GPUZ tells me this is a AMD Radeon 7600M. I know it is no power house, but I am pretty disappointed with the performance I am getting out of it. In comparison, my much older 9800 GTX+ is WAY faster. I have tried both the stock app, and the Lunatics apps. GPUZ tells me that the GPU is running at 97%. Is this what I can expect from this GPU or is there something else wrong?

---

You may wish to run AP on that GPU instead of MB work. NV GPU with CUDA app are most efficient at MB & ATI GPU with Open CL are most efficient at AP. At least that was the last news I heard on status of apps.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

Though it might be newer, the HD 7600M series appears to be based on the same Turks GPU core that was first introduced with the desktop HD 6570/6670 GPU. These are fairly lightweight in terms of performance and add the fact that mobile GPUs generally aim towards minimising power usage rather than maximising performance, it's not surprising that an older desktop GPU such as the 9800 GTX+ can still outstrip your HD 7600M GPU in terms of performance.

---

Soli Deo Gloria

Oddly enough the 9800 GTX+ is rated at 506 GFLOPS & an HD7670m is rated at 576 GFLOPS. So the mobile card does have more oomph.
However as I stated before the types of applications on each hardware have different efficiencies. For multibeam work an ATI GPU of the same GFLOPs rating as an Nvidia seems to run about about twice as long. Running astropulse it would go the other way.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

I'm not sure those GFLOPS ratings are all that accurate. I recognise GPGPU performance is not directly comparable with graphical performance, but the GPU hierarchy at Tom's (admittedly, only a guide and not definitive) suggests the 9800 GTX+ to be substantially better performing than the HD 6570/6670 that forms the basis of the HD 7600M series.

That aside, I'm aware that the current state of the applications are such that AMD GPUs seem to do better than comparable NV hardware on AstroPulse while the reverse is true for MultiBeam.

---

Soli Deo Gloria

There is an old thread about running MBs on an ATI card; I'm falling, I bought a parachute. From 100% AP, to 100% MB.
It's still about the same. ATI/AMD cards produce about 75% Less credit when running MBs verses APs. The difference with my ATI 6970 is typical.
7 minutes of MB = 40 points
21 minutes of AP = 550 points.
Or,
21 minutes of MB = 120 points.
21 minutes of AP = 550 points.
Which would you rather run?

Try an AP on the OP's NV 9800, it will take around 5-6 hours.
Try an AP on the OP's AMD 7600, it should take less than 2 hours.

There is a reason Raistmer recommends running MB on nVidia cards and APs on AMD cards.

The HD 7600M established an APR of 166 GFLOPS doing AP v6 tasks, and AP v7 should stabilize slightly higher. The 9800 GTX+ has an APR of 84 GFLOPS doing AP v7 tasks, so there are figures on Dragokatzov's systems supporting the idea that the 7600M is around twice as good for AP.

Even for SaH v7 work, one could argue that the 7600M is better because it is using about 1/7 the power of the 9800 GTX+.

Add that the laptop has a quad core i5-3210M processor which both outclasses the E8400 Core 2 duo in the other machine and uses less power, and I consider the laptop the superior crunching machine (assuming its cooling is good enough).

                                                                   Joe