Tries : 321
Update Log
2016/10/04 : Trying dropping data stack (won't fit, gets expensive with adders if doing indexed access). Have something figured out for DSP inputs. Rethinking ISA, have place holder for now to see that things fit in worst case. Thinking estimated 81% LUT usage so far will be too much, perhaps drop accumulator down from 40-bits to 36-bits or 32-bits? The problem I see with this design is that due to load/store being at the end of the pipeline, after a store, the next instruction will be starved of inputs. Next try (4) will likely focus on {load/store, mux input, mul, add} ordered pipeline, which will attempt to make {load,dsp}, {store,return}, {dsp,call/jmp}, VLIW instruction usage work well, but will likely require a forth like address register (to avoid 2 dependent loads before feeding the dsp).
Notes
Avnet AES-KU040-DB-G (XCKU040 Based Dev Board)
Bit Hacks
Bit Manipulation Instruction Sets
A Fast Carry Chain Adder for Virtex-5 FPGAs
GRVI Phalanx: A Massively Parallel RISC-V FPGA Accelerator Accelerator
On Hard Adders and Carry Chains in FPGAs
Principles of FPGA IP Interconnect
Update Log
2016/10/04 : Trying dropping data stack (won't fit, gets expensive with adders if doing indexed access). Have something figured out for DSP inputs. Rethinking ISA, have place holder for now to see that things fit in worst case. Thinking estimated 81% LUT usage so far will be too much, perhaps drop accumulator down from 40-bits to 36-bits or 32-bits? The problem I see with this design is that due to load/store being at the end of the pipeline, after a store, the next instruction will be starved of inputs. Next try (4) will likely focus on {load/store, mux input, mul, add} ordered pipeline, which will attempt to make {load,dsp}, {store,return}, {dsp,call/jmp}, VLIW instruction usage work well, but will likely require a forth like address register (to avoid 2 dependent loads before feeding the dsp).
Notes
Related Material
==================
CORE RESOURCES
==================
Core functional units,
16-bit x 8-entry return stack (1 port)
32-bit x 8-entry register file (2 ports, port 0 for DSP input, port 1 for BRAM address)
32-bit x 1024-entry BRAM (2 ports, port 0 for instruction fetch, port 1 for data)
===================
EXECUTION MODEL
===================
4 threads/core of execution running with guaranteed round-robin scheduling
Instructions are VLIW style with a fixed logical ordered set of operations
Order Operation
===== =========
1st mux inputs for DSP ............ uses loads from prior instruction
2nd DSP execution .................
3nd load/store to REG and BRAM .... can store DSP result
4th branch ........................ can branch to DSP result
=====================
PHYSICAL PIPELINE
=====================
Designing under the following constraints,
Loads from RAMs are not used until next stage
Each stage does only one LUT or ADD both of which can be vertically chained
DSP is fully pipelined
Outline,
DSP DSP DSP DSP DSP DAT ADR BLK BLK
Stage A B MUL C ADD P REG REG OUT RAM PC INS
===== === === === === === === === === === === ===
0 lut lut @
1 mul reg lut
2 add lut add
3 lut @! lut @! lut @
----- --- --- --- --- --- --- --- --- --- --- ---
DSP A B .... DSP input a and b arguments
DSP MUL .... DSP mul stage
DSP C ...... DSP input c argument
DSP ADD .... DSP add/op stage
DAT REG .... Register file data load/store
ADR REG .... Register file address register to BRAM address translation
BLK OUT .... BRAM data construct write value
BLK RAM .... BRAM data load/store
PC ......... Update program counter
INS ........ Fetch next instruction
=======
ISA
=======
Alphabet usage,
a ... 24-bit DSP input (sign extended)
b ... 16-bit DSP input (sign extended)
c ... 48-bit DSP input, accumulator
d ... 32-bit data register index
f ... 32-bit BRAM fetched value
i ... 10-bit immediate
j ... 10-bit program counter
m ... 2-bit BRAM memory mode
p ... 48-bit DSP output
s ... 3-bit BRAM address base register index
w ... 32-bit BRAM write value
Encoding,
11111111111111110000000000000000
