Hi,
MAC OS X provides more FP efficiency than Win7 and Cuda is still aviable and up to Date ( Cuda 3.2.17 last ).
I got some MAC's which still awaiting for a nice application. Others like CC still producing just errors and wont provide any solution.
Hope you solve my "problem". ;)

---

Does the GPU Grid project have any plans to support running work units on Mac OS? Thanks.

Sandor Hasznos

I think its very unlikely that the researchers would be able to support another operating system. At GPUGrid the work changes quite regularly (lots of development and testing) and its difficult enough to make apps/tasks work on Linux and Windows for so many GPU types (GeForce 200 through to Keplers).
Even if the researchers had the resources and expertise it would be hard to justify, for what is likely to be a very small number users, as it would prevent other work.
There is only so much a small team can do and if they try to do too much they will get nothing done.

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

On the other hand, Mac OS is based on Unix and Linux and all its varieties are based on Unix as well. In the scientific world a lot of Macs are used and more and more Macs are introduced. At Universities as well.

I don't have a Mac because they have left the Power PC and use Intel now. No problem with that, but they are years behind with the hardware nowadays.

The GPUGRID project could use a internship for a IT student to program it for them in a mac version. But updating new functionality and tests will be a problem then. Then there must be an internship permanently. I don't know if a University in Spain can and will fund such an internship. Would be a good opportunity for programming student though.

---

Greetings from TJ

Nowdays there's a non-negligible number of Macs around. However, most of them are laptops, so by definition not very interesting for GPU-Grid. And while it's based on Unix, development has driven Mac OS and Linux so far apart that they require separate drivers. As far as I remember there's never been driver interchangability between them.

Since Mac OS supports CUDA the initial port should ideally be straight forward, although reality might have something to say about this as well. I suspect the real problem here would be drivers. Just look at the current problems GPU-Grid faces.. it should work on Win and Linux, with any half-recent driver and on any half-modern GPU. Yet what we're seeing is.. something absolutely different. So before talking about yet another platform let's concentrate on fixing the current one first.

MrS

---

Scanning for our furry friends since Jan 2002