CSCE 2610: Examination- February 19, 2018 1:30pm-2:30pm Open Books-Open Notes Your Name: he wot lgos te wiein any and all assamptions your ene making Yow will receive partial credti are neasomaile wee ine aproblemthon the weights sagresa. For example, yow should not spend b.. (25%) You are told that on awerage, ARM processors see 309. load/store 20% branch, and S0% anthmeti logic instructions. You are also given that loadstone instructions take 3 eyoles, bewnch instructions take 2 cycles and arithmetic/logic instructions take 1 cycle. You are coasislkoring two possible bardware optimizations. cis Reduse execution of kodstore to 2 cycles. But this reduces clock speed by 20% Redluce execution of branch instructions to 1 cycle. But this reduces clock speed Fur each optiinication, calculate the overall performance improvement achieved. 2. (35%) The able below shows the instruction count (IC) for three programs running on three PI, Pe, and P3 with the clock rates 1.0 GHz, 2.5 GHz. and 2.0 GHz respectively. program coesisis of oaly Load/store, arithmetic, and branch instructions Load Store Arithmetic Branch 125 300 200 150 100 progrun 1 a Processor Pl 15 20 Ke Ror program 3 on Processor P3 150 a), (15% what is the CFt of each program? b), (10%)what are the execution times (or CPU times) of the programs? e). (10%) what are the MIPS ratings on each processor? (40%) Write ARM8 (or LE08)code for each ofthe following Csegments. Use registers XI, N2 A3 X4 r a h i and a. The base address of array d is in X0. 3. a) b) else dla) b-16