[Adoredtv] Ryzen 3000, Radeon 3000 Series LEAKS - It's Game On! - Page 100 - Overclock.net - An Overclocking Community

Forum Jump: 

[Adoredtv] Ryzen 3000, Radeon 3000 Series LEAKS - It's Game On!

Closed Thread
 
Thread Tools
post #991 of 1013 (permalink) Old 03-14-2019, 03:16 PM
I <3 narcissists
 
bigjdubb's Avatar
 
Join Date: Feb 2008
Location: Houston, TX
Posts: 5,594
Rep: 219 (Unique: 136)
Quote: Originally Posted by technodanvan View Post
Frankly, I have yet to see an impressive TR4 board either.
I really haven't looked at them in depth, there are certainly some expensive TR4 boards.

There's not really anything wrong with the AM4 boards but they got no pizazz. I wanted asus to make an x470 extreme with one of those dim.2 slots/riser cards, the little waste of an lcd screen on the io cover and a honkin' power delivery system. What we get instead is the CHVII that costs more than it's peers but doesn't actually offer anything more, not even any added fluff on the darn thing.

Gaming Rig
(12 items)
Couch Gaming Rig
(10 items)
Work rig
(11 items)
CPU
Ryzen 7 3700x
Motherboard
Gigabyte X570 Master
GPU
EVGA RTX 2080ti FTW3
RAM
G.Skill TridentZ RGB
Hard Drive
HP EX920 m.2
Hard Drive
Intel SSD6 m.2
Hard Drive
Intel SSD6 m.2
Power Supply
EVGA G3
Cooling
CORSAIR H150I PRO
Case
LianLi PC-O11 Dynamic
Operating System
Win 10 Home
Monitor
LG 32GK650G
CPU
Ryzen 5 3600
Motherboard
Asus x470i
GPU
Radeon VII
RAM
G.Skill TridentZ RGB
Hard Drive
Samsung 970 evo
Power Supply
InWin A1
Cooling
NZXT M22
Case
InWin A1
Operating System
Win10 Home
Monitor
Vizio 4k TV
CPU
AMD Ryzen 7 2700X
Motherboard
Asrock X470 Taichi Ultimate
GPU
GTX 750
RAM
Patriot Viper
Hard Drive
HP EX920 M.2
Power Supply
EVGA G3
Cooling
CORSAIR H150I PRO
Case
NZXT Source 210
Operating System
Windows 10 Pro
Monitor
BenQ PD3200Q
Monitor
LG 32UD59
▲ hide details ▲


bigjdubb is offline  
Sponsored Links
Advertisement
 
post #992 of 1013 (permalink) Old 03-14-2019, 05:03 PM
Indentified! On the Way!!
 
LancerVI's Avatar
 
Join Date: May 2012
Posts: 2,602
Rep: 153 (Unique: 114)
Quote: Originally Posted by bigjdubb View Post
I really haven't looked at them in depth, there are certainly some expensive TR4 boards.

There's not really anything wrong with the AM4 boards but they got no pizazz. I wanted asus to make an x470 extreme with one of those dim.2 slots/riser cards, the little waste of an lcd screen on the io cover and a honkin' power delivery system. What we get instead is the CHVII that costs more than it's peers but doesn't actually offer anything more, not even any added fluff on the darn thing.
Completely agree here. My Asus Prime X470 Pro is a good board, but it isn't a great board. I want real, high end option for AMD boards. Hopefully the mobo makers will come around. Ryzen is an unmitigated success. Time for full on support with full product trees.
LancerVI is offline  
post #993 of 1013 (permalink) Old 03-14-2019, 05:44 PM
Not a linux lobbyist
 
rluker5's Avatar
 
Join Date: Feb 2014
Location: Wisconsin
Posts: 1,788
Rep: 45 (Unique: 36)
Quote: Originally Posted by ajc9988 View Post
Your distinction between theoretical and practical is meaningless here. You are mistaking synthetic for theory. Synthetic is still practical IPC, theory is the mathematical computation of what is likely going to be the highest achievable IPC on paper. There is a difference.

Now, to address the synthetic comment, I addressed that by saying directly that IPC changes by task, task meaning program. What that means is that performance varies by program, which is already seen from game play benchmarks and variations due to the coding and compiling of the program. Easiest example of this is looking at Intel's compiler which purposely gimped AMD CPUs and told it to run a slower instruction set and was not optimized for AMD CPUs, which Intel got slapped with an antitrust violation for doing. That means specific program level optimizations for CPUs can and do make a difference. This is also why AMD has sued over some benchmark companies stacking the deck with tasks that don't necessarily mirror real life workloads and stacked the deck with tasks that Intel CPUs do well in, while reducing tasks that AMD does well in. This is different than the compiling optimizations and selection of instruction sets, etc.

