Linus benchmarks the 1080Ti on the 1800X and 7700K both are basically the same Ryzen does see higher FPS in GTA when it comes to minimums however all games were tested in 4K(Amd must be happy).
I get it its a expensive card and even i'm never going back to 1080P gaming for my desktop but at least include 1440P even Amd recommended that resolution to linus
Since i love high frame rate gaming 4K is still out of reach for me that's fine i spent way to much on my monitor lol
Edit power consumption in Crysis 3 is a little lower with the 1800X compared to a 7700K that is impressive.
another issue is Intel will be releasing 8th gen core chips this year, which is said to be %15 faster than previous chips. while AMD has just released a new chip barely on par with intel 6th-7th gen, with the release of the new series AMD will be falling short again.
So AMD has again some rough times ahead. They will be strugling with devolopers, also will be releasing new drivers and BIOS updates and when they finally have a stable product ready for the market, will face a new opponent. This can cause AMD to loose its edge they have today.
Right now they may be selling a lot of product yet a lot more people like myself will be waiting to see the final results to make a purchase. I have been purchasing AMD chips since Athlon and I will again bu a ryzen chip in the next months but not before I see the platform is stabilized by means of operating system, BIOS and software updates.
If by this year you mean, in the next 15 months. then sure.
As it stands, Intel announced 8th Gen coming in Q2 2018.
I also view that 15% increase as "optimistic" at best. Unless Intel has been lying to everyone since Sandy Bridge released and they haven't been "dogging" it because they have no realistic competition, Intel's performance increases by generation have been more around 5% as of late. I personally don't think their current iCore architecture can be pushed or optimized much further than it already is and aside from just upping stock clock speeds I don't see a 15% performance boost from further optimizations of their current architecture. What that would mean is non "K" processors would have a reasonable performance boost but the "K" variants wouldn't see much compared to current processors overclocked to the the same speeds. I really believe that Intel is already working on another architecture as iCore is already highly optimized and very near the apex of what the architecture is capable of.
Ok, so this *is from WTFBBQtech" so take it as you will however:
https://wccftech.com/amd-ryzen-4-core-benchmarks-intel-core-i7-7700k/
Those are benchmarks, Ryzen 1800X with half the cores and cache disabled to result in a 4 core, 8 thread 8mb cache processor, vs the 7700k. The speeds have been adjusted to be the same (7700K clocked down to 4ghz, the 1800X boosted to 4ghz).
This essentially eliminates any issues pertaining to thread copying between CCX modules. The results show Ryzen being *very close* to the 7700k in gaming once the CCX issue is taken out of the equation... If these benchmarks are accurate then it looks like the 4 core / 8 thread Ryzen part might be a great gaming cpu. It also suggests that the position of the 8 core / 16 thread part is likely to improve quite a bit once the thread scheduling is sorted out.
Ok, so this *is from WTFBBQtech" so take it as you will however:
https://wccftech.com/amd-ryzen-4-core-benchmarks-intel-core-i7-7700k/
Those are benchmarks, Ryzen 1800X with half the cores and cache disabled to result in a 4 core, 8 thread 8mb cache processor, vs the 7700k. The speeds have been adjusted to be the same (7700K clocked down to 4ghz, the 1800X boosted to 4ghz).
This essentially eliminates any issues pertaining to thread copying between CCX modules. The results show Ryzen being *very close* to the 7700k in gaming once the CCX issue is taken out of the equation... If these benchmarks are accurate then it looks like the 4 core / 8 thread Ryzen part might be a great gaming cpu. It also suggests that the position of the 8 core / 16 thread part is likely to improve quite a bit once the thread scheduling is sorted out.
Apart from the GPU bottleneck, the overclock/underclock is diminishing the turbo gap between both processors and distorting real performance at stock settings, not to mention that the quad-core RyZen uses defective silicon that gets lower clocks than their 8-core RyZen. The top quad-core has base clock of 3.5GHz. Also this test is assuming that one full CCX is disabled, but it is not evident that quad-cores will be only 4+0 ZP dies; they could be 2+2 dies.
