UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

cancel
Showing results for 
Search instead for 
Did you mean: 
Observer kaveirious
Observer
10,617 Views
Registered: ‎08-25-2007

ANN: YARDstick - custom processor development toolset

Dear friends,

I am very pleased and proud to announce YARDstick:
(http://electronics.physics.auth.gr/people/nkavv/yardstick),
a custom processor development toolset with an impressive list of features.

But, beware, the following amount of information might prove unhealthy to
low-life academic bozos that are unaware of real-life coding, are fans
of top-down programming (crap!, only professor's textbooks), UML (crap!), and hate
extreme programming.

So:

YARDstick is a novel design automation tool for custom processor development flows
that focuces on the hard part: generating and evaluating application-specific
hardware extensions. YARDstick is a powerful building block for ASIP
development, since it integrates application analysis, ultra-fast algorithms for custom
instruction generation and selection with user-defined compiler intermediate
representations. As of September 2007, YARDstick integrates retargetable compiler
features for the targeted IRs/architectures. Remarkable features of YARDstick are the
following:

- retargetable to used-defined IRs by machine description.
- can be targeted to low-level compiler IRs, assembly-level representations of virtual
  machines, or assembly code for existing processors.
- fully parameterized custom instruction generation and selection engine.
- lightning-fast code selector for multiple-input multiple-output patterns based on
  graph matching. It is known that the code selector scales very well with the instruction
  node count of basic block data-dependence graphs (successfully tested with custom
  instruction patterns of more than 30 nodes).
- virtual register assignment for virtual machine targets.
- an extensive set of backends including assembly code emitter, C backend, visualization
  backends for Graphviz and VCG (or aiSee), an XML format amenable to graph rewriting
  and others.

YARDstick comes along with a cross-platform GUI written in Tcl/Tk 8.5.

The ultimate goal of YARDstick is to liberate the designer's development infrastructure
from compiler and simulator idiosyncrasies. With YARDstick, the ASIP designer is empowered
with the freedom of specifying the target architecture of choice and adding new
implementations of analyses and custom instruction generation/selection methods.

At this moment, YARDstick is being heavily used for developing a new processor
architecture of mine with many never-being-seen features, mostly aiming FPGAs.
Status update report on the processor architecture should be expected near late
October 2007.

Typically, 2x to 15x speedups for benchmark applications (ANSI C optimized source code)
can be fully automatically obtained by using YARDstick depending on the
target architecture. Speedups are evaluated against a typical scalar RISC
architecture.

Detailed feature list:

  1. Analysis engines generating both static and dynamic statistics:
    - Data types
    - Operation-level statistics
    - Basic block statistics (ranking)
    - Performance estimations with/without custom instructions.

  2. Generation of CDFGs (Control-Data Flow Graphs).

  3. Backend engines:
    - ANSI C
    - dot (Graphviz)
    - VCG (GDL, aiSee)
    - XML (GGX for the AGG graph rewriting tool)
    - Retargetable assembly emitter for entire translation units (single files
    with multiple functions/procedures).
    - CDFG formats for various RTL synthesis tools.

  4. Custom instruction engines:
    - Full-parameterized MIMO custom instruction generation algorithm. Features:
      * Fast heuristic !!!
      * Configurable number of inputs
      * Configurable number of outputs
      * List of forbidden nodes
      * Node sorting strategies (3 different strategies!)
      * Transformation rule library for applying CFG transformation strategies

  5. Custom instruction selection:
    - Based on priority metrics (2 choices at the moment).

  6. Graph (and graph-subgraph) isomorphism features for eliminating redundant
  patterns. Multiple algorithms supported.

  7. Visualization of custom instructions, basic blocks, control-flow graphs and
  control-data flow graphs (basic block nodes expanded to their constituent
  instructions).

  8. Basic retargetable compiler features (alpha state):
    - Code selector for MIMO instructions (tested with large cases).
    - Virtual register assignment (allocation for a VM).
    - Hard register allocator in the works.

  9. Miscellaneous features:
    - single constant multiplication optimizer
    - elimination of false data-dependences in assembly-level CDFGs.
    - beautification options for visualization
    - interfacing (co-operation) with external tools such as peephole optimizers,
    profilers, code generators etc.
    - features related to the custom processor architecture (not to be disclosed
    yet)

Here is a list of application benchmarks that have been tested with YARDstick (compiler
features not fully tested):
  - ADPCM encoder and decoder (typically: 4x speedup)
  - Video processing kernels: full-search block-matching motion estimation, logarithmic
  search motion estimation, motion compensation
  - Image processing kernels: steganography (hide/uncover), edge detection, matrix
  multiplication
  - Cryptographic kernels: crc32, rc5, raiden (7x speedup, 12x for unrolled version)

At the YARDstick homepage:
http://electronics.physics.auth.gr/people/nkavv/yardstick/
you can find some additional material:

  - 2-page brochure
  - 2-page abstract for the DATE'07 University Booth
  - a more extended presentation on YARDstick

The above material refers to the status of April 2007.