fedcba9876543210fedcba9876543210
================================
......................iiiiiiiiii Immediate 10-bits
================================
.................mmsss.......... BRAM control
.................00............. f=[s]
.................01............. f=[s^i]
.................10............. [s]=w
.................11............. [s^i]=w
...................sss.......... Register file address register index
================================
...oooooaabbccddd............... DSP control
...ooooo........................ Opcode (todo)
........aa...................... DSP a input choice
..........bb.................... DSP b input choice
............cc.................. DSP c input choice
..............ddd............... Register file data register index
================================
ggg............................. PC control
???............................. No branch
???............................. Return
???............................. Call immediate
???............................. Call to p from end of prior instruction
???............................. Conditional jump immediate if p<0 from end of prior instruction
???............................. Conditional jump immediate if p>=0 from end of prior instruction
???............................. Jump immediate
???............................. Jump to p from prior instruction
============================
CURRENT LUT BUDGET USAGE
============================
Budget is 400 LUTs/core, adding as design is roughed out,
LUTs % usage
==== === =====
32 8 register file (4x 8-LUT SLICEM 32-entry x 8-bit 2 port RAM)
8 2 return stack ( 8-LUT SLICEM 32-entry x 16-bit 1 port RAM)
---- --- -----
20 5 DSP p output modifier
102 26 DSP c input
48 12 DSP b input
24 6 DSP a input
18 5 BRAM address generation
34 9 BRAM output generation
38 10 program counter
==== === =====
324 81 total
=====================
DSP BIT ALIGNMENT
=====================
DSP hardware does p=op(join(a,b),c) or p=c+mul(a,b) or p=c-mul(a,b)
25-bit a
18-bit b
48-bit c
48-bit p
Designing for the following,
222222222222222211111111111111110000000000000000
fedcba9876543210fedcba9876543210fedcba9876543210
================================================
s........................ 25-th bit is sign extended
.aaaaaaaaaaaaaaaaaaaaaaaa using only 24-bits of a
================================================
................00 2 lower bits are set to zero
bbbbbbbbbbbbbbbb.. 16-bits of b shifted left by 2
================================================
aaaaaaaaaaaaaaaaaaaaaaaabbbbbbbbbbbbbbbb00 extent of join(a,b) input (40-bit effective)
================================================
........................................00 2 lower bits are set to zero
........................................1. option to set to one for rounding
........cccccccccccccccccccccccccccccccc.. 32-bits of c shifted left by 2
cccccccc.................................. maintaining extra 8-bits of p
================================================
The reason for the 2-bit padding,
Can reuse the join(a,b) cases as a>>16 for the multiply input
=======================
DSP INPUT CHALLENGE
=======================
Need to support both
BRAM unpack in the LUT
DSP a and b inputs separate
DSP a and b inputs joined {MSB a, b LSB}
Which forces needing unpack options with >>16 (not going to fit)
Solving this by
Only supporting unpack in c and b
The {MSB a, b LSB} input is only going to support p (accumulator feedback)
The c input is reused for the non-accumulator input
This works out because only subtract in {MSB a, b LSB} cases is non-associative
The mul cases will still use c as an accumulator
Not supporting unpacking in the a input
=========================
DSP P OUTPUT MODIFIER
=========================
Placement in pipeline,
stage action
===== ======
0
1
2
3 modify p post DSP for input for next instruction
Can conditionally zero p if signed or unsigned,
inputs for pair of 5:1 functions in a LUT
=========================================
2 p bits
1 p sign bit (might want the overflow sign bit?)