That means that IPC varies program to program and means that seeing the performance on the specific program you use is the only way to get a deeper understanding of what the difference in the hardware is. That is why some reviewers measure it by game, bottleneck the CPU by using the most powerful GPU and using a lower frame rate, although below 1080p is NOT realistic in any sense of the word. This is why the settled practices are already teased out, which it was Intel that suggested to reviewers using 720p, which only a couple places actually used, one being PCPer, whose Ryan Shrout now works for Intel in a marketing position where he designs tests to over-inflate the performance of Intel CPUs for a living. Hmmm....

As to talking latency again, maybe you should go look up the rumored reduction in latency. You like to speculate on the increase of going off core die, surely you trying to find that value would solve your question whether AMD has a silver bullet or not. Same with an examination of trying to equalize the latency from the I/O die to the core chiplets on TR and Epyc (likely helps with the stale data potential).

As to IPC being too general, how? No one fully will use peak memory throughput or memory latency to determine the optimal settings for their use. They will either use performance on their specific software suite while tweaking settings or a synthetic benchmark which is similar to their use for setting up the system. How much can you change the values on each of the settings? Cache, can't change that. Pipeline and bandwidth on cache, generally no. So, what do you do? You overclock the ram to find stable, then measure INCREASE IN IPC to determine whether those settings work for your needs. So how is that too general?


Also, as to the impact of ram, until Zen, whose IF relies on Memory speed, the impact of ram on performance on Intel's side was generally 2%, with an outside of 5%, increase in performance. AMD, due to the clocks effecting IF speed, bandwidth, and latency, you could see a wider margin, like 10%, increase in performance. As you increased ram speed, so long as ram latency was equal to the slower ram speed or lower latency than the slower ram speed, you were receiving a knock on effect where the IF was then clocked higher, which lowered the IF latency. Teasing out how much of the performance uplift is attributed to each is an impossible task.

Because of that, it isn't getting lost or focusing on a lot of trees, it is explaining the role the various trees have in the forest, which explains how the forest is created and functions. It is attempting to show the interconnected nature and how differences in architecture would change the interactions between the parts.

Meanwhile, it's only a couple months between now and release, which is the true empirical analysis. So that will come soon enough.

So, with a 5ghz core clock and an increased ram latency of about 25%, do you think Zen2 anything will consistently beat the 7700k (stock) in these games in clearly cpu limited scenarios?: https://www.anandtech.com/show/13400...9600k-review/9
And how do you explain ipc when the cores don't scale?
I chose Anand because it was the first thing to come to mind, and the 9900k because the focus was neither on Ryzen or Kaby Lake.
And I'm comparing with gaming because 1. that is the most important reason for most people who put together computers for a cpu to be fast, 2. that is the one big thing people bring up about Zen not being as good as core with, and 3. it is ram latency related.

Ryzen wins in compute based productivity because of the number of cores that are good at it and Zen2 should be faster and have more cores. But that means little to an average guy like me who has no reason to upgrade any quad core cpu since Z97 other than video games.

But maybe I'm worrying over nothing. I hope Zen 2 is better. Intel doesn't seem to be coming out with much better anytime soon.