Expected enhancements to YARDstick in the near future:
  - linear-scan and integer-linear programming based register allocators
  - bitwidth analysis
  - CDFG->VHDL generation of custom instruction hardware
  - algorithm implementation for CDFG pipelining


Interested parties are welcome to contact me for details on how to get access
to a demo version of the YARDstick toolset.


Kind regards

Nikolaos Kavvadias
Computer Architecture Specialist - Compiler Developer
Ph.D. candidate
M.Sc. Eletronics Engineering
B.Sc. Physics

You may contact me at:
Nikolaos Kavvadias <nkavv@physics.auth.gr>
http://www.geocities.com/kaveirious/
http://electronics.physics.auth.gr/tomeas/en/kavvadias.html

USAGE EXAMPLES

--
-- A Custom instruction generated from an edge-detection filter
--

1. Auto-generated C simulation code.

void main_9(
  int    *vr234_s32
  ,int    *vr235_s32
  ,int    vr60_s32
  ,int    vr207_s32
  ,int    vr210_s32
  ,int    vr220_s32
)
{
  int    vr228_s32;
  int    vr229_s32;
  int    vr230_s32;
  int    vr231_s32;
  int    vr232_s32;
  int    vr233_s32;
  *vr235_s32 = 1;
  vr231_s32 = vr207_s32-vr220_s32;
  vr232_s32 = ((vr231_s32 < 0) ? -vr231_s32 : vr231_s32);
  vr233_s32 = vr60_s32<vr232_s32;
  vr228_s32 = vr207_s32-vr210_s32;
  vr229_s32 = ((vr228_s32 < 0) ? -vr228_s32 : vr228_s32);
  vr230_s32 = vr60_s32<vr229_s32;
  *vr234_s32 = vr230_s32|vr233_s32;
#pragma cycles_est_total = 2
}

2. Auto-generated VCG (GDL) graph

graph: { title: "main_9"

x: 30
y: 30
height: 380
width: 560
xspace: 20
yspace: 30
display_edge_labels: yes
layoutalgorithm: minbackward
port_sharing: no
node.borderwidth: 3
node.color: white
node.textcolor: black
node.bordercolor: black
edge.color: black

node: { title:"0" shape: ellipse label:" ior" color:yellow }
node: { title:"1" shape: ellipse label:" sl" color:yellow }
node: { title:"2" shape: ellipse label:" abs" color:yellow }
node: { title:"3" shape: ellipse label:" sub" color:yellow }
node: { title:"4" shape: ellipse label:" sl" color:yellow }
node: { title:"5" shape: ellipse label:" abs" color:yellow }
node: { title:"6" shape: ellipse label:" sub" color:yellow }
node: { title:"7" shape: ellipse label:" ldc" color:yellow }
node: { title:"8" shape: rhomb label:" 1" color:magenta }
edge: {sourcename:"8" targetname:"7"}
node: { title:"9" shape: triangle label:" vr234.s32" color:cyan }
edge: {sourcename:"0" targetname:"9"}
node: { title:"10" shape: triangle label:" vr235.s32" color:cyan }
edge: {sourcename:"7" targetname:"10"}
node: { title:"11" shape: box label:" vr60.s32" color:green }
edge: {sourcename:"11" targetname:"1"}
node: { title:"12" shape: box label:" vr207.s32" color:green }
edge: {sourcename:"12" targetname:"3"}
node: { title:"13" shape: box label:" vr210.s32" color:green }
edge: {sourcename:"13" targetname:"3"}
edge: {sourcename:"11" targetname:"4"}
edge: {sourcename:"12" targetname:"6"}
node: { title:"14" shape: box label:" vr220.s32" color:green }
edge: {sourcename:"14" targetname:"6"}

edge: {sourcename:"3" targetname:"2" label:"vr228.s32" }
edge: {sourcename:"2" targetname:"1" label:"vr229.s32" }
edge: {sourcename:"1" targetname:"0" label:"vr230.s32" }
edge: {sourcename:"6" targetname:"5" label:"vr231.s32" }
edge: {sourcename:"5" targetname:"4" label:"vr232.s32" }
edge: {sourcename:"4" targetname:"0" label:"vr233.s32" }

