[H] Ryzen benchmark question/suggestions

JDanser

Limp Gawd
Joined
Feb 9, 2012
Messages
246
If/when the [H] gets their grubby mitts on some Ryzen hardware to abuse, what are you wanting to see as far as hardware used, tests done, etc?

Personally, I'd like to see an older machine or 2 thrown in as a "holdout" setup. 2600k/3770k/9550/etc and some faster ddr3 to compare to would be neat. Plus, justification for dropping the dime would be neat.
 
Like to see stability tests (severe) when OC, OCing results (of course a later review) and temperature profile of the board to see hot spots.
Normal peripheral testing for USB 3.1/3.0 PCIe M2 speeds etc.
Sound quality and noise
Ram testing with different speeds and brands as much as possible - what works and what does not work
CPU coolers available (probably much later then initial launch) maybe some decent coolers will fit that will help in determining OC ability.

Then some real world testing and workloads such as virus scan, plus download while gaming scenarios. Maybe a unique test in maxing out the gpu pcie 3.0 lanes like two 1080's in SLI (since that will be 8x/8x) and then redo the M2 PCIe SSD tests to see any limitations.

Rendering benchmarks to use those cores. Game benchmarks including VR and using those USB ports to see if any issues.

Remember this is not just a cpu release but a whole new platform update.

I don't see the need to include older systems on the initial reviews, better to concentrate what is available and how they compare. Later though would be interesting to see how far things has come such as a Bulldozer system to a RyZen system but there are plenty of other sites that can waste that time doing those comparisons.
 
I'd like to see Total War: Warhammer benchmarks. NVME and SATA storage performance including RAID 0/5/10 benchmarks. Boot times. AIDA memory and cache numbers compared to Kaby Lake and Broadwell-E. Memory frequency and timing scaling. SLI 1080 numbers compared to Broadwell-E and Kaby Lake.
 
Remember this is not just a cpu release but a whole new platform update.

I don't see the need to include older systems on the initial reviews, better to concentrate what is available and how they compare. Later though would be interesting to see how far things has come such as a Bulldozer system to a RyZen system but there are plenty of other sites that can waste that time doing those comparisons.


It really is a significant platform update, and you are 100% correct that is needing to be a focus. for the hardcore hold out AMD guys, there is a lot to be covered. for the intel guys, it is how much parity is there now, if there is any. for those of still running SB/IB (and there is a bunch of us) a comparison is also in order, simply because is SL wasn't enough, KL really isn't bringing much more either. and finally we all need to know how our ePeen stacks up against it. But given the number of people who have held out on an upgrade to see what this brings, absolutely no reason not to look at the older stuff as well.
 
for those of still running SB/IB (and there is a bunch of us) a comparison is also in order, simply because is SL wasn't enough, KL really isn't bringing much more either. and finally we all need to know how our ePeen stacks up against it.

Nailed it. Normally to get some kind of idea about how older platforms stack up against newer platforms is hope that whatever review you're reading uses some kind of benchmark that you can reliably run on your machine. Otherwise you have to compare in generations (Okay, Sandy~==Ivy, Haswell > Ivy, Broadwell~==Haswell, Skylake > Broadwell, etc, etc.)
 
Nailed it. Normally to get some kind of idea about how older platforms stack up against newer platforms is hope that whatever review you're reading uses some kind of benchmark that you can reliably run on your machine. Otherwise you have to compare in generations (Okay, Sandy~==Ivy, Haswell > Ivy, Broadwell~==Haswell, Skylake > Broadwell, etc, etc.)

If he can get a hold of that 2600K again that he used for the IPC comparison, that would be ideal.

An IPC comparison of say 2600K and 7700K and Ryzen @ a fixed frequency on ST and MT apps. If one was going to bother with a comparison to the BD architecture, this is the only spot it would be worthwhile IMHO.

Then obvious stock and max OC with all 3 setups previously mentioned on several ST heavy and several MT heavy games. For MT if Overwatch/BF1 could be done (in multiplayer for the first two, I know it's not perfect but it's very much different than single player).
 
It would also be interesting if [H] could do some testing of the XFR modes on one of the chips with different cooling. For instance, they could compare clocks/performance using a "stock" cooler and maybe a good air cooler or a good water cooling setup. This would give us an idea of how cooling affects the clocks for XFR mode (assuming it works the way all the leaks have been saying).
 
Now that Ryzen ECC capability is apparently confirmed through mobo specs, I would like to know how much of a performance impact ECC will result in.
One would only run ECC if using a workstation or a server. I don't see the point of running ECC in a purely gaming box but then most of us game and work on the same box so I see your point.
 
One would only run ECC if using a workstation or a server. I don't see the point of running ECC in a purely gaming box but then most of us game and work on the same box so I see your point.

I am interested in this comparison, because my PC is very multipurpose. Gaming, Workstation, Plex Media Server, etc. It would also give insight into how well Ryzen would work for heavier server tasks when compared to similar Xeons. Especially since we know that Xeons don't overclock well, and Intel is notorious for segmenting their product lines (no ECC on most motherboards).
 