L5
(17 items)
Lea2
(11 items)
L7
(11 items)
CPU
5775c
Motherboard
Maximus VII Hero
GPU
Aorus 1080ti Waterforce
RAM
16 Gb Gskill Trident @ 2400,cas10,1.575v
RAM
8 Gb Gskill Trident @ 2400,cas10,1.575v
Hard Drive
1Tb Team ssd
Hard Drive
seagate barracuda 3T
Hard Drive
Optane 900p 480G OS
Optical Drive
Asus BW-16D1HT
Power Supply
EVGA Supernova 1300 G2
Cooling
Cooler Master MasterLiquid Pro 120 (cpu)
Cooling
2 140mm case fans, 2 120mm
Case
Fractal Design R4 (no window)
Operating System
W10 64 pro
Monitor
panasonic TC-58AX800U
Audio
Focal Elear, Nova 40, 598se, HE4xx, DT990pro w b.boost earpads
Audio
SoundbasterX AE-5, onboard
CPU
4770k
Motherboard
Asus Z87 Deluxe
GPU
Fury Nitro
RAM
8Gb klevv urbane 2133
Hard Drive
ROG Raidr 240Gb pcie
Hard Drive
1Tb WD blue
Power Supply
Pc Power&Cooling silencer Mk2 950w
Cooling
Deepcool Lucifer V2
Case
DIYPC P48-W
Operating System
W10 64 pro
Monitor
40"tv
CPU
4980hq
Motherboard
Asus H81T/CSM
RAM
8Gb 1600 samsung
Hard Drive
Samsung 850 evo 120gb
Power Supply
Skyvast 90w brick for hp pavilion something
Cooling
SilverStone Tek Super Slim
Case
SilverStone Tek PT13B
Operating System
W10 64 pro
Monitor
24" samsung 1080p
Keyboard
Logitech K400+
Other
Intel wifi ac card and noname antennas
▲ hide details ▲
rluker5 is offline  
Sponsored Links
Advertisement
 
post #994 of 1013 (permalink) Old 03-14-2019, 06:31 PM
New to Overclock.net
 
Grin's Avatar
 
Join Date: May 2017
Location: USA
Posts: 164
Rep: 5 (Unique: 5)
Do not mix an overall cpu performance with IPC. IPC calculations are based on the number of machine-level instructions required to complete a piece of code and the number of clock cycles required to complete it on the actual cpu. Processors with similar IPCs may have a different performance in different scenarios.

Last edited by Grin; 03-14-2019 at 06:39 PM.
Grin is offline  
post #995 of 1013 (permalink) Old 03-14-2019, 07:29 PM
New to Overclock.net
 
ajc9988's Avatar
 
Join Date: Mar 2015
Posts: 488
Rep: 33 (Unique: 26)
Quote: Originally Posted by rluker5 View Post
So, with a 5ghz core clock and an increased ram latency of about 25%, do you think Zen2 anything will consistently beat the 7700k (stock) in these games in clearly cpu limited scenarios?: https://www.anandtech.com/show/13400...9600k-review/9
And how do you explain ipc when the cores don't scale?
I chose Anand because it was the first thing to come to mind, and the 9900k because the focus was neither on Ryzen or Kaby Lake.
And I'm comparing with gaming because 1. that is the most important reason for most people who put together computers for a cpu to be fast, 2. that is the one big thing people bring up about Zen not being as good as core with, and 3. it is ram latency related.

Ryzen wins in compute based productivity because of the number of cores that are good at it and Zen2 should be faster and have more cores. But that means little to an average guy like me who has no reason to upgrade any quad core cpu since Z97 other than video games.

But maybe I'm worrying over nothing. I hope Zen 2 is better. Intel doesn't seem to be coming out with much better anytime soon.
For measuring IPC, I did mention that frequency needs to be constant between the tested hardware and that it varied by program. Looking at tests of processors with different frequencies and IPC is not doing an IPC comparison, but looking at overall.

Then, where are you getting an increase of 25%? From your estimate of what you think latency will be going to the I/O die, while ignoring the lower latency of IF, changes to the IMC, etc.? You are building that statement on assumptions to which you possess NO INFORMATION. Let's say that IF lowered the latency by 15ns and the IMC changes allowed to lower that latency as well. You'd wind up with zero change. I'm pulling numbers out of my butt here because NO ONE KNOWS, unless you are a company insider. That is my point, you are making bad assumptions without ground. Sure, it is reasonable to estimate there will be some latency going off core to the I/O die versus staying on die and straight out to memory, but you have to compare the latencies to the prior X950x processors, as that is the only comparable product with two chiplets. Reason I say that is because half the time on a 1950x or 2950x, you would have to first use IF to go to the other die, then out to memory, then return trip. With Zen 2, you are changing that so the memory controllers accessed by the CPUs is centrally located. That means that although the latency is higher for the first die where the IMC is no longer a direct link, it is WAY shorter than having to jump to the IMC on the second die, then go out, instead having both chips go to the I/O and then out on IF2 with double the bandwidth on the IF2 and much lower latency than previously, with IF2 connecting the core dies directly (something not present on EPYC, and may or may not be present on TR chips).