1 enable bit
1 signed or unsigned control bit
LUT area estimate,
LUTs usage
==== =====
20 p output modifier
===========================
DSP C INPUT
===========================
Need to also unpack BRAM load options so this gets expensive
Placement in pipeline,
stage action
===== ======
0 LUT DSP c input
1 register c
2 register pre-translated address for stage 0 of next cycle
3 register pre-translated address for stage 0 of next cycle
The c input expanded with unpack options, and control bits,
76543210fedcba9876543210fedcba9876543210 n LUT input count
======================================== = ===============
<-----------------------------iiiiiiiiii i 1-bit
pppppppppppppppppppppppppppppppppppppppp p 1-bit
<-------dddddddddddddddddddddddddddddddd d 1-bit
========================================
<-------aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa f 4-bits for MSB 8-bits of output
<-----------------------bbbbbbbbbbbbbbbb 7-bits for 2nd LSB 4-bits of output
<-----------------------cccccccccccccccc 15-bits for LSB 4-bits of output
00000000000000000000000000000000dddddddd
00000000000000000000000000000000eeeeeeee
00000000000000000000000000000000ffffffff
00000000000000000000000000000000gggggggg
000000000000000000000000000000000000hhhh
000000000000000000000000000000000000iiii
000000000000000000000000000000000000jjjj
000000000000000000000000000000000000kkkk
000000000000000000000000000000000000llll
000000000000000000000000000000000000mmmm
000000000000000000000000000000000000nnnn
000000000000000000000000000000000000oooo
========================================
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx needs 2-bit opcode control
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx needs 2-bit MSB of pre-translate address
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx........ needs 1-bit LSB of pre-translate address
................................xxxx.... needs 2-bit LSB of pre-translate address
....................................xxxx needs 3-bit LSB of pre-translate address
========================================
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx........ 12:1 function (2 LUT/bit) x 32-bit = 64 LUT
................................xxxx.... 15:1 function (4 LUT/bit) x 4-bit = 16 LUT
....................................xxxx 24:1 function (4 LUT/bit) x 4-bit = 16 LUT
LUT area estimate,
LUTs usage
==== =====
96 generate c
6 to register 5-bits x 2 stages of pre-translate address (rounded up)
---- -----
102 total
===========================
DSP B INPUT
===========================
Less expensive compared to c because of less bits
Placement in pipeline,
stage action
===== ======
0 LUT DSP b input
1
2
3
The b input expanded with unpack options, and control bits,
fedcba9876543210 n LUT input count
================ = ===============
<-----iiiiiiiiii i 1-bit
pppppppppppppppp p 1-bit
dddddddddddddddd d 1-bit
================
aaaaaaaaaaaaaaaa f 4-bits for MSB 8-bits of output
bbbbbbbbbbbbbbbb 7-bits for 2nd LSB 4-bits of output
cccccccccccccccc 15-bits for LSB 4-bits of output
00000000dddddddd
00000000eeeeeeee
00000000ffffffff
00000000gggggggg
000000000000hhhh
000000000000iiii
000000000000jjjj
000000000000kkkk
000000000000llll
000000000000mmmm
000000000000nnnn
000000000000oooo
================
xxxxxxxxxxxxxxxx needs 2-bit opcode control
xxxxxxxxxxxxxxxx needs 2-bit MSB of pre-translate address
xxxxxxxx........ needs 1-bit LSB of pre-translate address
........xxxx.... needs 2-bit LSB of pre-translate address
............xxxx needs 3-bit LSB of pre-translate address
================
xxxxxxxx........ 12:1 function (2 LUT/bit) x 8-bit = 16 LUT
........xxxx.... 16:1 function (4 LUT/bit) x 4-bit = 16 LUT
............xxxx 25:1 function (4 LUT/bit) x 4-bit = 16 LUT
LUT area estimate,
LUTs usage
==== =====
48 generate b
===========================
DSP A INPUT
===========================
No need to unpack for this input
Placement in pipeline,
stage action
===== ======
0 LUT DSP a input
1
2
3
Inputs can use simple 4:1 MUX,
76543210fedcba9876543210
========================
pppppppppppppppppppppppp this is p>>16
<-------------iiiiiiiiii
pppppppppppppppppppppppp
dddddddddddddddddddddddd
