M1 processor compatibility

Question

Hello,

Does the iTensor C++ version runs without any issue with Apple's new M1 processor? Or is there any possible decrease in performance?

Thank you!
Jun

Answer 1

Hi Jun,
Thanks for the interesting question. So I just tried the C++ version of ITensor on a Macbook Air with the M1 processor, and it was a smooth experience. I just cloned ITensor from Github, and then used the options.mk.sample file provided as the template for options.mk. The only change I needed to make was to use clang as the C++ compiler, rather than g++. Otherwise I got an error about the -fconcepts flag not being recognized. I did not try to install MKL and build ITensor against that; it would be an interesting thing to try and important to compare the speed to Apple's default "vecLib" framework (default BLAS provided by MacOS).

The performance was very good. The speed I got running the exthubbard sample code was comparable to a workstation at our office that has a 24-core, 3.4 GHz Xeon chip, whereas from what I have read, the M1 has 8 3.2 GHz cores. It was more than two times faster than my Intel Macbook Pro which has a 2.9 GHz chip!

I also installed Julia and the Julia version of ITensor, ITensors.jl, without any issue. That surprised me because I had read that Julia was not yet working for the M1, so perhaps it is using the Rosetta compatibility layer? Also I was able to install the MKL Julia package which we have found significantly speeds up ITensors.jl on Intel chips. It would be interesting to compare Julia with and without MKL on the M1 chip.

Best regards,
Miles

Comments

I should note that for a 2d DMRG calculation of the Heisenberg model (in Julia) with conserved quantum numbers, the 24-core workstation was faster than with the M1. So my guess is that this particular calculation made better use of the many cores, but we would need to do more extensive testing to be sure of the reason.

It could also very well be that the MKL.jl library doesn't provide the same sort of relative advantage over Julia's default of OpenBLAS on the M1 as on Intel chips.

---

Hi Miles,

Thanks a lot for the very thorough investigation! Good to know that it works very well and also has some room to be even better.
Personally, I'm with the C++ version so it looks like the M1 chip would not be a problem for me :)
Thanks again, and hope to see you in the near future!

Best,
Jun

---

Hi Miles,

I couldn't succeed in the installation of ITensor in the MacBook Air M1.
Following your steps and all installation steps from http://itensor.org/docs.cgi?vers=cppv3&page=install
I am still getting an error: tensor/all.h file not found
Could you help me, please? Should do I do something beyond it?

The best,
Caroline

---

Hi Caroline,
Which step did you get this error on? Were you compiling the library or a code based on the library?

Also I notice you wrote tensor/all.h but it should say itensor/all.h. Did you mean itensor instead of tensor?

Lastly have you considered using the Julia version of ITensor? Just asking in case it is an option for you because it is easier to install and has more and newer features than the C++ version.

Miles

---

Hi Miles,

Thanks for the fast reply.

I got the error when I am trying compiling a code based on the library.

Sorry, I meant itensor/all.h

I notice the problem is in #incluse "itensor/all.h". So if I forced it by putting all the file path: users/local/itensor/all.h, there is another error: itensor/all_mps.h. It seems it cannot find the path of the folder with all files .h.

In Julia, it works well. However, if I work with C++, it will save me time.

The best,
Caroline

---

Thanks for the additional information.

Is this error happening for the step of compiling the library itself? Or a code based on the library?

---

It is happing when I try to run the file myappname.cc from tutorial/project_tamplate. Or even when trying to run Hello_ITensor.cpp.

---

I see –  did you adjust the LIBRARY_DIR variable inside the Makefile? Does it point at the correct directory where you placed the ITensor library?

---

Yes, I did.

LIBRARY_DIR=/Users/myfolder/itensor

Should I change here too:
include $(LIBRARY_DIR)/this_dir.mk
include $(LIBRARY_DIR)/options.mk

---

Let's discuss this more via email – I think I will need to ask you more questions and that will be quicker.

---

Ok, no problem. Thanks.