Why is FSB Slower than CPU internal speed?
2012-04-17 21:05
399 查看
see:http://www.tomshardware.com/forum/251796-29-slower-internal-speed
The FSB is only responsable for transporting data, therefore doesnt need to run nearly as fast as the CPU, which takes said data, processes it, then sends it back on the FSB.
the FSB does not need to work at the same speed as the cpu. It only transports data, it does not process it. If the FSB was the same speed as the cpu, the fsb would be waiting on the cpu most of the time.
Dear Hive Mind,
I think this belongs in here, because it's a pretty remedial question. If I had two processors with equivalent characteristics (say, an Intel E2220 and um, an E8200 that had been underclocked to 2.4Ghz, so two 2.4Ghz Core 2 Duo processors for use in a program that makes no use of cache), but one with a 800MT/s FSB and another with a 1333MT/s FSB, would there be a significant performance difference between them? This is something that I've never really understood, where in the CPU>FSB>RAM link is the bottleneck? Since cache means nothing in this situation, I'm tempted to get the E2220, but I'm just a little wary about the lower FSB.
Also, does the situation somehow change if I were to overclock both of these processors (again to the same clock frequency)? I assume not, since the FSB would rise proportionally with the overclock, but as I said, a little hazy on this area.
Answer:
The bottleneck in the ram-fsb-cpu bottleneck CAN be the fsb, if the multiplier is too high then the ram can't supply enough instructions to the cpu to use it at its full potential. Its also important to note that ram would not use the entire FSB. The timings actually refer to how long it takes to get data and use it - its not instantaneous. There is a balance that needs to happen between FSB and Latencies to get optimum performance
The FSB is only responsable for transporting data, therefore doesnt need to run nearly as fast as the CPU, which takes said data, processes it, then sends it back on the FSB.
the FSB does not need to work at the same speed as the cpu. It only transports data, it does not process it. If the FSB was the same speed as the cpu, the fsb would be waiting on the cpu most of the time.
Dear Hive Mind,
I think this belongs in here, because it's a pretty remedial question. If I had two processors with equivalent characteristics (say, an Intel E2220 and um, an E8200 that had been underclocked to 2.4Ghz, so two 2.4Ghz Core 2 Duo processors for use in a program that makes no use of cache), but one with a 800MT/s FSB and another with a 1333MT/s FSB, would there be a significant performance difference between them? This is something that I've never really understood, where in the CPU>FSB>RAM link is the bottleneck? Since cache means nothing in this situation, I'm tempted to get the E2220, but I'm just a little wary about the lower FSB.
Also, does the situation somehow change if I were to overclock both of these processors (again to the same clock frequency)? I assume not, since the FSB would rise proportionally with the overclock, but as I said, a little hazy on this area.
Answer:
The bottleneck in the ram-fsb-cpu bottleneck CAN be the fsb, if the multiplier is too high then the ram can't supply enough instructions to the cpu to use it at its full potential. Its also important to note that ram would not use the entire FSB. The timings actually refer to how long it takes to get data and use it - its not instantaneous. There is a balance that needs to happen between FSB and Latencies to get optimum performance
相关文章推荐
- Why malloc+memset is slower than calloc?
- Why is one loop so much slower than two loops?
- stackoverflow : Why C++ output is too much slower than C?
- Why OpenCV is slower than Matlab in matrix related operation,如:特征值分解,奇异值分解
- stackoverflow : Why C++ output is too much slower than C?
- stackoverflow : Why C++ output is too much slower than C?
- 7 Reasons Why BSD Is Better than Linux
- Why Git is Better than X
- Why is epoll faster than select?
- Why attitude is more important than IQ
- keras: Why is the training loss much higher than the testing loss?
- Why Git is better than SVN
- Why AlloyFinger is so much smaller than hammerjs?
- Why Notebook CPU stuck at Low Speed (为什么笔记本CPU降频工作)?
- Why is processing a sorted array faster than an unsorted array(Stackoverflow)
- Why is truth() faster than bool()? Part I
- JAVA常见问题其一:Why is it faster to process a sorted array than an unsorted array?
- Why ConcurrentHashMap is better than Hashtable and just as good as a HashMap
- Why is IMAP better than POP?
- Why iperf dual traffic in 100Mbps envrionment is much less than 100M?