core I7 940 or 960. Benchmarking is not included in this since benchmarking is designed to use all the cores

Interesting that you phrase it like that, because maxing out all the cores (the goal of certain synthetic tests) is not necessarily the fastest way to perform a given task (the goal of most real applications).
 
Exactly.
The first step was to make a core superscalar (Pentium), so that multiple instructions of a thread could be executed in parallel.
The next step was to make a core execute out-of-order (Pentium Pro), basically buffering instructions and reordering them, so that more parallelism could be explored.
The third step is Synchronous Multi Threading (SMT), known as HyperThreading Technology in Intel's implementation (Pentium 4). Basically you feed two streams into the instruction buffer at the same time... Since instructions of different threads are independent of eachother by definition, there will be more parallelism than with a single thread (where you may have a lot of dependencies on instructions, and no matter how you reorder, the extracted parallelism will be limited). This exploits the massively superscalar backend to a much higher degree.

Well then, what about HT on the atom? Isn't the atom an in-order processor?
 
How do you get max to use 100% of all your cores? I had backburner and 3dsMax working on a 100m poly scene the other day and it wouldn't max out all the cores...

Do you have HT turned on?

Don't have an I7, should have mentioned that in my post. I know max uses 100% of my Phenom when rendering, and I've heard it'll fully load more than that. How many cores it will actually use I'm not sure, guess it's why I said it uses 4+.
 
The two most intensive things I do on my CPU are distributed computing and RAR compression, with the occasional video edit as well. DC projects will use anything you give them, and can scale infinitely so long as the server can send you enough work units. For video editing it will depend on what you're rendering. Simply 100% loading your CPU won't always happen. For example my DC projects use a lot of floating point math, and some arethmatic as well, so when I'm running them they spike my CPU up and down from full to null while they swap data in and out to process, and as the instruction mix changes. If your software isn't coded to account for this, and find the CPU something to do while it's idling, then it will behave the same way. I will post a task manager screen later to show what I mean if anyone doesn't get he idea
 
foobar conversion
i time stretch some podcasts.. with 4 core it chews 4 files at a time. or flac to mp3 conversion.same deal
 
Well then, what about HT on the atom? Isn't the atom an in-order processor?

Yes, that works differently from the Pentium 4 and Core i7 that I was describing.
I believe it's a case of time-based multiplexing. On even cycles, it feeds instructions from thread 0, on odd cycles, it feeds instructions from thread 1, in a nutshell (Larrabee also does this, but it does it 4-way).
 
Yes, that works differently from the Pentium 4 and Core i7 that I was describing.
I believe it's a case of time-based multiplexing. On even cycles, it feeds instructions from thread 0, on odd cycles, it feeds instructions from thread 1, in a nutshell (Larrabee also does this, but it does it 4-way).

The Xeon 5500 series uses 2-way Hyper Threading AFAIK.
 
Well then, what about HT on the atom? Isn't the atom an in-order processor?

Yeah, and slow as hell.

[Offtopic] For some reason, I still laugh at this day the time that you posted " your printer has better graphic performance than anything GMA" , and that was two years ago loll
 
Back
Top