i never saw the point in synthetic benchmarks besides they look pretty but i really do not care if my gpu gets a certain score in 3dmark or my cpu gets a certain score in sandra. What i would rather know is how well gpu does in the witcher 2 or the cpu in handbrake/
Watched a previous tek video and never knew about the optimisations for intel in a benchmark program
Funny thing is on another youtube channel (who shall remain nameless ) there was a live q and a and you can post questions in the forum i asked about synthetic benchmarks and about doing more real world and he answered the questions before mine and after so skipped my question
I like using some synthetic benchmarks to check for stability. I doubt synthetic benchmarks will die out for this reason alone. I find the benchmarks that take a few seconds are helpful as well. It allows me to see what I have improved. The problem with using games for benchmarking is a lot can change with video driver updates. Just my opinon.
I don't think synthetic benchmarks will ever die. To this day I prefer them over game benchmarks for GPU's because you can never play the same thing twice in a game. You do the best you can do to keep it the same, but that's all you can do. Whereas synthetic benchmarks are the exact same tests every single time.
true but as long as that same performance difference in a synthetic benchmarks translates to the games i play though i do have to say some games are biased to different things like some games seem to work better with nvidia then amd and vice a versa
I am not sure how you would get the best of both worlds though games with built in benchmarking tools are pretty handy
But anywho, I think synthetics are also helpful when trying to compare apples to oranges. Example: friend A has a Intel/Nvidia build with Corsair SSD and RAM whereas friend B has an AMD rig with Kingston SSD and G.Skill Ram. Instead of individually benchmarking each different piece or just looking at how each specific game and/or application runs a quick look at synthetic scores can give you a ballpark figure as to how each friends rig is going to perform. Remember though, always take synthetics with a bag of salt.
synthetics won't die out mainly because people love to compare hardware. And synthetic benchmarks are the only way to properly benchmark hardware without any other variables besides the actual hardware.
yeah all good points i have to say.Guess my problem is sites that seem to weigh to heavily towards synthetic and just have the token real world stats.I guess its about finding the balance
Did post on youtube a while back about this but have to say the stress testing is a pretty good idea for stability but also for testing wattage on how much a system draws when its stressed