And there's the rub, test with CPU limited situations (graphics turned down) which is what FPS players like myself do and AMD is behind consistently. Again, I hope that changes with Zen2 and am stoked to see what AMD does.
It isn't all over the place, it is consistent. Pretty much all tests say the same, Intel wins in gaming: https://www.gamersnexus.net/guides/3423-intel-i7-7700k-revisit-benchmark-vs-9700k-2700-9900k
Of course if a test is more GPU bound Intels faster gaming CPU can't pull away. Don't use cherry picked benchmarks.
It isn't all over the place, it is consistent. Pretty much all tests say the same, Intel wins in gaming: https://www.gamersnexus.net/guides/3423-intel-i7-7700k-revisit-benchmark-vs-9700k-2700-9900k
Of course if a test is more GPU bound Intels faster gaming CPU can't pull away. Don't use cherry picked benchmarks.
> Doesn't watch video.
> Doesn't comment on video.
> Posts a link that reviews the processor on "flagship" games people hardly play
> Doesn't have an in game video showing core utilization and FPS variance in real-time side by side which highlights my point
> Shows static graphs that don't capture dynamic varying behavior
From your own link :
when there is a deviation, we want it to be fewer than an 8-12ms change frame-to-frame as that tends to be about the human perceptual range for frametime consistency.
All of the CPUs produce the the same general frame-time consistency which was my point .. What you can actually perceive. You're the limit in the pipeline not the GPU/CPU anymore.
1000/12ms = 83
1000/8ms = 125
The 2700x is within the perceptible frame limitations.
That's all that matters. A 9900k under LN2 cooling will produce even higher frame rates. Do I care? No because I can hardly perceive the difference. People pridefully state they don't care how or why something is the case... You posted a link on your own which highlights the consequence of this posture. You and your biological perception are the limit in the pipeline not the CPU/GPU anymore. I know some people like to believe they have super human capabilities and are Cyborgs but you're really not.
> mfw 200 FPS is insufficient
> mfw you spot the 4790k you have in your gaming machine w/ 4 cores from 2014 producing 200fps and someone is bragging about "performance"
> mfw it's clear none of these games properly scale across 8 cores
> What's the current max refresh rate on monitors?
> What's the frame 'hold time'?
> Pointless performance and accolades
I've had enough of this debate. It's pointless and ridiculous. If a comment isn't AMD related, I'm not responding anymore.
If you use some low latency 3200 c14 ram on the 2700x what fast paced games Can't keep 60 fps 99/99.9% of the time?So a long winded way of saying Intel is faster, but not enough that you would notice.
Still waiting for example where Ryzen actually wins, which was the previous claim.
Sad how AMD is barely trying with their mobile offerings, but I suppose the odds are stacked against them in that market. Raven Ridge suffered from poor battery life relative to Intel's offerings and other issues, but the opportunity was there to now capitalize on their node advantage and they didn't take it. Their 12nm APUs won't make too much of a difference and Renoir will be far too late.
They didn't pay attention or they ignore stuff like that intentionally if they expect a full announcement. We will get a launch window announcement, maybe. Even Adored says March to April for the actual launch...AMD will launch server first (EPYC 2) and then Desktop Ryzen 3, so why did anyone expect for AMD to announce Desktop Ryzen 3 in CES 2019 when they havent officially launched EPYC 2 ??
AMD will launch server first (EPYC 2) and then Desktop Ryzen 3, so why did anyone expect for AMD to announce Desktop Ryzen 3 in CES 2019 when they havent officially launched EPYC 2 ??
AMD will launch server first (EPYC 2) and then Desktop Ryzen 3, so why did anyone expect for AMD to announce Desktop Ryzen 3 in CES 2019 when they havent officially launched EPYC 2 ??
AMD will launch server first (EPYC 2) and then Desktop Ryzen 3, so why did anyone expect for AMD to announce Desktop Ryzen 3 in CES 2019 when they havent officially launched EPYC 2 ??
thisYou pre-announce products, to stifle sales of competing product already on the market (9900K).
In effect they would be saying, don't buy that 9900K yet, wait and see what Ryzen 3 has for you...
(and for GPUs, don't buy RTX, when our 7nm GPUs will be coming soon...)
Only for the mobile chips: https://www.amd.com/en/press-releas...plete-mobile-portfolio-new-ryzen-athlon-and-aNo official AMD press release yet?
Not Ryzen 3000, but Epyc threads haven't been bumped in a while, so whatever. First Rome QS spotted. 1.4GHz base, 2.2GHz boost...although it apparently ran at 0.90GHz in this test.
You pre-announce products, to stifle sales of competing product already on the market (9900K).
In effect they would be saying, don't buy that 9900K yet, wait and see what Ryzen 3 has for you...
(and for GPUs, don't buy RTX, when our 7nm GPUs will be coming soon...)
Not Ryzen 3000, but Epyc threads haven't been bumped in a while, so whatever. First Rome QS spotted. 1.4GHz base, 2.2GHz boost...although it apparently ran at 0.90GHz in this test.
I think it probably wasn't running at 0.9GHz. If I recall correctly, SiSoft Sandra isn't very accurate with reported clockspeeds.At 900MHz, and on a core/core basis, this would be about 17% faster than a 7980XE and 20% faster than a dual Xeon 8180.
The multimedia test use AVX2, so it s an apple/apple comparison.
If this is a real thing every chiplet has two 4-core CCXs where each CCX will have 16MB L3.
L3 is shared across the two CCXs of a Zeppelin die.
You are obviously right, dunno what I was thinking (probably cores instead of complexes)Wrong, it's per CCX.
L3 is shared across the two CCXs of a Zeppelin die.