What began as a simple internal discussion about the validity of claims that Windows 10 scheduling might be to blame for some of Ryzen's performance oddities, and that an update from Microsoft and AMD might magically save us all, has turned into a full day with many people chipping in to help put together a great story. The team at PC Perspective believes strongly that the Windows 10 scheduler is not improperly assigning workloads to Ryzen processors because of a lack of architecture knowledge on the structure of the CPU.
In fact, though we are waiting for official comments we can attribute from AMD on the matter, I have been told from high knowledge individuals inside the company that even AMD does not believe the Windows 10 scheduler has anything at all to do with the problems they are investigating on gaming performance.
In the process, we did find a new source of information in our latency testing tool that clearly shows differentiation between Intel's architecture and AMD's Zen architecture for core to core communications. In this way at least, the CCX design of 8-core Ryzen CPUs appears to more closely emulate a 2-socket system. With that, it is possible for Windows to logically split the CCX modules via the Non-Uniform Memory Access (NUMA), but that would force everything not specifically coded to span NUMA nodes (all games, some media encoders, etc) to use only half of Ryzen. How does this new information affect our expectation of something like Naples that will depend on Infinity Fabric even more directly for AMD's enterprise play?
Ok, so this *is from WTFBBQtech" so take it as you will however:
https://wccftech.com/amd-ryzen-4-core-benchmarks-intel-core-i7-7700k/
Those are benchmarks, Ryzen 1800X with half the cores and cache disabled to result in a 4 core, 8 thread 8mb cache processor, vs the 7700k. The speeds have been adjusted to be the same (7700K clocked down to 4ghz, the 1800X boosted to 4ghz).
This essentially eliminates any issues pertaining to thread copying between CCX modules. The results show Ryzen being *very close* to the 7700k in gaming once the CCX issue is taken out of the equation... If these benchmarks are accurate then it looks like the 4 core / 8 thread Ryzen part might be a great gaming cpu. It also suggests that the position of the 8 core / 16 thread part is likely to improve quite a bit once the thread scheduling is sorted out.
Apart from the GPU bottleneck, the overclock/underclock is diminishing the turbo gap between both processors and distorting real performance at stock settings, not to mention that the quad-core RyZen uses defective silicon that gets lower clocks than their 8-core RyZen. The top quad-core has base clock of 3.5GHz. Also this test is assuming that one full CCX is disabled, but it is not evident that quad-cores will be only 4+0 ZP dies; they could be 2+2 dies.
Running the parts at same speed / thread count does highlight just how close the two can be. Also I think it highlights the problems the *ccx* structure is causing for gaming.
As to the pcper quote- I read that and basically they have totally missed the point. Everything is pointing to the issue being due to *thread copying* across the CCX blocks rather than incorrect use of SMT. That is still something that could be improved upon with software imo- although it's unclear how much the benefit would be. It also ignores the point that disabling SMT does yield a gain in quite a few cases so there's something not quite right. That said I acknowledge that any gains for the above aren't going to be huge.
I see your point on the 4 core not necessarily being constructed from a single ccx if it is a harvested 8 core part. Still- I'm also sure we will see *native quad core Ryzen cpus* once the APU comes out (i.e. apu with the gpu block disabled). Also I was under the impression that Ryzen 3 was a separate die?
Ok, so this *is from WTFBBQtech" so take it as you will however:
https://wccftech.com/amd-ryzen-4-core-benchmarks-intel-core-i7-7700k/
Those are benchmarks, Ryzen 1800X with half the cores and cache disabled to result in a 4 core, 8 thread 8mb cache processor, vs the 7700k. The speeds have been adjusted to be the same (7700K clocked down to 4ghz, the 1800X boosted to 4ghz).
This essentially eliminates any issues pertaining to thread copying between CCX modules. The results show Ryzen being *very close* to the 7700k in gaming once the CCX issue is taken out of the equation... If these benchmarks are accurate then it looks like the 4 core / 8 thread Ryzen part might be a great gaming cpu. It also suggests that the position of the 8 core / 16 thread part is likely to improve quite a bit once the thread scheduling is sorted out.