Next, you assume memory is the issue, but if you compare the score of the 2700X to the 9900K in World of Tanks (which this is before the AMD optimization went in place for WoT in the past month or two, so those numbers are WORTHLESS as a practical matter because AMD CPUs received a significant uplift in performance with that update, but I digress), you have a 30% delta (which is now much lower, and of which, this is an outlier on game performance comparing frames out of the entire selection of games benchmarked, which ranges from 1-2% performance difference up to nearly 20%, with many settling into single digit and low double digit (think low teens) for the most common distribution, once again getting back to my statement earlier in posts showing 11% overall performance deficit in gaming and 22% at overall productivity when comparing a 2700X and a 9900K. This means even though Intel has lower memory latency and higher frequency AND higher IPC (which memory latency is compounded into in regards to its effects), the delta for gaming is now easier to close than the productivity suites designed for better use of multithread workloads. It also shows the variance by program on tasks for the CPU.

So you are using wild aspersions on latency and the impact thereof, using outdated, cherry picked data to try to prove a point (as bad as anecdotal evidence), misunderstanding the meaning of IPC and interplay with frequency and other factors to give overall performance, using overall performance instead of an IPC calculation, etc. In other words, you don't control for anything and are using improper data to try to prove a point that the data does not support. Intel's frequency alone is over 10% of AMD's fastest, meaning having that same relative speed increase on a Zen or Zen+ design should allow, mostly, for the 11% average difference to disappear, if one were being rational. When you multiply the projected percent frequency increase by the projected IPC increase over first gen or second gen, you get an estimate of the overall uplift in performance, which should be 25%+ (not 25% IPC, using 13% multiplied by a 10% frequency increase gives you 24% approximate uplift in performance).

What are you trying to accomplish? Are you truly trying to understand this better, or trying to prove me wrong? That last paragraph suggests the latter, but I could be wrong... If truly trying to wrap your mind around this, let's continue.

Quote: Originally Posted by Grin View Post
Do not mix an overall cpu performance with IPC. IPC calculations are based on the number of machine-level instructions required to complete a piece of code and the number of clock cycles required to complete it on the actual cpu. Processors with similar IPCs may have a different performance in different scenarios.
Thank you for the direct statement here. Although I did explain this a couple posts back, I really like this straightforward statement's wording which is very accurate and poignant.

Desktop (Click to show)
Asus Maximus VIII Extreme, G.Skill 4133MHz - Running 3733MHz (MB limitation; I've had 4000, but...)@[email protected] 14-17-17-36-280 with Corsair CMXAF2 Ram fans, EVGA 980 Ti Classified (OC specs to come), [email protected]@1.42V everyday stable, Thermaltake Water 3.0 Ultimate AIO (push/pull 6xNoctua iPPC 3000), EVGA 1600 Ti PSU (for later SLI setups), Thermaltake X9 Cube, Intel 600P NVME
Laptop (Click to show)
P770ZM / Chi Mei 1080p 17.3" TN matte - OC to 100Hz / Intel Core i7 4790K - 4.5GHz 1.130V Static / Nvidia Geforce 980M 8GB (current OC limit of 330MHZ Core/410MHZ VRam/ [email protected]) / Kingston HyperX 32GB DDR3 [email protected] CR-2T custom SPD / Samsung 850 EVO 500GB M.2 / Intel Wifi Dual Band 7265ac+BT / Win 7 Pro x64 - custom by NTLite / 330W PSU/Prema BIOS
ajc9988 is offline  
post #996 of 1013 (permalink) Old 03-14-2019, 07:33 PM
New to Posting
 
nonametoclaim's Avatar
 
Join Date: Jul 2017
Posts: 172
Rep: 6 (Unique: 6)
didnt feel like searching to see if this was already posted or not(pretty sure i have the flu)

but some more leaks have surfaced, this one shows march release which would be awesome but im sure will be debunked

https://www.guru3d.com/news-story/ov...ed-online.html

CPU
Ryzen 5 2600
Motherboard
MSI AM4 B450I Gaming Plus AC
GPU
MSI Radeon RX 580 8GB
RAM
Patriot Viper 4 2x8gb 3000mhz
Hard Drive
500gb Mushkin Source M.2
Hard Drive
1tb Mushkin Source SSD
Power Supply
Corsair CX 650
Power Supply
2016 International Prostar 2200w Inverter
Cooling
Coolermaster LC120E AIO
Case
Silverstone Sugo13
Operating System
Microsoft Windows 10 Home
Monitor
Insignia 43" 4K HDR
Monitor
Samsung Odyssey+
▲ hide details ▲
nonametoclaim is offline  
post #997 of 1013 (permalink) Old 03-14-2019, 08:00 PM
New to Overclock.net
 
