[quotemsg=19460340,0,482859]The GTA V benchmark is consistent if you record only the last (5th) part of the ingame-benchmark. It is also long enough and a lot more real than all other parts of this in-game benchmark. I run it scripted three times - with a hot card. The results are in each case very stable. And I'm not using, what GTA is writing into the file. We are monitoring the output with our software.
All benchmark runs were made with heated cards in a closed system, maybe you read my review about the new bench table. That means, my results are mostly worse than in other reviews on an open bench table (Boost). And exactly this is another reason, why such results are not really comparable. It give you a direction, a trend - but nothing more
From my sight:
It is very difficult to compare the same card on different systems. But different cards (with the same GPU inside) on different systems is even more a big mess. It is simply not fair to think, that card A on system B is faster or slower than card C on system D (and vise versa). I like comparisons under 100% equal conditions, but all this crossover-comparisons are a pain.
I try to test all products of a roundup with frozen drivers and windows updates to have over this period the same conditions. The reader is updating his system as often as he can and some driver changes may cause dramatically different results. I had in the past very often such situations, when ppl compared their fresh stuff with a six month old roundup and called me a Nvidia or AMD fanboy, depending at the drivers and the preferred product.
As I wrote in the roundup:
The goal of such a roundup is in the first line a fair technical analysis and comparison of different solutions, not a collection of some benchmark bars. Benchmarks you can find tons in the net, bars in each color and size that you prefer. But the main difference between this tested cards are not the bars, but other fundamental things like cooling, power draw and finally: issues and broken standards.
But if it is sooooo important - let me check, how we can add Firestrike or something similar in the next reviews. Just preparing the first 1080 Ti custom for this.[/quotemsg]
[quotemsg=19468488,0,992934][quotemsg=19457023,0,1069610]Darnit, the Zotac and Gigabyte mini models weren't covered. I was curious to see how those two compare vs. the EVGA model since I'm working with limited space inside my case.[/quotemsg]
FWIW, I bought the Zotac Mini over the summer when nothing else was in stock, and I'm very pleased. It's a little loud when the fan revs up, but despite the simplistic cooler it runs at reasonable temperatures -- and I was relieved to see that it has heatsinks over the VRMs.
That last point is gratifying, because until I read this article I assumed the EVGA mini would have been superior. Now I begin to doubt it, though of course FormatC's little mod probably makes up the difference.[/quotemsg]
Facebook
Twitter
Instagram