Apart from the GPU bottleneck, the overclock/underclock is diminishing the turbo gap between both processors and distorting real performance at stock settings, not to mention that the quad-core RyZen uses defective silicon that gets lower clocks than their 8-core RyZen. The top quad-core has base clock of 3.5GHz. Also this test is assuming that one full CCX is disabled, but it is not evident that quad-cores will be only 4+0 ZP dies; they could be 2+2 dies.
Running the parts at same speed / thread count does highlight just how close the two can be. Also I think it highlights the problems the *ccx* structure is causing for gaming.
As to the pcper quote- I read that and basically they have totally missed the point. Everything is pointing to the issue being due to *thread copying* across the CCX blocks rather than incorrect use of SMT. That is still something that could be improved upon with software imo- although it's unclear how much the benefit would be. It also ignores the point that disabling SMT does yield a gain in quite a few cases so there's something not quite right. That said I acknowledge that any gains for the above aren't going to be huge.
I see your point on the 4 core not necessarily being constructed from a single ccx if it is a harvested 8 core part. Still- I'm also sure we will see *native quad core Ryzen cpus* once the APU comes out (i.e. apu with the gpu block disabled). Also I was under the impression that Ryzen 3 was a separate die?
Yes, we will surely see alternative quad-core CPUs from disabling the iGPU in a Raven Ridge die, but (i) those Athlon-like CPUs aren't the R5 CPUs, and (ii) those Athlon-like CPUs have half the L3 cache of the R5 brothers for power reasons, and clocks are unknown (there are even rumors that the Raven Ridge dies use a different process node than Summit Ridge dies).
About CCX--CCX latency, PCPer makes it clear:
Some of you may believe that there could be some optimization to the Windows scheduler to fix this issue. Perhaps keep processes on one CCX if at all possible. Well in the testing we did, that was also happening. Here is the SMT ON result for a lighter (13%) workload using two threads:
See what's going on there? The Windows scheduler was already keeping those threads within the same CCX. This was repeatable (some runs were on the other CCX) and did not appear to be coincidental. Further, the example shown in the first (bar) chart demonstrated a workload spread across the four cores in CCX 0.
If there are few threads, those are already scheduled on the same CCX. If there are more threads than cores then they have to be scheduled on both CCX. If the threads are independent (like in throughput workloads) then there is no problem, because each CCX works almost independently of the other. If the threads are dependent then the latency issue appears again.
There is no magic trick that will solve the gaming/latency problem. There is a fundamental design flaw in Zen, and as I mentioned before, this will be not solved until Zen+, if it is solved at all!
Ok, so this *is from WTFBBQtech" so take it as you will however:
https://wccftech.com/amd-ryzen-4-core-benchmarks-intel-core-i7-7700k/
Those are benchmarks, Ryzen 1800X with half the cores and cache disabled to result in a 4 core, 8 thread 8mb cache processor, vs the 7700k. The speeds have been adjusted to be the same (7700K clocked down to 4ghz, the 1800X boosted to 4ghz).
This essentially eliminates any issues pertaining to thread copying between CCX modules. The results show Ryzen being *very close* to the 7700k in gaming once the CCX issue is taken out of the equation... If these benchmarks are accurate then it looks like the 4 core / 8 thread Ryzen part might be a great gaming cpu. It also suggests that the position of the 8 core / 16 thread part is likely to improve quite a bit once the thread scheduling is sorted out.
Apart from the GPU bottleneck, the overclock/underclock is diminishing the turbo gap between both processors and distorting real performance at stock settings, not to mention that the quad-core RyZen uses defective silicon that gets lower clocks than their 8-core RyZen. The top quad-core has base clock of 3.5GHz. Also this test is assuming that one full CCX is disabled, but it is not evident that quad-cores will be only 4+0 ZP dies; they could be 2+2 dies.
Running the parts at same speed / thread count does highlight just how close the two can be. Also I think it highlights the problems the *ccx* structure is causing for gaming.