ajc9988's Avatar
 
Join Date: Mar 2015
Posts: 488
Rep: 33 (Unique: 26)
Quote: Originally Posted by nonametoclaim View Post
didnt feel like searching to see if this was already posted or not(pretty sure i have the flu)

but some more leaks have surfaced, this one shows march release which would be awesome but im sure will be debunked

https://www.guru3d.com/news-story/ov...ed-online.html
Yeah, there won't be a March release. At GDC on March 20th, AMD is presenting more information on Zen 2 and how to code for it, as well as potential info on Navi coming to the GPU optimization presentation later in the day. As such, I wouldn't get too excited yet. Just wait a week to find out more about Zen 2.

Desktop (Click to show)
Asus Maximus VIII Extreme, G.Skill 4133MHz - Running 3733MHz (MB limitation; I've had 4000, but...)@[email protected] 14-17-17-36-280 with Corsair CMXAF2 Ram fans, EVGA 980 Ti Classified (OC specs to come), [email protected]@1.42V everyday stable, Thermaltake Water 3.0 Ultimate AIO (push/pull 6xNoctua iPPC 3000), EVGA 1600 Ti PSU (for later SLI setups), Thermaltake X9 Cube, Intel 600P NVME
Laptop (Click to show)
P770ZM / Chi Mei 1080p 17.3" TN matte - OC to 100Hz / Intel Core i7 4790K - 4.5GHz 1.130V Static / Nvidia Geforce 980M 8GB (current OC limit of 330MHZ Core/410MHZ VRam/ [email protected]) / Kingston HyperX 32GB DDR3 [email protected] CR-2T custom SPD / Samsung 850 EVO 500GB M.2 / Intel Wifi Dual Band 7265ac+BT / Win 7 Pro x64 - custom by NTLite / 330W PSU/Prema BIOS
ajc9988 is offline  
post #998 of 1013 (permalink) Old 03-14-2019, 08:17 PM
New to Posting
 
nonametoclaim's Avatar
 
Join Date: Jul 2017
Posts: 172
Rep: 6 (Unique: 6)
Quote: Originally Posted by ajc9988 View Post
Yeah, there won't be a March release. At GDC on March 20th, AMD is presenting more information on Zen 2 and how to code for it, as well as potential info on Navi coming to the GPU optimization presentation later in the day. As such, I wouldn't get too excited yet. Just wait a week to find out more about Zen 2.
honestly im in no need of an upgrade, my current $800(at build) rig is over twice the performance of my rig from 2012 that cost me a fortune to build. but i absolutely love watching AMD casually toss out bang for the buck CPUs that creep on intel's money milkers and watch team blue yank stuff off the production line and toss it into the market just to claim top performer.

but a 5ghz 12/24 c/t does sound quite nice

just hoping navi can bring in something for my SG13 that rivals a 1080 ti

CPU
Ryzen 5 2600
Motherboard
MSI AM4 B450I Gaming Plus AC
GPU
MSI Radeon RX 580 8GB
RAM
Patriot Viper 4 2x8gb 3000mhz
Hard Drive
500gb Mushkin Source M.2
Hard Drive
1tb Mushkin Source SSD
Power Supply
Corsair CX 650
Power Supply
2016 International Prostar 2200w Inverter
Cooling
Coolermaster LC120E AIO
Case
Silverstone Sugo13
Operating System
Microsoft Windows 10 Home
Monitor
Insignia 43" 4K HDR
Monitor
Samsung Odyssey+
▲ hide details ▲
nonametoclaim is offline  
post #999 of 1013 (permalink) Old 03-14-2019, 08:34 PM
Not a linux lobbyist
 
rluker5's Avatar
 
Join Date: Feb 2014
Location: Wisconsin
Posts: 1,788
Rep: 45 (Unique: 36)
Quote: Originally Posted by ajc9988 View Post
For measuring IPC, I did mention that frequency needs to be constant between the tested hardware and that it varied by program. Looking at tests of processors with different frequencies and IPC is not doing an IPC comparison, but looking at overall.

