Hi Jason,

Thanks a lot. I’m glad symjit has been useful.

I’ve added many features in the last year. However, the plan forward is
toward a maintenance phase with focus on correctness. There is already a
large set of tests that each version has to pass before it is released. I
hope to expand this corpus based on actual real world use cases.

Moreover, being a compiler, symjit is dependent on the underlying
architecture and operation system. I wonder if there a survey of the sympy
community showing what the most common hardware and os are.

Thanks,

Shahriar

On Thu, Nov 20, 2025 at 1:17 AM Jason Moore <[email protected]> wrote:

> Nice work Shahriar!
>
> I added symjit as a backend to PyDy recently and the code generation
> computation time is significantly faster than anything we have in SymPy.
> And the code it generates is very close to Fortran and C speeds. Shahriar's
> approach to generate machine code directly is very novel.
>
> We should add it as an optional backend to lambdify.
>
> Jason
> moorepants.info
> +01 530-601-9791
>
>
> On Wed, Nov 19, 2025 at 11:34 PM Shahriar Iravanian <
> [email protected]> wrote:
>
>> Dear All,
>>
>> I just released version 2.8.0 of Symjit (
>> https://github.com/siravan/symjit/). Symjit is a just-in-time (JIT)
>> compiler for SymPy. A quick review of major new features introduced in the
>> last few releases:
>>
>> v2.3: Adding multi-threading for vectorized code.
>> v2.4: Implementing common sub-expression elimination (CSE).
>> v2.5: Introducing a new intermediate-representation, allowing better
>> optimization (level-1).
>> v2.6: Using a graph-coloring algorithm for register allocation, improving
>> optimization (level-2).
>> v2.7: Supporting RISC-V architecture.
>> v2.8: Adding explicit looping constructs (Sum and Product). Allowing
>> calls to Python and other Symjit functions.
>>
>> As always, suggestions and comments are highly appreciated!
>>
>> -- Shahriar
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "sympy" group.
>> To unsubscribe from this group and stop receiving emails from it, send an
>> email to [email protected].
>> To view this discussion visit
>> https://groups.google.com/d/msgid/sympy/c0501194-38d6-43ab-a9ee-ec47d08419c7n%40googlegroups.com
>> <https://groups.google.com/d/msgid/sympy/c0501194-38d6-43ab-a9ee-ec47d08419c7n%40googlegroups.com?utm_medium=email&utm_source=footer>
>> .
>>
> --
> You received this message because you are subscribed to a topic in the
> Google Groups "sympy" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/sympy/2pPTSM0fV8g/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to
> [email protected].
> To view this discussion visit
> https://groups.google.com/d/msgid/sympy/CAP7f1Ajg-MSotMLzfCpW-L062wHSCfSybCgNgTXLz3P-KXMiAw%40mail.gmail.com
> <https://groups.google.com/d/msgid/sympy/CAP7f1Ajg-MSotMLzfCpW-L062wHSCfSybCgNgTXLz3P-KXMiAw%40mail.gmail.com?utm_medium=email&utm_source=footer>
> .
>

-- 
You received this message because you are subscribed to the Google Groups 
"sympy" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/sympy/CABEBXTanwiMSpqf4a2qP9jmQXg_DchuOuyji%3DbBBugGo9hRuyQ%40mail.gmail.com.

Reply via email to