LUT area estimate,
LUTs usage
==== =====
24 generate a
===========================
BRAM ADDRESS GENERATION
===========================
Supports the feature of variable-bit width windows into the 4KB of ram
Placement in pipeline,
stage action
===== ======
0 fetch base address from register file
1 optionally XOR immediate
2 translate into BRAM address
3
Implementation requires XOR control to be single bit in opcode
BRAMs always in 32-bit port mode,
fedcba9876543210
================
.xxxxxxxxxx00000 - requires 10-bit address
Address register,
fedcba9876543210 access
================ ======
00....xxxxxxxxxx 1024 x 32-bit
01...xxxxxxxxxxx 2048 x 16-bit
10..xxxxxxxxxxxx 4096 x 8-bit
11.xxxxxxxxxxxxx 8192 x 4-bit (supported for read only)
Address register value to BRAM address translation
Uses a 6:1 function for each bit,
bits meaning
==== =======
4 address shifted left {0,1,2,3} bits
2 the 'fe' address bits
LUT area estimate,
LUTs usage
==== =====
8 optional XOR (16-bits at 2-bits per LUT), rounding up for ending register
10 translate (10-bits x 1 LUT)
---- -----
18 total
===========================
BRAM OUTPUT GENERATION
===========================
Shifts DSP p output for store, and compute byte write mask
Placement in pipeline,
stage action
===== ======
0
1
2
3 LUT new output here
Permutations (showing address and byte write mask for store),
11111111111111110000000000000000
fedcba9876543210fedcba9876543210
================================
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa - 32-bit adr=00....xxxxxxxxxx write=1111
................bbbbbbbbbbbbbbbb - 16-bit adr=01...xxxxxxxxxx0 write=0011
cccccccccccccccc................ - 16-bit adr=01...xxxxxxxxxx1 write=1100
........................dddddddd - 8-bit adr=10..xxxxxxxxxx00 write=0001
................eeeeeeee........ - 8-bit adr=10..xxxxxxxxxx01 write=0010
........ffffffff................ - 8-bit adr=10..xxxxxxxxxx10 write=0100
gggggggg........................ - 8-bit adr=10..xxxxxxxxxx11 write=1000
............................hhhh - 4-bit adr=11.xxxxxxxxxx000
........................iiii.... - 4-bit adr=11.xxxxxxxxxx001
....................jjjj........ - 4-bit adr=11.xxxxxxxxxx010
................kkkk............ - 4-bit adr=11.xxxxxxxxxx011
............llll................ - 4-bit adr=11.xxxxxxxxxx100
........mmmm.................... - 4-bit adr=11.xxxxxxxxxx101
....nnnn........................ - 4-bit adr=11.xxxxxxxxxx110
oooo............................ - 4-bit adr=11.xxxxxxxxxx111
Shift value for store,
Requires 3:1 MUX per bit, 32 LUTs
Generate write enable for store,
Requires same 4-bits per function,
2 lower address bits
2 upper address bits
2 LUTs (5:1 function sharing inputs, 2 outputs per LUT)
LUT area estimate,
LUTs usage
==== =====
32 shift value for store
2 generate write enable
---- -----
34 total
===================
PROGRAM COUNTER
===================
10-bit program counter (PC)
Only lower 8-bits of PC increment on linear execution
Requires only an 8-bit PC+1 computation (one slice)
Placement in pipeline,
stage action
===== ======
0 register
1 register
2 increment PC
3 LUT new PC based on DSP p output and instruction opcode
PC function inputs per output bit (map to 13:1 function at 2 LUTs/bit),
bits meaning
==== =======
1 next PC if not branching (computed in prior stage)
1 top of return stack
1 immediate absolute branch address
1 DSP P output register (computed branch target in prior clock)
1 DSP P output register sign bit (for conditional branch)
3 bits from instruction opcode
LUT area estimate,
LUTs usage
==== =====
20 13:1 function for next 10-bit PC computation including instruction decode
8 PC+1 adder for 8 lower bits of PC
10 for 2 stage registers (2-bits/LUT)
---- -----
38 total
========================
INSTRUCTION PIPELINE
========================
Todo, remember to count cost to pipeline opcode bits through stages
===============
EXTRA NOTES
===============
=================================================
IDEA LIST
=================================================
Want to be able to use address register fetch for DSP source
Would be nice to get to 16-bit immediate to be able to do self modify immediates