As to the pcper quote- I read that and basically they have totally missed the point. Everything is pointing to the issue being due to *thread copying* across the CCX blocks rather than incorrect use of SMT. That is still something that could be improved upon with software imo- although it's unclear how much the benefit would be. It also ignores the point that disabling SMT does yield a gain in quite a few cases so there's something not quite right. That said I acknowledge that any gains for the above aren't going to be huge.
I see your point on the 4 core not necessarily being constructed from a single ccx if it is a harvested 8 core part. Still- I'm also sure we will see *native quad core Ryzen cpus* once the APU comes out (i.e. apu with the gpu block disabled). Also I was under the impression that Ryzen 3 was a separate die?
Yes, we will surely see alternative quad-core CPUs from disabling the iGPU in a Raven Ridge die, but (i) those Athlon-like CPUs aren't the R5 CPUs, and (ii) those Athlon-like CPUs have half the L3 cache of the R5 brothers for power reasons, and clocks are unknown (there are even rumors that the Raven Ridge dies use a different process node than Summit Ridge dies).
About CCX--CCX latency, PCPer makes it clear:
Some of you may believe that there could be some optimization to the Windows scheduler to fix this issue. Perhaps keep processes on one CCX if at all possible. Well in the testing we did, that was also happening. Here is the SMT ON result for a lighter (13%) workload using two threads:
See what's going on there? The Windows scheduler was already keeping those threads within the same CCX. This was repeatable (some runs were on the other CCX) and did not appear to be coincidental. Further, the example shown in the first (bar) chart demonstrated a workload spread across the four cores in CCX 0.
If there are few threads, those are already scheduled on the same CCX. If there are more threads than cores then they have to be scheduled on both CCX. If the threads are independent (like in throughput workloads) then there is no problem, because each CCX works almost independently of the other. If the threads are dependent then the latency issue appears again.
There is no magic trick that will solve the gaming/latency problem. There is a fundamental design flaw in Zen, and as I mentioned before, this will be not solved until Zen+, if it is solved at all!
That's what everyone is fearing, and trying to believe is false: Ryzen has a flaw for consumer workloads, especially games, and it can only be fixed by engineering, or very clever programming. I believe it will get better with Zen+ on the hardware side, but developers might have a hard time reengineering their software to work on a different multi-core paradigm.
But this video points to the problem in games: PC Perspective tested using custom C++ code, and Windows worked right. But games are not aware of this dual-CPU approach, and thread scheduling goes wrong in them. Also, games may spam 10 threads, which breaks the possibility of staying inside one CCX. That can be fixed by software, but how many developers will do it?
There are two issues that affect Ryzen overclocking in a way that the benchmark scores reported do not reflect real performance. HWBOT has officially banned certain submissions and investigating both the RTC bias and the sleep bug
CanardPC got access to the first six-core and quad-core models: R5 1600, R5 1400, and R3 1200. Base clocks 3.2GHz and lower, there is no all-core turbo and the single-core turbo is 3.4GHz in the best case. Most of the discussion is in French, but you can access one English tweet here
<Link removed by moderator> <As this is the English language version of Tom's Hardware please post links to English language content only>
@juan, AMD themselves have announced 1600x specifications of 3.6ghz base, 4ghz turbo. That is imo going to be the stand out gaming part for ryzen- given the same clocks it should match the r7s in games as 6 core / 12 threads is ample, but at a substantially lower price.
Also, even in the worst case that gaming can't be 'fixed' on ryzen, it does in fact perform pretty well. I mean the jump from fx is enormous in most cases, irrespective of the 'bugs'. I also think some developers may well find good ways to utilise at the extra threads- however it's going to be a much more 'case by case' basis. I guess that means for high end gaming Intel will remain king, but for users like myself who use their machines for multiple purposes ryzen does offer some nice incentives and the games numbers I'm seeing don't look to pose a real problem.
Juan if that pans out right, we should be seeing a price drop soon enough (If people actually stop being fanboys).