Yeah an overclockable server chip would be fun for sure. I too am watching that space.
 
I have been holding off on my new build to see how well Ryzen compares with Intel so that I can make the correct decision as I imagine many others are. I am building exclusively for gaming and will be replacing my i5 2500k with a 7600k/7700k or a Ryzen CPU. What I want to see is benchmarks for the most popular games comparing the 7600k and 7700k to AMD's offering at similar price points. I'm most interested in the OC vs OC benchmarks because I'll be overclocking whichever one I go with. I would like to see these comparisons using a GTX1070 and 16gb RAM at 1440p that is what will be in my new build.

I have been holding off for so long that I am very eager to make my decision and purchase the rest of my build. I know it's wishful thinking but hopefully these reviews and benchmarks can be up ASAP once the Ryzen processors are available for purchase.
 
A multi-purpose rig that can do all trades can be a rig that does nothing well. Sometimes it is better to build rigs to specific purposes if very important for professional reasons. Then again if you don't game much but do need that type of break I would think most faster server chips will do just fine on a number of games. If one is looking for the fastest gaming rig for a 144mhz high end monitor then high IPC with high clock speed is probably the best route a.k.a I7 7700K or OC I7 5960x especially with dual GPU's which may not make the best server or rendering rig.
 
A multi-purpose rig that can do all trades can be a rig that does nothing well. Sometimes it is better to build rigs to specific purposes if very important for professional reasons. Then again if you don't game much but do need that type of break I would think most faster server chips will do just fine on a number of games. If one is looking for the fastest gaming rig for a 144mhz high end monitor then high IPC with high clock speed is probably the best route a.k.a I7 7700K or OC I7 5960x especially with dual GPU's which may not make the best server or rendering rig.

Not as good as Ryzen for video streaming. Which means if you do multiplayer game play Ryzen would take advantage of the streaming aspect better than 6700k or 7700k.
 
Similar situation with 2 systems: i5 2500K for gaming, an i7-2600K for production. Both are old but given the small advances in CPU speed -- do they both need to be upgraded? I'm not expecting a lot of noticeable difference. My i5 is OC'd to 4.5Ghz and is using a GTX 980 GPU connected to an Asus Rog Swift 2560x1440 monitor. What kind of improvement can an overclocked Ryzen give me over my current system in popular games compared to a 7600K? Or is a GPU upgrade the most bang for the buck? I'd love to see gaming FPS benchmarks based on common overclock speeds of the older chips vs the new chips.

I have been holding off on my new build to see how well Ryzen compares with Intel so that I can make the correct decision as I imagine many others are. I am building exclusively for gaming and will be replacing my i5 2500k with a 7600k/7700k or a Ryzen CPU. What I want to see is benchmarks for the most popular games comparing the 7600k and 7700k to AMD's offering at similar price points. I'm most interested in the OC vs OC benchmarks because I'll be overclocking whichever one I go with. I would like to see these comparisons using a GTX1070 and 16gb RAM at 1440p that is what will be in my new build.

I have been holding off for so long that I am very eager to make my decision and purchase the rest of my build. I know it's wishful thinking but hopefully these reviews and benchmarks can be up ASAP once the Ryzen processors are available for purchase.
 
A multi-purpose rig that can do all trades can be a rig that does nothing well. Sometimes it is better to build rigs to specific purposes if very important for professional reasons. Then again if you don't game much but do need that type of break I would think most faster server chips will do just fine on a number of games. If one is looking for the fastest gaming rig for a 144mhz high end monitor then high IPC with high clock speed is probably the best route a.k.a I7 7700K or OC I7 5960x especially with dual GPU's which may not make the best server or rendering rig.

Yes - Separate rigs or dedicated servers will certainly do each thing "better", but at what point do we say "close enough"? I don't have the space or cash to spend money on a separate server, workstation and gaming PC. For me (and I suspect many others), the value is how many usable features I can get for the price. ECC will slow the computer down some, sure. But how large of an impact are we talking? Performing this analysis will also give insight into how Zen performs at different Ram speeds.

I want to have my cake and eat it too, and I am cautiously optimistic about AMD. If it fails, I will likely pick up a Xeon chipset and go from there. (Please don't suck)
 
Yes - Separate rigs or dedicated servers will certainly do each thing "better", but at what point do we say "close enough"? I don't have the space or cash to spend money on a separate server, workstation and gaming PC. For me (and I suspect many others), the value is how many usable features I can get for the price. ECC will slow the computer down some, sure. But how large of an impact are we talking? Performing this analysis will also give insight into how Zen performs at different Ram speeds.

I want to have my cake and eat it too, and I am cautiously optimistic about AMD. If it fails, I will likely pick up a Xeon chipset and go from there. (Please don't suck)

Right on! That's why virtualization performance would be a good metric.

P.S. Romeomium... I'm not stalking you man. Coincidence.
 
Back
Top