Then, where are you getting an increase of 25%? From your estimate of what you think latency will be going to the I/O die, while ignoring the lower latency of IF, changes to the IMC, etc.? You are building that statement on assumptions to which you possess NO INFORMATION. Let's say that IF lowered the latency by 15ns and the IMC changes allowed to lower that latency as well. You'd wind up with zero change. I'm pulling numbers out of my butt here because NO ONE KNOWS, unless you are a company insider. That is my point, you are making bad assumptions without ground. Sure, it is reasonable to estimate there will be some latency going off core to the I/O die versus staying on die and straight out to memory, but you have to compare the latencies to the prior X950x processors, as that is the only comparable product with two chiplets. Reason I say that is because half the time on a 1950x or 2950x, you would have to first use IF to go to the other die, then out to memory, then return trip. With Zen 2, you are changing that so the memory controllers accessed by the CPUs is centrally located. That means that although the latency is higher for the first die where the IMC is no longer a direct link, it is WAY shorter than having to jump to the IMC on the second die, then go out, instead having both chips go to the I/O and then out on IF2 with double the bandwidth on the IF2 and much lower latency than previously, with IF2 connecting the core dies directly (something not present on EPYC, and may or may not be present on TR chips).

Next, you assume memory is the issue, but if you compare the score of the 2700X to the 9900K in World of Tanks (which this is before the AMD optimization went in place for WoT in the past month or two, so those numbers are WORTHLESS as a practical matter because AMD CPUs received a significant uplift in performance with that update, but I digress), you have a 30% delta (which is now much lower, and of which, this is an outlier on game performance comparing frames out of the entire selection of games benchmarked, which ranges from 1-2% performance difference up to nearly 20%, with many settling into single digit and low double digit (think low teens) for the most common distribution, once again getting back to my statement earlier in posts showing 11% overall performance deficit in gaming and 22% at overall productivity when comparing a 2700X and a 9900K. This means even though Intel has lower memory latency and higher frequency AND higher IPC (which memory latency is compounded into in regards to its effects), the delta for gaming is now easier to close than the productivity suites designed for better use of multithread workloads. It also shows the variance by program on tasks for the CPU.

So you are using wild aspersions on latency and the impact thereof, using outdated, cherry picked data to try to prove a point (as bad as anecdotal evidence), misunderstanding the meaning of IPC and interplay with frequency and other factors to give overall performance, using overall performance instead of an IPC calculation, etc. In other words, you don't control for anything and are using improper data to try to prove a point that the data does not support. Intel's frequency alone is over 10% of AMD's fastest, meaning having that same relative speed increase on a Zen or Zen+ design should allow, mostly, for the 11% average difference to disappear, if one were being rational. When you multiply the projected percent frequency increase by the projected IPC increase over first gen or second gen, you get an estimate of the overall uplift in performance, which should be 25%+ (not 25% IPC, using 13% multiplied by a 10% frequency increase gives you 24% approximate uplift in performance).

What are you trying to accomplish? Are you truly trying to understand this better, or trying to prove me wrong? That last paragraph suggests the latter, but I could be wrong... If truly trying to wrap your mind around this, let's continue.


Thank you for the direct statement here. Although I did explain this a couple posts back, I really like this straightforward statement's wording which is very accurate and poignant.
That 25% was from the ram latency on a leaked Zen2 userbenchmark where the latency was 15 or 20 ns higher than the crappy ram Hynix spec, where my laptop with kaby lake and very similar crappy Hynix runs 5ns less. 15ns/60ns = .25, 20/80=.25 . A very rough approximation of a plausible range, but all I have heard so far.

I didn't cherry pick, Anand was first pick. It was a guess. And I was going by the average of averages and mins of all of the games at 720p. Guru3d was 2nd and Ryzen looked worse for gaming. Maybe you have a better cpu limited (not 4k ultra or 1080p ultra with a 470 so there is a spread in performance per cpu larger than random testing noise) source that isn't reviewing either of the compared cpus.

And I'm trying to understand better because I'm hoping for an upgrade. But I want my use case - gaming, web browsing, streaming, shopping, light office, etc to justify it. That's all. I plan on getting a single gpu of the next Nvidia 80 series when it comes out as well if it is a bit over 2080ti perf at about 200w and $800. Still on Z97 era stuff and would like if something came out on the cpu side that could overpower 4k60 without running over 80c on an aio, or a big air cooler. I have a concern over ram latency because I've seen it's effects, but you just dismiss them, even though the main reason the 2nd Ryzen series was faster than the first was latency improvement. My stuff already games like a 7700k and I don't feel like spending 1k on a new mobo, cpu, ram for something over a sidegrade that will be all hot and noisy unless I go for watercooling.
Not getting any new info so I guess I can wait like everyone else.

