AMD Bobcat (Zacate)
Configuration
AMD E-350 1600 MHz.
- L1 Data cache = 32 KB. 64 B/line, 8-WAY. Parity Protected, Copyback, Advanced 8-stream prefetcher.
- L1 Instruction cache = 32 KB, 64 B/line, 2-WAY, Parity Protected.
Fetch up to 32 bytes per cycle.
- L2 cache size = 512 KB. 64 B/line, 16-WAY. ECC protected, Half speed clocking for
power reduction.
- dual x86 instruction decoder: Scans up to 22 bytes, Decodes up to 2 x86 instructions per cycle,
The decoder can directly map 89% of x86 instructions to a single microOp, an additional
10% to a pair of microOps, and more complicated x86 instructions (<1%) are microcoded (Dynamic
Instruction Counts).
- Branch Predictor: Predicts up to two branches per cycle, Remembers branch
instruction locations, Return Stack Address Predictor, Indirect Dynamic Address Predictor.
- 2 ALU, LOAD, STORE, 2 * FPU.
- Integer Execution: A dual port integer scheduler feeds 2 ALUs,
A dual port address scheduler feeds a load address unit, and a store address unit.
Physical Register File uses maps and pointers to reduce power by minimizing data copying/movement.
- FPU: A centralized FP scheduler feeds two 64-bit FP execution stacks,
MMX and Logical units are replicated in both stacks, The FP Mul Unit can perform 2 SP
multiplies per cycle, The FP Add Unit can perform 2 SP additions per cycle,
A physical register file is used to reduce power.
- Out-of-Order Load Store Unit:
Loads bypassing loads, Loads bypassing stores, Stores bypassing loads,
Bypass tracking and dependency correction, Hazard predictor, Fast store forwarding,
Fast critical word fill forwarding.
- 64-bit DDR3.
- Bus Unit: 8-outstanding data accesses, 2-outstanding fetch accesses,
Eviction Buffers, Fill Buffers, Write combining buffers, Coherency management.
4 KB pages mode (64-bit Windows, 64-bit soft)
- DATA L1 TLB size = 40 items. Miss penalty = ?.
- DATA L2 TLB size = 512 items. Miss penalty = ?.
- Instruction TLB size = 512 items.
Size |
Latency |
Description |
32 K | 3 | TLB + L1 |
160 K | 20 | +17 (L1 miss) |
512 K | 27? | +7? (L1 TLB miss) |
2 M | 27? + ? ns | +RAM |
... | 27? + ? + ? ns | +? (L2 TLB miss) |
2 MB pages mode (64-bit Windows, 64-bit soft)
- DATA L1 TLB size = 8 items. Miss penalty = ?.
- DATA L2 TLB size = 64 items. Miss penalty = ?.
- Instruction TLB size = 8 items.
Integer pipeline:
# |
Name |
Stage |
Description |
1 | Fetch0 | Fetch | |
2 | Fetch1 | |
3 | Fetch2 | |
4 | Decode0 | Decode mCode Dispatch | |
5 | Decode1 | |
6 | Decode2 | |
7 | Pack | |
8 | FDeck | |
9 | Disptch | |
10 | Schedule | |
11 | RegRead | |
12 | ALU | |
13 | Writeback | |
Links
Bobcat at Wikipedia
Bobcat. AMD's New Low Power x86 Core Architecture. Brad Burgess. Hot Chips.