image/svg+xml Front End InstructionCache Tag µOP CacheTag L1 Instruction Cache32KiB 8-Way InstructionTLB Instruction Fetch & PreDecode(16 B window) Instruction Queue MOP MicroCodeSequencerROM(MS ROM) Decoded Stream Buffer (DSB)(µOP Cache)(1.5k µOPs; 8-Way)(64 B window) BranchPredictor(BPU) Allocation Queue (IDQ) (128, 2x64 µOPs) L2 Cache1 MiB 16-Way Unified STLB Execution Engine Memory Subsystem L1 Data Cache32KiB 8-Way Data TLB SchedulerUnified Reservation Station (RS)(97 entries) Integer Physical Register File(180 Registers) Vector Physical Register File(168 Registers) Port 0 Port 1 Port 5 Port 6 Port 2 Port 3 Port 4 Port 7 INT ALU INT DIV INT Vect ALU INT Vect MUL FP FMA AES Vect String FP DIV INT ALU INT MUL INT Vect ALU INT Vect MUL FP FMA Bit Scan INT ALU LEA INT ALU Branch AGU Load Data AGU Load Data AGU (56 entries) Store Buffer & Forwarding 64B/cycle µOP µOP µOP µOP µOP µOP µOP µOP (50, 2x25 entries) 16 Bytes/cycle µOP µOP µOP µOP µOP µOP Macro-Fusion MOP MOP MOP MOP MOP MOP MOP MOP MOP MOP Micro-Fusion 64B/cycle 64B/cycle StackEngine(SE) Adder Adder Adder ≤4 µOPs µOP µOP µOP ComplexDecoder 4-Way Decode SimpleDecoder SimpleDecoder SimpleDecoder ≤4 µOPs MUX ≤5 µOPs Loop StreamDetector (LSD) Register Alias Table (RAT) 4 µOP Branch Order Buffer(BOB) (48-entry) Rename / Allocate / Retirement ReOrder Buffer (224 entries) Zeroing Idioms Move Elimination Ones Idioms Line Fill Buffers (LFB)(10 entries) Store Data 64B/cycle 64B/cycle 512bit/cycle Load Buffer(72 entries) ≤6 µOPs EUs µOP µOP µOP µOP µOP µOP µOP µOP Common Data Buses (CDBs) Int Int Vect FP Load Store Branch To L3 64B/cycle 512b fused INT Vect ALU INT Vect MUL FP FMA 512b