L5
(17 items)
Lea2
(11 items)
L7
(11 items)
CPU
5775c
Motherboard
Maximus VII Hero
GPU
Aorus 1080ti Waterforce
RAM
16 Gb Gskill Trident @ 2400,cas10,1.575v
RAM
8 Gb Gskill Trident @ 2400,cas10,1.575v
Hard Drive
1Tb Team ssd
Hard Drive
seagate barracuda 3T
Hard Drive
Optane 900p 480G OS
Optical Drive
Asus BW-16D1HT
Power Supply
EVGA Supernova 1300 G2
Cooling
Cooler Master MasterLiquid Pro 120 (cpu)
Cooling
2 140mm case fans, 2 120mm
Case
Fractal Design R4 (no window)
Operating System
W10 64 pro
Monitor
panasonic TC-58AX800U
Audio
Focal Elear, Nova 40, 598se, HE4xx, DT990pro w b.boost earpads
Audio
SoundbasterX AE-5, onboard
CPU
4770k
Motherboard
Asus Z87 Deluxe
GPU
Fury Nitro
RAM
8Gb klevv urbane 2133
Hard Drive
ROG Raidr 240Gb pcie
Hard Drive
1Tb WD blue
Power Supply
Pc Power&Cooling silencer Mk2 950w
Cooling
Deepcool Lucifer V2
Case
DIYPC P48-W
Operating System
W10 64 pro
Monitor
40"tv
CPU
4980hq
Motherboard
Asus H81T/CSM
RAM
8Gb 1600 samsung
Hard Drive
Samsung 850 evo 120gb
Power Supply
Skyvast 90w brick for hp pavilion something
Cooling
SilverStone Tek Super Slim
Case
SilverStone Tek PT13B
Operating System
W10 64 pro
Monitor
24" samsung 1080p
Keyboard
Logitech K400+
Other
Intel wifi ac card and noname antennas
▲ hide details ▲
rluker5 is offline  
post #1000 of 1013 (permalink) Old 03-14-2019, 10:34 PM
New to Overclock.net
 
ajc9988's Avatar
 
Join Date: Mar 2015
Posts: 488
Rep: 33 (Unique: 26)
Quote: Originally Posted by rluker5 View Post
That 25% was from the ram latency on a leaked Zen2 userbenchmark where the latency was 15 or 20 ns higher than the crappy ram Hynix spec, where my laptop with kaby lake and very similar crappy Hynix runs 5ns less. 15ns/60ns = .25, 20/80=.25 . A very rough approximation of a plausible range, but all I have heard so far.

I didn't cherry pick, Anand was first pick. It was a guess. And I was going by the average of averages and mins of all of the games at 720p. Guru3d was 2nd and Ryzen looked worse for gaming. Maybe you have a better cpu limited (not 4k ultra or 1080p ultra with a 470 so there is a spread in performance per cpu larger than random testing noise) source that isn't reviewing either of the compared cpus.

And I'm trying to understand better because I'm hoping for an upgrade. But I want my use case - gaming, web browsing, streaming, shopping, light office, etc to justify it. That's all. I plan on getting a single gpu of the next Nvidia 80 series when it comes out as well if it is a bit over 2080ti perf at about 200w and $800. Still on Z97 era stuff and would like if something came out on the cpu side that could overpower 4k60 without running over 80c on an aio, or a big air cooler. I have a concern over ram latency because I've seen it's effects, but you just dismiss them, even though the main reason the 2nd Ryzen series was faster than the first was latency improvement. My stuff already games like a 7700k and I don't feel like spending 1k on a new mobo, cpu, ram for something over a sidegrade that will be all hot and noisy unless I go for watercooling.
Not getting any new info so I guess I can wait like everyone else.
OK, now that I know where you got the information from, I can address it.

First, I'll post AdoredTV addressing it:

It's a good video. Excellent explanation of caches at the start of the video as well.

Now, on to the issue of it being a single channel of 4GB SR ram clocked at 2666 CL19, which would be glacial compared to my [email protected] in quad channel SR sticks. If you think that won't add some latency to you, I don't know what you are thinking. Further, as shown, the cache behavior varies in three examples, one of which had tight latency values. It shows that none of them are trustworthy, but that two may have been outliers compared to the other bench that had tight latencies.

