Yes it is huge problem for people who needs high fps wants spend money on premium memory, but don't want to see buy a discrete GPU to get way higher performance.
Compared with Epyc(1) it should be pretty easy to improve latency, with the current Epyc 75% of the memory channels are remote with really high latency (around 130-150 ns depending on the memory). Infinity Fabric 2 has a lot of improvements to latency.
Currently on Epyc(1) we have 4x8 core = 32, with 8 cores per chip, 2 CCX per Chip and 4 cores per CCX.
We have no way of knowing if each chiplets has 2 CCX with 4 cores like current chips or they are moving to an 8 core CCX.
Personally I believe in an 8 core CCX, but it is a guess since...
The arm CPU in the secure processor is an inorder ARM Cortex-A5 core, Spectre and Meltdown like flaws do not apply to inorder CPU cores.
The "only" ARM cores impacted are modern high performance ones.
You would have been better off if you har the statements from AMD and not just random articles on the internet. AMD never denied Spectre 1 nor 2, people who claim otherwise is either dishonest or ignorant.
The chance of 2nd generation infinity band is zero. They might very well do some minor tweaks reducing latency.
Precession boost 2 is a given, increasing the average clockspeed.
Slight tweaks to the memory controller and improved AGESA will increase memory speeds and improve compatibility...
Navi "scalability" is from a 2016 capsaicin event slide. Some people interpret the "scalability" as modular in reality nobody seems to know what it means.
CPU usage(in the rendering pipeline) per frame is generally speaking constant no matter the resolution, if the frame rates drops due to higher resolution the CPU usage drops too.
AI, physics etc might run independent of the frame rate.
Let me summarise:
Topweasel: nvidia used to scale fine with more cores, but with current drivers they don't scale above four cores anymore.
Carfax83: you are wrong, look at these old benchmarks nvidia drivers does scale with more cores.
Mostly as expected, the R5 crushes the i5 in multithreaded applications and trades blows in games.
Some interesting benchmarks in the review like r480 vs gf1060 in 1080p Rocket League on Ryzen.
I agree that the 5.5% performance improvement in 1080p overall is insignificant and does not make a difference in real life experience (it is like a 250 MHz increase). But if you play Fallout 4, Far Cry Primal or Total War: Warhammer or similar you might well fell the difference.
Games at 1080p do show a difference and in some games the difference is very significant. A lot of reviews over-emphasized low resolution gaming - faster memory will make a (perceived) difference here.
MajinCry is onto to the fact that if the workload is single threaded, the thread is jumping between CCX modules and slow memory is used then the performance is going to be suboptimal with Ryzen.
The problem is that he does not seem to care about the fact that most modern workloads (including...
The issues with the Ryzen motherboards are mostly related to memory clocks, Bristol Ridge is limited to DDR4-2400 anyway.
The supply issues are worst with the high end boards, an earlier release of Bristol Ridge would only have improved the availability of entry level boards.
Nobody is expecting games to implement user mode scheduling, the potential 3-7% performance improvement is simply not worth it.
There is a difference between a lack of optimisation and a bug, it is pretty big stretch to claim that the windows scheduler is buggy due to the missing CCX...
You completely misunderstood the F1 2016 issue. The issue is that F1 2016 is that when it is is installed it saves the CPU topology in the settings, the settings is backed up in steam and if you reinstall on a new system the settings will downloaded again with the topology for the old CPU.
A...
They are not replacing Polaris, Vega will slot in higher in their line up. They will continue to sell P10 for at least another year (it might get rebranded at some point).
Most likely fp16 vs Fiji(Fury non X), that is possible with tdp of 230W(higher with a higher than expected clock speed).
Sendt fra min SM-G928F med Tapatalk
Those buzzwords don't tell us much about gaming performance. 2x peak performance per clock due to double rate fp16, twice the bandwidth and 8x the memory per stack per pin due to HBM2 etc.. hopefully we will get some real performance info at the event.
Sendt fra min SM-G928F med Tapatalk
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.