-
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.
AMD THREADRIPPER VS INTEL SKYLAKE X
- Thread starter faceman123
- Start date
Caporegime
- Joined
- 9 Nov 2009
- Posts
- 25,333
- Location
- Planet Earth
Looks like Alienware has announced 16 core Ryzen and 10 core Intel Skylake X Area 51 systems:
https://videocardz.com/70234/dell-announces-alienware-area-51-threadripper-edition
https://videocardz.com/70234/dell-announces-alienware-area-51-threadripper-edition
None of them advertise what is separately on the Chipset, Threadripper has 64 on the CPU, Skylake-X has 28 - 44.
For example: Your Ryzen has 24 on the CPU, plus 4 USB 3.1 also on the CPU.
Ryzen has an Integrated Southbridge, it doesn't actually need a traditional Chipset, however it does have one for extra expansion, the X370 Chipset has an additional 2X USB 3.1 Type C, 6X Type A, 6X 2.0, 2 to 6X SATA 3 or 2X SATA Express and 8X PCIe2 lanes.
Do you know this for a fact or are you just assuming this is a fact? It’s been stated in more than one E3 YouTube video that 48 CPU lanes are used for the slots. It does not matter to me as 48 is still more than I need anyway.
"The slots" you mean PCIe, for dedicated Graphics? 48 dedicated for Graphics, thats no different to Intel, not all the PCIe lanes on the Intel chips are dedicated to Graphics, its divided up between NVMe and other IO.
64 and in Intel's case 44 is the total.
Those aren't proper IOPS benchmarks just throughput. There are a number of sites that have done storage benchmarks these two I think have the more detailed break downs:
https://www.pcper.com/reviews/Proce...nd-Zen/Ryzen-Chipsets-and-Storage-Performance
http://www.tweaktown.com/articles/8073/amd-ryzen-ssd-storage-performance-preview/index3.html
And to quote from another site "And that’s it. As you can see the chipset does pretty good in RAID0 but when using it for mirroring/RAID1 it falls flat on it’s face."
There are areas the platform does very well at such as sequential read - but some areas that are important for day to day performance such as larger queue IOPS and certain small file operations that are some of the most used in daily tasks (booting the OS, loading games and applications, on the fly real time loading, etc.) the platform falls flat on its face - someone posted a screenshot from Samsung magician showing IOPS half what I get on my 4820K and I have one generation older version of the drive they were testing and that isn't an abnormal result.
This has quite a lot of implications for workstation/enterprise use with the bigger CPUs with lots of cores and the kind of stuff people might put thread ripper upto where IOPS performance can be a lot more of a consideration than more mainstream desktop users who probably won't even notice the difference unless they are running an appropriate Intel rig alongside the AMD one.
Just found this, if its correct then it uses 4 lanes for the PCH: https://www.reddit.com/r/Amd/comments/6ee5qw/thanks_to_threadrippers_64_pcielanes_new_systems/
Meh, ^^^ 4 out of 64 to drive a Chipset 
Oh come on Roff, there isn't anything at all in that about Ryzen having Poor IOPS, nothing at all really of any concern. In someways it did better than X99, not that it was anything meaningful.
There are slight variations from platform to platform, none of them did anything notably worse than the other.
Infact his conclusion.
Why say something is there when it isn't? why would you do that?
Oh come on Roff, there isn't anything at all in that about Ryzen having Poor IOPS, nothing at all really of any concern. In someways it did better than X99, not that it was anything meaningful.
There are slight variations from platform to platform, none of them did anything notably worse than the other.
Infact his conclusion.
Why say something is there when it isn't? why would you do that?
Last edited:
Maybe because the IOPS performance is OK?
For my work fast drives with good IOPS make a big difference as I write POS import software that processes massive amounts of small(<10MB) files and posts it into databases. I have not tested it on my AMD PC yet so I don’t know how it performs but VM’s running SQL Server run fast.
I believe that screenshot came from me. A user here blamed it for his performance in games. When I posted the screenshot showing similar speeds and the fact I have no problems gaming we determined the problem is elsewhere.
Boot times, load times etc are not a problem.
His problem may or may not be elsewhere but IOPS being significantly down isn't a good thing for overall system performance regardless.
If you look at the queue depth graph for IOPS the picture isn't a rosy one for higher end use and even for desktop use it is going to have implications for how fast stuff starts and there are many areas where the performance is 20-50% down on equivalent Intel CPUs - if it was the case for a new Intel CPU people would be screaming blue murder.
Those details are NOT OK if you understand the implications - like so many things people will gloss over it because they don't like what I'm saying and then down the road it will come to light proper - remember things like microstutter?
If you look at the queue depth graph for IOPS the picture isn't a rosy one for higher end use and even for desktop use it is going to have implications for how fast stuff starts and there are many areas where the performance is 20-50% down on equivalent Intel CPUs - if it was the case for a new Intel CPU people would be screaming blue murder.
Those details are NOT OK if you understand the implications - like so many things people will gloss over it because they don't like what I'm saying and then down the road it will come to light proper - remember things like microstutter?
LOL, That’s what I thought. I expected the PCH to use more but as you say, a lot has moved away from the chipset.Meh, ^^^ 4 out of 64 to drive a Chipset
Microstutter like the 7700k has yet has better IOPS performance? For the majority this isn't even an issue.
Understand the implications? as only you do, right? i understand it enough to know that in reality its makes little difference.
You're making a mountain out of a mole hill to force an agenda.
No I was referring to past things like multi GPU microstutter where I was voicing concerns about it while people were in denial months before the people like PCPer started to properly expose the problem and things were fixed.
You really do have a short memory - so many times we've been at this point before and every time so far without fail I've turned out to have a point in the long run. One would think you'd be a little more cautious of flat out writing off there being any potential implications by now.
Oh and before I get ahead of myself - we don't know how Threadripper will perform in this respect yet but if it holds out it will be a much bigger issue on the kind of uses that platform would be put to compared to desktop use where it is less than ideal but not a game breaker.
It is amusing after all this time you still flat out deny anything I say that you don't like to hear and eventually come around to it later.
It is amusing after all this time you still flat out deny anything I say that you don't like to hear and eventually come around to it later.
Listen to yourself.....