Edit: I don't understand it that much, and I don't even know if the benchmarks are correct.
Yes, that PcPer link was given above and discussed in this same page.
cdrkf :
@juan, AMD themselves have announced 1600x specifications of 3.6ghz base, 4ghz turbo. That is imo going to be the stand out gaming part for ryzen- given the same clocks it should match the r7s in games as 6 core / 12 threads is ample, but at a substantially lower price.
If you check the twitter discussion I gave the specs for the 1600X and the 1500X. If you check his response he said me there is no trace of those models and he speculates the timeline has changed.
I am not fully convinced that the R5 1600X will match the R7 1800X on games. Hardware.fr found that the six-core Broadwell was 4% behind the eight-core Broadwell despite having 6% higher clocks, and the 6C Ryzen has same clocks than the 8C Ryzen. If those percentages apply to Ryzen then the 1600X could be 10% behind the 1800X, but in the other hand the 1600X seems to have more L3 cache per core and that could reduce the gap. It is difficult to estimate accurate performance in this case.
Getting ready to pull the trigger on a 1700 build. The problems with 'gaming' are overblown, it's still a very solidly performing chip, and I want the extra threads. Just gotta get some more play-dough together.
Getting ready to pull the trigger on a 1700 build. The problems with 'gaming' are overblown, it's still a very solidly performing chip, and I want the extra threads. Just gotta get some more play-dough together.
Yes, the gaming issues are kind of overblown. And you're right; for the price is a very decent power house as an "all rounder" CPU. Nothing to feel ashamed from the Intel platforms, specially using something decent with a B350 chipset.
EDIT: Forgot to mention, the *baseline* performance it displays in most day-1 reviews is good and, even though I do expect the software patches to not do any miracles, they will make things better, so the more reason to have a bit of confidence in shelling some money in AMD now (maybe not stocks for now? heh).
And any news about the APU? I'm more interested in that lil' beauty now, more than the rest of the Ryzen lineup, haha.
Juan if that pans out right, we should be seeing a price drop soon enough (If people actually stop being fanboys).
Edit: I don't understand it that much, and I don't even know if the benchmarks are correct.
Yes, that PcPer link was given above and discussed in this same page.
cdrkf :
@juan, AMD themselves have announced 1600x specifications of 3.6ghz base, 4ghz turbo. That is imo going to be the stand out gaming part for ryzen- given the same clocks it should match the r7s in games as 6 core / 12 threads is ample, but at a substantially lower price.
If you check the twitter discussion I gave the specs for the 1600X and the 1500X. If you check his response he said me there is no trace of those models and he speculates the timeline has changed.
I am not fully convinced that the R5 1600X will match the R7 1800X on games. Hardware.fr found that the six-core Broadwell was 4% behind the eight-core Broadwell despite having 6% higher clocks, and the 6C Ryzen has same clocks than the 8C Ryzen. If those percentages apply to Ryzen then the 1600X could be 10% behind the 1800X, but in the other hand the 1600X seems to have more L3 cache per core and that could reduce the gap. It is difficult to estimate accurate performance in this case.
That's a fair point- it's possible a 6c Ryzen might drop a bit of performance. Still, we're not talking much if it does drop a bit. The key (as always) is price, my take away from all this is whilst Ryzen doesn't really compete head to head with Skylake (which let's be honest few of us expected it to), it's close enough that priced well it makes a lot of sense. I do agree with many here in the fact that the main issue isn't the performance is terrible, but rather AMD shouldn't have pitched it as a Skylake competitor. To compare it in current titles against the 7700k was a very poor decision (and if they were going to do so- they should have had a few patched titles ready to back up their claims- I do think things like Ashes will potentially fly with a bit of work as it's the sort of title that can leverage a high core count cpu).
Ok, so this *is from WTFBBQtech" so take it as you will however:
https://wccftech.com/amd-ryzen-4-core-benchmarks-intel-core-i7-7700k/
Those are benchmarks, Ryzen 1800X with half the cores and cache disabled to result in a 4 core, 8 thread 8mb cache processor, vs the 7700k. The speeds have been adjusted to be the same (7700K clocked down to 4ghz, the 1800X boosted to 4ghz).
