VehiclesFashionRecipesBlogsHuntTravelsSportFunHandmadeITEducation
Mini-Games
x

x
zakruti.com » IT - Software » PC World
Discussing Benchmarking Approaches With TechTechPotato

Discussing Benchmarking Approaches With TechTechPotato

FBTwitterReddit

video description

Rating: 4.0; Vote: 1
Gordon chats about hardware benchmarking practices with Dr. Ian Cutress of TechTechPotato and Anandtech fame. In this video they go over everything from how they choose benchmarks to whether or not they can trust company provided benchmarks. daverhodus: Benchmarking is overemphasized. A review should be more about the experience of using a product. Running a bunch of benchmarks doesn't always relate directly with a person using a product day-to-day. ie in a real system doing real things that people do.
Date: 2022-12-16

Comments and reviews: 6


I have to say it!!! It doesn't matter who you choose anymore.. (Benchmarks are for shite!!!) Yes they do give you an Idea of how a part performs but these days everything is so freaking fast that it doesn't really matter.. AMD and Intel are in a performance ty.. You would never be able to tell the difference between a new Intel PC or a New AMD PC in everyday use, same with Radeon and Nvidia.. In a blind test n If you didn't tell people what hardware they were using they would not be able to tell!! That's how close they all are in performance.. Look I have an ASUS 15 laptop with a 5500U and 16gigs of 3200Mhz and I have high end desktop 5900X/6800XT, And there are times when the laptop opens apps and plays things quicker than my desktop.. Even though the Desktop is like 5 times more powerful.. Honestly I could care less about Benchmarks anymore.. AMD Intel and Nvidia all are putting out great products that are all on par with each other.. Finally everyone knows that if you are into 4K gaming on a pretty OLED, The CPU matters very little.. You can still pair a 4080 or a 7900XTX with a 10900K or a 5800X and get the same frames no matter if you are using a 10900K or 13900K, it doesn't matter.. ( I Thank God that we can stay a couple Gens behind in CPU's and still get the same frames at 4K with a High end GPU.. )
reply

Have to say I'm a fan of Prime95 as a benchmark, not as a stress test, not as an application. Some only see it as a way to burn power, but to me it heavily exercises FP64 performance (especially AVX family) and depending on the test sizes, you alter the working data size so can gain information about caches and memory performance too. The main downside is without some understanding on what it is doing, you can't relate it back to hardware.
I do have an interest in finding prime numbers so to me it is a real use case, but I'm not interested in the type Prime95 focuses on. The math library written for Prime95 is also used in other similar applications (LLR, PFGW).

reply

My biggest pet peeve is when certain reviewers use specific games that have been coded to leverage specific hardware without highlighting this important detail/ information. This can easily tilt the overall review towards one company over another. The new and uninformed viewers will believe this as a real overall representation of the performance. I wish most reviewers approached benchmarks with a more balanced approach while disclaiming when certain software was specifically developed to use ABC's company hardware. A real unbiased, balanced, disclosed approach.
reply

My issue is when Reviews deliberately sabotage BIOS settings and or tweaking a games settings to help or hurt one CPU or GPU over another. Example most benchmarks are run under Ultra and another under Basic all to manipulate results artificially.
reply

My rig is for gaming, and I like the single player games with the eye candy so I watch benchmarks for games like god of war, cyberpunk, and control
reply

I hate when reviewers have bottlenecks on a different part of their system. You CANT do a GPU review if you're in any way cpu limited and vice versa.
reply
Add a review, comment






Other channel videos