GPU with general purpose operating system

[engr_joni_ee]

Hi,

I guess Intel core i5 and core i7 both have 4 processor cores on their chip. The only difference is that there is multi-threading option in core i7. Am I right here ? I am wondering to have multi-core processing system with general purpose operating system like with Nvidia GeForce GTX 1080 which it self cost around 500-600 USD. Is it possible to have such a system with multi-cores with general purpose operating system like Windows or Linux etc ?

GTX 1080 has 2560 CUDA cores, how these cores differs from the cores of intel i5 or i7 ?

 

[betwixt]

My i7 has 8 cores.

I don't think Nvidia give technical information about their hardware but I'm sure they don't have a general purpose instruction set like an i3/i5/i7. I think of Nvidia cores as being like bit-slice processors (showing my age!), each very fast but with limited capabilities. Even if they could run a standard OS, they would not have the hardware needed to interface to peripherals.

Brian.

 

[kripacharya]

....like bit-slice processors...

Wow. Haven't come across that term in... umm.. decades? :thumbsup:

 

[ThisIsNotSam]

Wow. Haven't come across that term in... umm.. decades? :thumbsup:

Sadly enough, I know exactly what the term is and remember studying it in computer architecture 101.

 

[kripacharya]

Sadly enough, I know exactly what the term is and remember studying it in computer architecture 101.

Well.... John... then you are younger than the both of us ;-)

Brian needs to update his pic. Uhh.. so do I.

 

[engr_joni_ee]

Hi, I guess some core i7 have 8 cpu cores while others have 4 cpu cores, please correct me here.

Back to the original question: If there is no such general purpose OS exist for GPU based system then what about connecting a GTX 1080 to PCI Express slot of motherboard having core i5 or i7 ? Would it increase the speed when it comes to perform heavy simulations ? Would it help to run software like Xilinx Vivado which takes almost 30+ GB space in hard drive and is considered to be a time consuming to synthesis and bit generation ?

 

[betwixt]

Would it increase the speed when it comes to perform heavy simulations ?

Only if the program specifically asked the GPU to perform certain tasks. The CPU and GPU are not parallel devices that share the load, the GPU is a peripheral device that can autonomously perform certain tasks but it has to be told what they are by the main CPU. The GPU is also optimized to 'dual-port' some or all of it's memory so it can be read to the screen at the same time as it is being updated by the GPU, this is of little benefit to the program you run other than getting the results to your eyes a little quicker.

As far as instructions go, the GPU has little concept of a conventional program and you can't program it at low level (unless you have insider knowledge) but it can for example very rapidly calculate flood fills when given the boundary points of the flood region.

You could think of it this way:
To draw a 100x100 pixel blue box on the screen, the CPU alone might have to start at the top left corner and turn the blue bits of the mapped memory location on and the red & green bits off, then move to the next location and repeat, each time checking it was still within the boundary and if necessary moving to the next line down the screen and repeating until it reached the bottom right corner. That means everything is repeated 100x100 = 10,000 times before the CPU could move on to do something else. It would only be able to write to the video memory during the sync period (off the screen edge) so for most of the time it would be held up because the displaying hardware would have control of the memory.

Using a GPU, the CPU would tell it the coordinates of the box and the color to fill it then let the GPU take over, just a few instructions then it could continue more useful tasks.

Brian.

 

[engr_joni_ee]

Hi, is there any software exist at commercial level which tells the GPU to share the load of CPU having Windows OS ?

 

[ads-ee]

Hi, is there any software exist at commercial level which tells the GPU to share the load of CPU having Windows OS ?

There might be some niche product out there that I haven't heard about, but I guarantee you won't find something that will speed up a Vivado synthesis.

Projects such as take advantage of multi-core CPU and GPU and require both as such project use matrix operations, which GPUs are specifically design to accelerate.