This essentially eliminates any issues pertaining to thread copying between CCX modules. The results show Ryzen being *very close* to the 7700k in gaming once the CCX issue is taken out of the equation... If these benchmarks are accurate then it looks like the 4 core / 8 thread Ryzen part might be a great gaming cpu. It also suggests that the position of the 8 core / 16 thread part is likely to improve quite a bit once the thread scheduling is sorted out.
Apart from the GPU bottleneck, the overclock/underclock is diminishing the turbo gap between both processors and distorting real performance at stock settings, not to mention that the quad-core RyZen uses defective silicon that gets lower clocks than their 8-core RyZen. The top quad-core has base clock of 3.5GHz. Also this test is assuming that one full CCX is disabled, but it is not evident that quad-cores will be only 4+0 ZP dies; they could be 2+2 dies.
Running the parts at same speed / thread count does highlight just how close the two can be. Also I think it highlights the problems the *ccx* structure is causing for gaming.
As to the pcper quote- I read that and basically they have totally missed the point. Everything is pointing to the issue being due to *thread copying* across the CCX blocks rather than incorrect use of SMT. That is still something that could be improved upon with software imo- although it's unclear how much the benefit would be. It also ignores the point that disabling SMT does yield a gain in quite a few cases so there's something not quite right. That said I acknowledge that any gains for the above aren't going to be huge.
I see your point on the 4 core not necessarily being constructed from a single ccx if it is a harvested 8 core part. Still- I'm also sure we will see *native quad core Ryzen cpus* once the APU comes out (i.e. apu with the gpu block disabled). Also I was under the impression that Ryzen 3 was a separate die?
Yes, we will surely see alternative quad-core CPUs from disabling the iGPU in a Raven Ridge die, but (i) those Athlon-like CPUs aren't the R5 CPUs, and (ii) those Athlon-like CPUs have half the L3 cache of the R5 brothers for power reasons, and clocks are unknown (there are even rumors that the Raven Ridge dies use a different process node than Summit Ridge dies).
About CCX--CCX latency, PCPer makes it clear:
Some of you may believe that there could be some optimization to the Windows scheduler to fix this issue. Perhaps keep processes on one CCX if at all possible. Well in the testing we did, that was also happening. Here is the SMT ON result for a lighter (13%) workload using two threads:
See what's going on there? The Windows scheduler was already keeping those threads within the same CCX. This was repeatable (some runs were on the other CCX) and did not appear to be coincidental. Further, the example shown in the first (bar) chart demonstrated a workload spread across the four cores in CCX 0.
If there are few threads, those are already scheduled on the same CCX. If there are more threads than cores then they have to be scheduled on both CCX. If the threads are independent (like in throughput workloads) then there is no problem, because each CCX works almost independently of the other. If the threads are dependent then the latency issue appears again.
There is no magic trick that will solve the gaming/latency problem. There is a fundamental design flaw in Zen, and as I mentioned before, this will be not solved until Zen+, if it is solved at all!
That's what everyone is fearing, and trying to believe is false: Ryzen has a flaw for consumer workloads, especially games, and it can only be fixed by engineering, or very clever programming. I believe it will get better with Zen+ on the hardware side, but developers might have a hard time reengineering their software to work on a different multi-core paradigm.
But this video points to the problem in games: PC Perspective tested using custom C++ code, and Windows worked right. But games are not aware of this dual-CPU approach, and thread scheduling goes wrong in them. Also, games may spam 10 threads, which breaks the possibility of staying inside one CCX. That can be fixed by software, but how many developers will do it?
https://www.youtube.com/watch?v=40h4skxDkh4
NUMA support is still coming.
Also, it has been discovered that IF runs at half the rate of your RAM. So, higher RAM frequencies not only boost IMC performance, but also IF transfer rates internally. So...as RAM becomes validated at higher frequencies, IF performance will increase as well.
Juan if that pans out right, we should be seeing a price drop soon enough (If people actually stop being fanboys).