=================================================
ADDRESS REGISTER XOR INSTEAD OF ADD IMMEDIATE
=================================================
Planning on [address ^ immediate] addressing
This removes an adder from the design
XOR is the same as [address + immediate] for an n-bit immediate
When lower n-bits of address are zero
Means data must be aligned to the nearest pow2 of maximum immediate offset
Using the following terms,
ggggggoooo
g = group bits (address bits choose the group of data)
o = offset bits (address bits are zero, immediate chooses element in group)
For bits in address which are not cleared (ie the group bits),
Setting bits in the immediate results in accessing a neighbor group
Regardless of the starting group in the address register
It is possible to roll through all aligned groups
But ordering is different based on starting group address
Example of group bits for address crossed with immediate
00 01 10 11
+-------------
00 | 00 01 10 11
01 | 01 00 11 10
10 | 10 11 00 01
11 | 11 10 01 00
===================================
BRAM VARIABLE BIT-WIDTH WINDOWS
===================================
Trying to support transparent pack/unpack of variable bit-widths from BRAM
Want zero impact to ISA, no special instructions
Instead dividing address range into windows of different bit-widths
Each address range addresses at a multiple of the bit-width
Effectively the high bits of address choose the bit-width
Store path limited to {8,16,32}-bit
Only using BRAM byte write mask to avoid any {read, modify, write}
Fixed signed vs unsigned configuration,
size choice
====== ======
32-bit signed (but doesn't matter)
16-bit going to go with signed (needed for vector or audio)
8-bit unsigned (keeps implementation simple)
4-bit unsigned for sure (sprites?)
==========================================
WORKING THROUGH OPTIONS DSP OPERATIONS
==========================================
Opcode forms,
p = c op ((a << 16) + unsigned(b))
p = c + (a * b)
p = c - (a * b)
Where op can be the following,
and .....
nand ....
nor .....
not .....
or ......
xnor ....
xor .....
Where the following can also be applied,
extra set c bit -1 to 1 (for rounding)
(a * b) can be forced to zero (nop)
((a << 16) + unsigned(b)) can be forced to zero (nop)
((a << 16) + unsigned(b)) can be forced to all ones
===============================================
WORKING THROUGH OPTIONS FOR A,B,C DSP INPUT
===============================================
DSP inputs (as they appear in the core),
24-bit a
16-bit b
40-bit c
Possible inputs,
10-bit immediate
16-bit top of return stack
32-bit register file load (from prior instruction)
32-bit BRAM load (from prior instruction)
40-bit DSP p output
====================
FAST ABS MIN MAX
====================
These need to work on the N-bit accumulator without precision loss
So using multiply stage is out
Min and max, where a is the accumulator, and b is the limit,
min(a, b) = ((a - b) & ((a - b) < 0 ? ~0 : 0)) + b
max(a, b) = ((a - b) & ((a - b) < 0 ? 0 : ~0)) + b
Want to be able do the following,
acc -= b;
acc = acc < 0 ? acc : 0; // want to fold this into prior operation
acc += b;
Have to either LUT or register p in stage 3,
Could LUT p to zero if signed or unsigned based on control bit
This works out to 2-bits/LUT (pair of 5:1 functions with same input)
So 20 LUTs total (same as just registering)
Plus likely need to decode control and enable from opcode in prior pass
inputs
------
2 p bits
1 p sign bit (might want the overflow sign bit?)
1 enable bit
1 signed or unsigned control bit
This enables min and max to work in 2 instructions without branching
Absolute value this way is not friendly for accumulator,
abs(a) = max(a, -a)
Instead leverage the "free" branching (single cycle abs),
...; if(acc>=0) goto then; // add branch to end of prior op
acc = -acc; // works since acc is now in a:b for 1ops and 2ops
then:
Avnet AES-KU040-DB-G (XCKU040 Based Dev Board)
Bit Hacks
Bit Manipulation Instruction Sets
A Fast Carry Chain Adder for Virtex-5 FPGAs
GRVI Phalanx: A Massively Parallel RISC-V FPGA Accelerator Accelerator
On Hard Adders and Carry Chains in FPGAs
Principles of FPGA IP Interconnect