fix cuTNPerm & cuGeSvd without is_U=false & attempt to fix cutt call …

[jeffry1829]

…& add stacktrace print in cytnx_error

[jeffry1829]

I recommend to merge this even though I can't figure out how to fix cuTT. As cuTT is actually contained in cuTensor and cuQuantum, and we won't need it.

[yingjerkao]

I recommend to merge this even though I can't figure out how to fix cuTT. As cuTT is actually contained in cuTensor and cuQuantum, and we won't need it.

Is it possible to drop cuTT dependence completely?

[jeffry1829]

Yes, I can do that

Ying-Jer Kao @.***>於 2024年10月1日 週二，下午4:33寫道：

I recommend to merge this even though I can't figure out how to fix cuTT. As cuTT is actually contained in cuTensor and cuQuantum, and we won't need it.

Is it possible to drop cuTT dependence completely?

— Reply to this email directly, view it on GitHub https://github.com/Cytnx-dev/Cytnx/pull/480#issuecomment-2385131768, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABUUA2MNBF5POKH6X6UF4TLZZJM4BAVCNFSM6AAAAABOYMERX6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOBVGEZTCNZWHA . You are receiving this because you authored the thread.Message ID: @.***>

[jeffry1829]

cuTensor contraction performance has been significantly improved by chooseing TGETT algo