Precision 7530 & Precision 7730 owner's thread

Texture / Pixel fill rates are completely different than clock speeds. They are roughly double on the 7730's version of the P3200

What makes Max-Q "Q" is their special software and meeting some noise limit spec Nvidia outlines, and some other licensing crap. Otherwise OEMs have always been free to Firmware limit their GPUs as they see fit... Which really sucks, and we try to uncover the culprits and let it be known. (Razer Blade Pro 17" GTX 1080 is a classic example of this)

 

The pixel and texel fillrates listed on the main GPU-Z tab are calculated based on GPU base clock and ROP/TMU count respectively. It lists 709MHz as the base clock, but it actually runs at 1250MHz under load, therefore the fillrates are not accurate.

 

Ok, gotcha

 

Here is the Hwinfo shot after Heaven on the above settings.

Let me know if you guys need anything else

 

Very nice. Can you run 3DMark Fire Strike and show the scores? Thanks for obliging.

 

Could you run a 3DMark Fire Strike bench and take a screenshot of the results page? Your screenshot proves that the P3200 in the 7730 has a ridiculous 40% higher power limit. I'd like to see exactly how close to the GTX 1070 the unfettered, un-crippled P3200 is.

Heh, birds of a feather.

 

Deleted.

 

Very curious to see the Fire Strike graphics score... I'm predicting anywhere between 12000-12500.

The only reason I can see Dell aggressively limiting the P3200 is to improve the thermal stability of simultaneous workloads. I have a feeling that if the P52 were to attempt such a workload, such as some CPU+GPU intensive games, it would throttle both pieces of hardware whereas the 7530 would only need to throttle the CPU.

This is the only rational conclusion I can come to. You'd like to think Dell's engineers are smart and have their reasons. Transparency is the only thing missing here. If @Dell-Mano_G could get us a word, that'd be terrific.

 

Here you go, think i got it right

 

Not quite that much. The P52 is about 5W higher and the 7730 is about 15W higher.

As @Mazed's 12.4K FS Graphics score shows, the normal TDP P3200 is still not close a 1070 variant. 1070MQ 80W 14K+, 1070MQ 90W 15K+, 1070N 115W 17K+.

It seems to be held back quite a bit by its memory. 1792 CUDA cores running at 1550MHz+ is actually more compute power than some of the 1070MQ's, certainly the 80W variants running at 1300MHz under load, but the 1070's have a big BW advantage. If one can use Nvidia Inspector to raise mem offset by +500MHz or more stably on the P3200, I'm thinking it significantly closes the gap with the 1070MQ 80W.