Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Important question on verification

Status
Not open for further replies.

masai_mara

Advanced Member level 4
Joined
Aug 13, 2004
Messages
118
Helped
8
Reputation
14
Reaction score
2
Trophy points
1,298
Activity points
1,426
Hi folks,
I am thinking of making a change in jobs. I have been working mainly on design and functional verification but almost all of them were on FPGA's. I now intend to go into ASIC design/verification. Could you people please point me out the necessary things to update myself on. I dont think it matters much for design but verification of asics invloves a lot more than those targeted for fpga's. I will appreciate any feedbacks on this aspect. does new methodologies like ABV or systemC matter much?? anything that I need to concentrate in particular?


thanks in advance for your replies,
steve
 

gogogo

Member level 2
Joined
Jan 22, 2005
Messages
43
Helped
1
Reputation
2
Reaction score
0
Trophy points
1,286
Activity points
322
some books are useful to you.
<< Writing Testbenches--Functional Verification of HDL Models>>
<<system on chip verification methology>> Kluwer
 

realtek

Member level 5
Joined
Mar 15, 2004
Messages
89
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,288
Activity points
870
many different with FPGA ,
you need DC, PT,and write a lot of testbench
,model, call library, co-work with analog/layout team,
power issue, clock tree ........

(I suggest you should read synopsys SOLD (DC/PT part), better do some tutorial for beginning)


but it's more interesting when tapeout and IC works , good luck!!
 

william_qiu

Newbie level 5
Joined
Apr 19, 2005
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,367
Lots of things should be considered. The most important thing to be remebered is for FPGA you may change your design as will but for ASIC the chance will be rare.
 

rprince006

Member level 4
Joined
Jul 16, 2004
Messages
79
Helped
2
Reputation
4
Reaction score
0
Trophy points
1,286
Activity points
841
Hi Steve,

Basically, verification is a difficult task, and there are more and more new tools,
languages, and techniques developed for it, making the technical berriers even
higher.

there are system level, chip level, and module level verifications. However,
I have some work experience with chip and module level verification. For functional verification, traditional methods are HDL test-bench based, you need
to know how to organize a verification environment using HDL. the two books
listed above are good ones to read, plus, you may find it good to know about
HDL API (VerilogPLI or VHDL FLI). However, it seems a transition to new techniques in functional verification, like HVLs(specman, vera), formal methods.
(it is said systemVerilog is good both for design and verification)

For system level, you may need an ESL, like SystemC or c/c++.

you may also need to know some script languages, like unix shell, perl, etc.
 

foster_cn

Member level 4
Joined
Jan 14, 2003
Messages
73
Helped
3
Reputation
6
Reaction score
1
Trophy points
1,288
Activity points
442
for scripts, as rirince006 mentioned, TCL is also very important, especially for synopsys tools
 

masai_mara

Advanced Member level 4
Joined
Aug 13, 2004
Messages
118
Helped
8
Reputation
14
Reaction score
2
Trophy points
1,298
Activity points
1,426
Thanks for all the answers. I do know some scripting in TCL and have developed test benches mainly in VHDL for verification of module level blocks and also some chip level test benches for medium sized cores. So are there any major points to keep in mind when I shift to ASICS. what is the generally practiced mehtodology for verification of ASICS and how does it differ from FPGA's apart from being more rigorous. for example what sort of coverage do we aim for?

thanks,
steve
 

rprince006

Member level 4
Joined
Jul 16, 2004
Messages
79
Helped
2
Reputation
4
Reaction score
0
Trophy points
1,286
Activity points
841
Hi,

I have less knowledge about FPGA, and the difference between FPGA and ASIC.

Using HDL to write test-bench is a tranditional way, and it is difficult to cover all
functional tests with only hdl based metthds.

Basically, there are two kind of coverage: code coverage and functional coverage,
and you need specific tools to collect coverage info and evaluate your effort.
 

Kulprashant

Member level 3
Joined
Mar 28, 2005
Messages
59
Helped
3
Reputation
6
Reaction score
1
Trophy points
1,288
Activity points
1,813
hi,
Writing Testbenches--Functional Verification of HDL Models jarnick.

with regards,
kul
 

mason_y

Newbie level 1
Joined
Dec 10, 2003
Messages
1
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
8
Other useful book:
L. Bening, H. Foster, "Principles of Verifiable RTL Design", Kluwer Academic Publishers,. 2001, ISBN: 0792373685
good luck!
 

sarath51

Full Member level 2
Joined
Dec 11, 2002
Messages
134
Helped
11
Reputation
22
Reaction score
4
Trophy points
1,298
Activity points
985
If u'e concern is in verification it is better to focus on Systerm verilog where which has enhanced features for assertion new constructs for verification
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top