0
Want this question answered?
Be notified when an answer is posted
Add your answer:
Earn +20 pts
What has the author Steven J E Wilton written?
Steven J. E. Wilton has written: 'Block transfers in a shared memory multiprocessor' 'Architectures and algorithms for field-programmable gate arrays with embedded memory'
Difference between multicomputer and multiprocessor?
Multiprocessors have a single physical address space (memory) shared by all the CPUs whereas multicomputers have one physical address space per CPU Multiprocessors have a single physical address space (memory) shared by all the CPUs whereas multicomputers have one physical address space per CPU
Difference between distributed system and multiprocessor system?
The difference between distributed system and multiprocessor system is whether the processing units in the system share the main memory. If yes, then the system is multiprocessor system; otherwise, it's a distributed system.
Compare distributed shared memory and centralized shared memory?
yes
What does Asura software do?
Hi all, The ASURA software is a Distributed shared memory multiprocessor. It is a large scale, cluster-based, distributed, shared memory, multiprocesor being developed at Kyoto university and Kuboto corporation. Up to 128 clusters are interconnected to form an ASURA system of up to 1024 processors. The basic concept of the ASURA design is to take advantage of the hierarchical structure of the system. Implementing this concept, a large shared cache is placed between each cluster and the inter-cluster network. The shared cache and the shared memories distributed among the clusters form part of ASURA's hierarchical memory architecture, providing various unique features to ASURA
A bus based multiprocessor uses Snoopy cache to achieve coherent memory. will semaphore works on this process?
yes
What are the devices which contain multiprocessor and multicomputers?
multiprocessor and multicomputer
Which data type perimits sharing memory among different types of data?
Types cannot share memory; only instances of a type (objects or variables) can share memory. This is achieved through the use of shared resource handles or "smart pointers". We can also use C-style pointer variables to share memory, however they are problematic because as soon as the shared memory is released, all pointers to it become invalid but there's no way to tell if a pointer is valid or not. Even without using shared memory, there is no notion of "ownership" as far as pointers are concerned. In the absence of resource handles, the best strategy is to allocate memory in an outer scope and then create shared pointers within an inner scope. The shared pointers MUST NOT release the memory they refer to. When we return to the outer scope, the shared references will have fallen from scope, thus we can safely release the memory using the same pointer variable we used to allocate the memory. In this way we get some notion of ownership. With resource handles we don't have this problem because the shared memory cannot be released accidently until the very last reference falls from scope. The last reference need not be the one we originally used to allocate the memory. Sharing memory between threads is best kept to an absolute minimum. Although we can use locks to prevent data races, we achieve better performance with lock-free code and there's less risk of deadlock (where two threads are waiting on each other to release the memory they've locked). Shared memory is often unavoidable, but don't use it just because you can, only use it when it is appropriate to do so; there are always alternatives.
What is the difference between dedicated and shared graphics?
Some graphics cards have memory on board the card itself. Others rely on the memory chips on the system board for memory , thus sharing memory. Suppose a PC has a specification of 256MB of memory and 64MB shared graphics. The memory available for your operating system and applications will be 192MB. If a PC had the same specifications but the 64MB were dedicated, the 64MB required to run the card would use this memory, leaving the full 256MB for your operating system and applications. In short, shared = compromise, dedicated = better.
What is centralized shared memory architecture?
MIMD(Multiple Instruction streams, Multiple Data streams) multiprocessors fall into two classes Centralized Shared memory and distributed shared memory Centralized Shared Memory Architecture has at most a few dozen microprocessors chips, that shares a single centralized memory. They have large caches, single memory, with multiple banks. The single memory has a symmetric relation to all the processors and uniform access time from any processor.
Why use multiprocessor in supercomputers?
More processing means faster, better computing.
Is dedicated video memory still better than a significantly bigger shared video memory Like the Acer AS5530-5041 with 256 dedicated AND 640 shared versus the Acer AS7520-5181 with 1792 shared only?
Yes, dedicated memory is always better then shared (or integrated) edit answer after editing the question: but what is the shared memory option is much bigger than the dedicated option? (See edited question)