I already explained 720P was rejected as unrealistic, because NO ONE USES 720P anymore! Your adherence to a point INTEL fished to tech journalists is really.... And wanting to embellish performance benefits through making it look more spread out than any experience likely is ... sad! Especially when I said use a 2080 Ti to try to get the CPU to bottleneck at 1080P is proper, as you want the MOST powerful GPU you can use to remove the GPU being the limit to see how much the CPU is hitting performance. So why did you bring up a 470? For someone saying earlier that I needed to focus on non-synthetics, but real-world use, that seems ... yeah.

As to what you want, you just have to wait until June to July and it will be tested in all scenarios. You can make your choice on empirical data. Also, AMD graphics cards are loud, but the CPUs are not bad. And are you talking rendering on the CPU or rendering on the GPU, which 4K60 would be a GPU limit on anything lower than a 2080 Ti and above over the CPU. As you increase resolution, you increase the GPU load, which decreases the load on the CPU. That is the reason as you go to 1440p and 4K, you see the frame rates get closer or sync to within a frame or so of each other.

So just wait for reviews and make your choice on the hard data in front of you.

Edit:
Here is what my latency is in that benchmark on a 1950X:
https://www.userbenchmark.com/UserRun/15351327 (58.7ns latency)
Here is a person with a similar setup, except running 3.95GHz instead of 4.2GHz and ram running at 3200CL14 (likely stock XMP) (85.8ns latency)
https://www.userbenchmark.com/UserRun/15477346

Part of the difference, about 20ns, is due to setting up the interleaving of the ram channels and ranks to lower the latency. When you select channel interleaving, 512b size, etc., it knocks off that amount. So, the difference between the chips most comparable to a 16-core mainstream is that. If you then look at what real time latency 2666CL19 has, you get 14.25ns, going against 3200CL14 giving 8.82ns or 3466CL14 at 8.12ns, which doesn't go into the other timings of the ram that effect the total memory latency.

So the settings for memory and the timings can easily add 10ns or more latency to the score, especially using a single channel, single rank dimm with no channel or rank interleaving, which would give the 20ns or so penalty I showed with my own rig and someone else's rig above. Put them together, it looks like you would be in the territory of these new chips, doesn't it?

That is why trying to read too much into one leaked benchmark regarding the latency will NOT give you a true picture of what is going on with the capabilities of the silicon. Engineering Samples created and tested 5 months before a product launches will not give you the final results you can expect. You don't know what is and isn't tuned yet, what all is being tested, etc. This is why I ignored anyone at the time the benches were making the rounds screaming their head off about this latency. The majority can be explained and shown to be similar to what current chips have, after making the above adjustments.

Here is that first one with the peak of 96.92ns (about 11ns slower than the 3200CL14 example of the 1950X)
https://www.userbenchmark.com/UserRun/14076820

Here is the bugged run with the rise too early in cache latency which hit 100ns
https://www.userbenchmark.com/UserRun/14273098

Also, with knowing the cache was separated from the memory speed, we have ZERO idea what they had the IF2 speed tuned to, meaning we do not know if the engineering sample had a purposely tuned down IF2 to test something else with the chips, trying to remove errors that can occur due to faster IF2 speeds that can cause cache errors or something similar.

So don't buy too heavily into writing off the chip on things easily explainable, especially if the rumor of officially supporting 3200MHz is true and that is achievable with CL14. Between that and interleaving, getting memory latency to the 60ns levels seems to be in the cards without too much effort.

Desktop (Click to show)
Asus Maximus VIII Extreme, G.Skill 4133MHz - Running 3733MHz (MB limitation; I've had 4000, but...)@[email protected] 14-17-17-36-280 with Corsair CMXAF2 Ram fans, EVGA 980 Ti Classified (OC specs to come), [email protected]@1.42V everyday stable, Thermaltake Water 3.0 Ultimate AIO (push/pull 6xNoctua iPPC 3000), EVGA 1600 Ti PSU (for later SLI setups), Thermaltake X9 Cube, Intel 600P NVME
Laptop (Click to show)
P770ZM / Chi Mei 1080p 17.3" TN matte - OC to 100Hz / Intel Core i7 4790K - 4.5GHz 1.130V Static / Nvidia Geforce 980M 8GB (current OC limit of 330MHZ Core/410MHZ VRam/ [email protected]) / Kingston HyperX 32GB DDR3 [email protected] CR-2T custom SPD / Samsung 850 EVO 500GB M.2 / Intel Wifi Dual Band 7265ac+BT / Win 7 Pro x64 - custom by NTLite / 330W PSU/Prema BIOS

Last edited by ajc9988; 03-15-2019 at 07:13 AM.
ajc9988 is offline  
Closed Thread

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off