To: andreas_wonisch who wrote (30112 ) 3/2/2001 11:26:05 PM From: jcholewa Read Replies (1) | Respond to > I'm not sure if I understand you correctly. Do you mean that PC1600 CL2 (not CL2.5) latency is not the same as PC150 2 cycles, then its latency in nanoseconds (you know, actual time units) depends on the core frequency.PC1600 memory runs on a 100MHz clock. This mean 100,000,000 cycles per second. This means ten billionths of a second per cycle, or ten nanoseconds per cycle . So 2 cycles (the best case for PC1600) at this speed would be twenty nanoseconds . 2.5 cycles (the worst case for PC1600 CAS Latency) would be twenty-five nanoseconds .PC133 memory runs on a 133MHz clock. 133,333,333.33 cycles per second, or 7.5 nanoseconds per cycle . The best case CL, 2 cycles, equates to fifteen nanoseconds , which is faster than the 20ns of the CL2 PC1600. The worst case for PC133's CL, 3 cycles, equates to 22.5 nanoseconds , still better than CL2.5 PC1600 latency of 25ns.time because it is operating on a higher clock than PC1600!!! This is really, really important. PC133 CL3 has lower latency than PC1600 CL2.5. PC133 CL2 has lower latency by an even greater margin than PC1600 CL2. The only way PC1600 can win here is if you get it in CL2 but you get the PC133 in CL3. But that's an unfair fight.CAS, CAS2RAS and RAS ratings (like the chipset and soforth, and maybe other stuff in the memory itself).
