Benchmarks with a CPU bottleneck are a good indication of the future potential of a CPU

>benchmarks with a CPU bottleneck are a good indication of the future potential of a CPU

The historic facts sure don't line up with that.

everyone whos not retarded knows that low res benchmarks are no better than synthetics

Guess people with 4 threads are going to enjoy their 26 fps next year as the trend continues..

Most people are retarded, tho.

>Most people are retarded, tho.
sadly, it's the reality of the world

Worse than that, at least synthetic benchmarks can still be an indicator of performance. Game engines are fucking shit at indicating general performance they're usually poorly threaded if they're even threaded at all plus the devs are dumb as fuck and usually write spaghetti. Also they might use a certain compiler that favors a certain brand.

low resolution game benchmarks haven't been used widely for over 6-7 years and now they suddenly showed up.

lowres benches are purely IPC and clockspeed benchmarks.

Thankfully there's actually applications that test a CPU properly like sysbench, spec2006 gcc subtests and apache so not all is lost.

Even at high rest like 1440p there isn't too much of a GPU bottleneck and you can see...

Ryzen still sucks ass

They evidently did something wrong in the 2017 benchmark. A 20% CPU increase out of nowhere after 3 years of slight improvements makes literally no fucking sense. Also what kind of retarded fanboy would wait 5 fucking years for a CPU to finally show results?

It's one factor among many. Another factor might be devs making greater use of more threads as time goes on.
That doesn't mean CPU bottlenecking isn't useful for determining performance. Only that it's not the whole picture.

CPU bottlenecks in low-res gaming continues to be an effective gauge of future gaming performance. Among other tests, including synthetics.

They're showing up now because "professional hardware reviewers" (youtubers) are saying it's a way to see the real performance of a CPU in gaming and give an indicator of how it'll bottleneck in the future.

Funny enough, synthetic non-gaming benchmarks are a better indication of the future potential performance of a CPU in games if the game is optimized for it.

Nice, you're retarded and missed the whole point of the image in the OP. It's showing how more threads/cores became more important than higher per-core performance over time.
Newer games are using more cores and SMT.

Hmm..
Guess everything depends on what games someone cherrypicks.

Yeah it's slightly confusing. What it shows is a change in suite of games over time.

In later 2017 they changed to newer games than they used in very early 2017.
The first 2017 is basically 2016.

>CPU bottlenecks in low-res gaming continues to be an effective gauge of future gaming performance. Among other tests, including synthetics.
Looks like an Intel official decided to come down here for damage control.

Both synthetic and this suck shit, to be honest.

>Also what kind of retarded fanboy would wait 5 fucking years for a CPU to finally show results?
That's the other side of the coin.
Yes, the 8350 is a better CPU for gaming today. When people are already beginning to upgrade out of it.
For gaming, the 8350 was a bad investment. You needed to wait until today for it to be a good investment. But now you're stuck with an old CPU, mobo, chipset, etc. You very likely wanted it to be a good chip for the first 4 years more than the last 2.

The video cited in the OP in no way diminishes the value of low-res gaming benches. He only claims that they can't account for better utilization of cores/threads in the future, which both true and a separate issue.

ohhhhhh nooooooo ryzen lost to the 6900k by 3 fps amd is finished

j/k neck yourself fgt

Benchmark at low res dont show stutter issue. at 4k res bad min can ruin your day

Cpu scale with GPU power not as people expected

Least it was five years with the same socket. AM4 will continue this. You may get a 4core, then upgrade to an 8 core 16t zen+ in a year or two. Intel on the other hand fucks you over every chance they get.

AMD is now preforming more then well enough to buy into, and to tell Intel to go fuck themselves.

No it's showing changes in how much games need more threads over more single core performance.

Same GPU at 2013-2014 performance stay the same. new gpu performance got better

If I bottleneck CPU at 4k gaming, I'll just start 8k gaming.

same gpu (titan) performance got better, gap closed

go back to Sup Forums

look at the yellow bar fagget

Fucking this. It's really weird after all these years.

You realize the best Ryzen CPU is losing to an Intel CPU that costs $160 less

Synthetic benches do not test gaming performance though

Literally the best way to test pure gaming CPU performance is to knock the resolution down and turn off all the details

Barely behind a GPU that costs over twice as much and better than all the cheaper ones, there.

Sorry about your blindness, user.

It's because AMD mentioned it probably. So reviewers are taking to looking at Steam Hardware Survey to decide that most single monitor gaming is done at 1080p. Second place is 1366 x 768.

So a bunch of laptops on Steam means 1080p gaming is really a good reason to run desktop CPU comparisons because it shows that Ryzen can only do 280 fps vs 350 fps on 1080p low detail with Titan X Pascal. Especially 1024MB and ~2048MB VRAM are actually the two most popular VRAM amounts on there.

Are you drunk? That made no sense.

I meant CPU, not GPU.

...

Still not fair results since SMT is only working in a single game on Windows10 while HT is working in the majority of 2016 and 2017 releases.
But considering the 1400X will be like $200, still a good start. Better buy over a 7400 or even 7600k when it comes out, I bet.

You realize the best Intel CPU is losing to an Intel CPU that costs $750 less? Retard.

Is this the designated shitting on Intel thread?