Hello. I am using a GTX 760 2GB now, and I run some BOINC projects on my PC. Some of the projects can use CUDA. I am thinking of upgrading my graphic card now. On the futuremark page (https://www.futuremark.com/hardware/gpu), GTX 760 only gains 6420 points and GTX 1060 6GB has 12960 points. I don't understand how these marks were measured. Does it mean GTX 1060 will calculate the same BOINC task 2 times faster than GTX 760? But I checked the specifications and found GTX 760 has 1152 CUDA cores @980 MHz and 1060 has 1280 CUDA cores @1506 MHz, so the actual calculation capacity of GTX 1060 should be less than 2 times of GTX 760?
-
Multi-million dollar Cheyenne supercomputer auction ends with $480,085 bid — buyer walked away with 8,064 Intel Xeon Broadwell CPUs, 313TB DDR4-2400 ECC RAM, and some water leaks
Which factors determine the speed of CUDA calculation?
- Thread starter Jasis
- Start date
Solution
Martell1977
Splendid
- Oct 26, 2010
- 3,846
- 3
- 24,960
The difference is not in how many cores, but the efficiency of the architecture of those cores. Same as why an i7-950 has 4 cores/8 threads, but is slower than a i7 4790, which is the same cores/threads. They refined the architecture to do more with the same or less resources on chip, then increased the memory amount and speed.
TRENDING THREADS
-
-
News Firefox user loses 7,470 opened tabs saved over two years after they can’t restore browsing session- Started by Admin
- Replies: 58
-
News Apple debuts M4 processor in new iPad Pros with 38 TOPS on neural engine
- Started by Admin
- Replies: 76
Latest posts
-
-
-
-
Question Getting alot of kmode exception errors after updating bios
- Latest: CodeConnoisseur
Facebook
Twitter
Instagram