• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Intel to launch 6 core Coffee Lake-S CPUs & Z370 chipset 5 October 2017

Lots.

1.475vcore , board has a little bit of vdroop so its sometimes 1.462-1.475. Been running for years no issues.

Could u list your bios settings please bud? Or just take screenshots with ya phone? I've never really pushed mine and wouldn't mind trying as I'm gonna upgrade soon anyway..

Would appreciate cheers bud
 
I don't know why people stuck with x264, not x265/H.265/HEVC on either CPU or GPU for superior quality stream. I did endless hours, days, weeks of tests with all methods and settings found Nvidia NVENC H.265 has same image quality as CPU H.265 on slower setting at same bitrate. Encoded Star Wars: The Force Awaken Blu-ray on NVENC H.265 at 6000kps bitrate took about 16 mins but encode on CPU H.265 never get done because it estimated to complete in 10 hours then it increased to 15 hours, 24 hours etc so I dont bothered encode it on CPU as it take forever no matter how many cores. I tried encoded 117MB Star Wars: The Last Jedi MPEG4 trailer on NVENC H.265 took amazing 6 secs to compacted 22MB at 1500kbps bitrate, CPU H.265 encoded took about 16 mins both GPU and CPU encoded had same image quality of original trailer which used higher 7308kbps bitrate. I read lots of interesting things about next generation AOMedia AV1 codec and tried encoded x264 and H.265 at 300-500kbps for fun but image quality was horrible blurred with artifacts. Encoded it with AV1 at 300kbps would see very clean image quality at tiny 7MB file size that will put x264, x265 and HEVC codec to shame.

http://www.streamingmedia.com/Artic...ard-Joins-Alliance-for-Open-Media-117634.aspx
https://www.elecard.com/news/results-of-elecards-latest-benchmarks-of-av1-compared-to-hevc
https://www.elecard.com/videos

Twitch 6000kbps bitrate on x264 is tiny compare to Crowd 1080p AV1 stream at 93962kbps around the quality of 4K Ultra HD Blu-ray bitrate.

Been watched AV1 test samples on Elecard Player blew me away with amazing image quality at very low and high bitrates, it is the only player at the moment capable to decode AV1 playback.

Compare Sintel trailer encoded in AV1, H.265 and H.264 at 300kbps bitrate:


Cant wait to use AV1 when software and browsers will have it enabled around Jan 2018 when AOMedia freeze AV1 codec experiment and release final version by 31 Dec 2017. :D

First off, you cant use NVENC with an AMD GPU.. and secondly, were talking about streaming here, casual streaming.. Twitch are never gona allow 93962kbps bitrate for streaming lol

For recording, well thats another topic for another thread :)

Lets see how the benchmarks pan out anyway, maybe Quick Sync has a better quality on the iGPU than before, but then it probably wont :)
 
It's going to be interesting to see some streaming benchmarks with X264 and how the 8700K compares to R7s.
X264 scales well with higher thread counts, but it also makes some use of AVX2, which is somewhat gimped on the Zen cores (2x 128bit FMAs for Zen vs 2x 256bit FMAs for Coffee Lake). Technically the Ryzens should still be more suited for streaming but I wonder if the 8700K's higher AVX throughput will help.
 
I don't have performance bias enabled and presumably neither do the other reviews which show it scoring higher.

Admittedly not a huge difference on the multi, the single core just seems lower than it should be.

Exactly, so a 1800X at stock (3.6GHz) scored 100 points less for the 200MHz all core deficit. Seems about right?

My 1800x will score 1850ish @4.1 without the bias enabled, that’s either 3466 or 3600 on the RAM (usually 3466 CL17).

Single thread is pretty poor on Ryzen though which is the major difference - I think it’s likely the 8700k will come close still in multi but far ahead single... which is why I’ll probably get one.
 
My 1800x will score 1850ish @4.1 without the bias enabled, that’s either 3466 or 3600 on the RAM (usually 3466 CL17).

Single thread is pretty poor on Ryzen though which is the major difference - I think it’s likely the 8700k will come close still in multi but far ahead single... which is why I’ll probably get one.

Seems like you have a very good sample there!
 
Most of the gaming pro boards offer Tbase 10Gb/s. Although aquantia arn't really proper 10Gb/s and It's strange they never used an Intel NIC.

It's shame Coffeemaker hasn't got ECC support.
 
What is the use case for having three Ethernet ports at the expense of more USB ports? I'm not even sure why a "gaming" motherboard needs two Ethernet ports...
 
What is the use case for having three Ethernet ports at the expense of more USB ports? I'm not even sure why a "gaming" motherboard needs two Ethernet ports...

One use for 10G Tbase is for building a server and integrating it into a current network without the need to build the network from scratch. Its a bit pointless building a server without ECC suport though and if your building a gaming system it's a not needed at all. TBH T base might actually introduce a little latency.
 
Seems like you have a very good sample there!

Just think the 1800x is much easier to work with as you would expect is the better silicon, for 24/7 use I wouldnt call 4.1 stable as a rock but runs 4.0 fine since day one. The memory speed is what made if increase from about 1750 to 1850... Not that i've really tried, maybe i'll try and break 1900 later ;)
 
Just think the 1800x is much easier to work with as you would expect is the better silicon, for 24/7 use I wouldnt call 4.1 stable as a rock but runs 4.0 fine since day one. The memory speed is what made if increase from about 1750 to 1850... Not that i've really tried, maybe i'll try and break 1900 later ;)

Yeah I was more impressed with the 3600, assuming you didn't set the trfc to something ridiculous.
 
Just think the 1800x is much easier to work with as you would expect is the better silicon, for 24/7 use I wouldnt call 4.1 stable as a rock but runs 4.0 fine since day one. The memory speed is what made if increase from about 1750 to 1850... Not that i've really tried, maybe i'll try and break 1900 later ;)

Yup same for my 1900x in Dual channel mode, so effectively like a Ryzen 7 memory speed is what gets you to 1850 ballpark. Like your self single core is not the strongest at 173 but it's not exactly slow :)

Hit a wall at 3600 also, tried faster RAM but couldn't get past it, actually best performance seems to come at 3466 with tight timings. So I'll be selling my faster DDR4 as it doesn't do a lot for TR probably great for this Intel chip though.
 
Yup same for my 1900x in Dual channel mode, so effectively like a Ryzen 7 memory speed is what gets you to 1850 ballpark. Like your self single core is not the strongest at 173 but it's not exactly slow :)

Hit a wall at 3600 also, tried faster RAM but couldn't get past it, actually best performance seems to come at 3466 with tight timings. So I'll be selling my faster DDR4 as it doesn't do a lot for TR probably great for this Intel chip though.

I haven't played around that much but I think to see gains past 3300Mhz you would need to increase Ryzen clock speed in a lot of situations. Mainly running games.
 
Anandtech published a menory scaling article on ryzen yesterday. It doesn't provide gains as big as the AMD slides would have you believe.
 
Back
Top Bottom