}

3. Auto-generated Graphviz (dot) file

digraph main_9 {

node [fontname=Courier,fontsize=14,style=filled];
  0    [shape=ellipse,label="ior",fillcolor=yellow]
  1    [shape=ellipse,label="sl",fillcolor=yellow]
  2    [shape=ellipse,label="abs",fillcolor=yellow]
  3    [shape=ellipse,label="sub",fillcolor=yellow]
  4    [shape=ellipse,label="sl",fillcolor=yellow]
  5    [shape=ellipse,label="abs",fillcolor=yellow]
  6    [shape=ellipse,label="sub",fillcolor=yellow]
  7    [shape=ellipse,label="ldc",fillcolor=yellow]
  8    [shape=diamond,label="1",fillcolor=magenta]
  8 -> 7;
  9    [shape=triangle,label="vr234.s32",fillcolor=cyan]
  0 -> 9;
  10    [shape=triangle,label="vr235.s32",fillcolor=cyan]
  7 -> 10;
  11    [shape=invtriangle,label="vr60.s32",fillcolor=green]
  11 -> 1;
  12    [shape=invtriangle,label="vr207.s32",fillcolor=green]
  12 -> 3;
  13    [shape=invtriangle,label="vr210.s32",fillcolor=green]
  13 -> 3;
  11 -> 4;
  12 -> 6;
  14    [shape=invtriangle,label="vr220.s32",fillcolor=green]
  14 -> 6;

  3 -> 2    [label="vr228.s32"];
  2 -> 1    [label="vr229.s32"];
  1 -> 0    [label="vr230.s32"];
  6 -> 5    [label="vr231.s32"];
  5 -> 4    [label="vr232.s32"];
  4 -> 0    [label="vr233.s32"];

}

0 Kudos
4 Replies
Explorer
Explorer
10,613 Views
Registered: ‎08-13-2007

Re: ANN: YARDstick - custom processor development toolset

Hi
 
your project would have WAY MORE interest if you would provide some real downloads, not link to download 2 page PDF booklets only.
 
Antti Lukats
0 Kudos
Observer kaveirious
Observer
10,608 Views
Registered: ‎08-25-2007

Re: ANN: YARDstick - custom processor development toolset

Hi Antti

i always read forum messages, your opinion matters.

The story about YARDstick is that it represents about 2.5 years of work (Summer '04 till now). It has grown to about ~25Klines of C,C++ and Tcl/Tk code and has many (some of them rare to found) capabilities that are extremely useful for embedded/custom processor development teams.

I have hoped that, what is the best piece of software written in my department EVER, would earn me there a (permanent) research/lab assistant position. But it did not. So, right now i'm trying to see what would the open market think of YARDstick. Greek companies are laughable (and very low-tech) so was no hard decision to rule them out.

I believe that your remark is right, in the sense that i should make available more information on YARDstick. I'm not sure about a full-access demo, but maybe a restricted-access tech demo or a CGI-scripted webpage would be a first step. And of course, some complete examples linked from my webpage.

But tell me, what are "real" downloads? I have not yet set up a business and would first like to see what do development teams really miss and are willing to pay/cooperate for.

I believe that in some aspects, YARDstick is well ahead competition (although i donnot necessarily see anyone as competition).
1. Retargetability of IRs.
2. Many features regarding custom instruction generation and associated estimations.
3. Interfacing of external/third party tools and libraries. The more are supported the better :)

Ah, what do you think of the leaflets?


Kind regards
Nikolaos Kavvadias
PS: I'm looking for good co-ops as well.

0 Kudos
Explorer
Explorer
290 Views
Registered: ‎10-16-2018

Re: ANN: YARDstick - custom processor development toolset

Hi @kaveirious ,

"The requested URL /people/nkavv/yardstick was not found on this server"

Thanks

0 Kudos
Highlighted
Observer kaveirious
Observer
231 Views
Registered: ‎08-25-2007

Re: ANN: YARDstick - custom processor development toolset

Hi ahmed_alfadhel,

 

you are probably referring to the old university page for YARDstick. There is a maintained page here: http://www.nkavvadias.com/yardstick/

 

Please note that the tool has not been released (except for a brief demo stint) and there are no current plans to release an updated version. The most important reason is that it will be a big migration and maintenance effort and I don't have time for it.

 

Best regards

Nikolaos Kavvadias