Edit: I don't understand it that much, and I don't even know if the benchmarks are correct.
Yes, that PcPer link was given above and discussed in this same page.
cdrkf :
@juan, AMD themselves have announced 1600x specifications of 3.6ghz base, 4ghz turbo. That is imo going to be the stand out gaming part for ryzen- given the same clocks it should match the r7s in games as 6 core / 12 threads is ample, but at a substantially lower price.
If you check the twitter discussion I gave the specs for the 1600X and the 1500X. If you check his response he said me there is no trace of those models and he speculates the timeline has changed.
I am not fully convinced that the R5 1600X will match the R7 1800X on games. Hardware.fr found that the six-core Broadwell was 4% behind the eight-core Broadwell despite having 6% higher clocks, and the 6C Ryzen has same clocks than the 8C Ryzen. If those percentages apply to Ryzen then the 1600X could be 10% behind the 1800X, but in the other hand the 1600X seems to have more L3 cache per core and that could reduce the gap. It is difficult to estimate accurate performance in this case.
That's a fair point- it's possible a 6c Ryzen might drop a bit of performance. Still, we're not talking much if it does drop a bit. The key (as always) is price, my take away from all this is whilst Ryzen doesn't really compete head to head with Skylake (which let's be honest few of us expected it to), it's close enough that priced well it makes a lot of sense. I do agree with many here in the fact that the main issue isn't the performance is terrible, but rather AMD shouldn't have pitched it as a Skylake competitor. To compare it in current titles against the 7700k was a very poor decision (and if they were going to do so- they should have had a few patched titles ready to back up their claims- I do think things like Ashes will potentially fly with a bit of work as it's the sort of title that can leverage a high core count cpu).
Reviewers compared it to the 7700K, because YT reviewers...
AMD compared it squarely to the 6900K, which is a pretty valid comparison.
Ryzen has already almost and in some made up it's 1080p gaming deficit and will only get better. Ryzen and gaming are all about fast memory . The way Ryzen works a couple of other clocks are directly tied to the memory speed one of them being the HT link/south bridge and has a direct effect on gaming. Most review sites either didn't have the support or bother taking the time to tune memory and just ran it between 2133 and 2400. That has a big negative effect cause it slows other clocks down also. Anyway look at the link 3-4 different sites saying going from 2400 to 3200 mem speed with Ryzen gets you 10-15% in 1080p gaming.Gskill has a Ryzen certified 3488 kit coming soon bet that thing will make Ryzen sing. http://www.overclock.net/t/1625187/the-ryzen-gaming-performance-gap-is-mostly-gone
Ryzen has already almost and in some made up it's 1080p gaming deficit and will only get better. Ryzen and gaming are all about fast memory . The way Ryzen works a couple of other clocks are directly tied to the memory speed one of them being the HT link/south bridge and has a direct effect on gaming. Most review sites either didn't have the support or bother taking the time to tune memory and just ran it between 2133 and 2400. That has a big negative effect cause it slows other clocks down also. Anyway look at the link 3-4 different sites saying going from 2400 to 3200 mem speed with Ryzen gets you 10-15% in 1080p gaming.Gskill has a Ryzen certified 3488 kit coming soon bet that thing will make Ryzen sing.
http://www.overclock.net/t/1625187/the-ryzen-gaming-performance-gap-is-mostly-gone
That's good to hear. I can't wait until AMD fixes issues with high memory frequencies so this is possible for more folks.
Ryzen has already almost and in some made up it's 1080p gaming deficit and will only get better. Ryzen and gaming are all about fast memory . The way Ryzen works a couple of other clocks are directly tied to the memory speed one of them being the HT link/south bridge and has a direct effect on gaming. Most review sites either didn't have the support or bother taking the time to tune memory and just ran it between 2133 and 2400. That has a big negative effect cause it slows other clocks down also. Anyway look at the link 3-4 different sites saying going from 2400 to 3200 mem speed with Ryzen gets you 10-15% in 1080p gaming.Gskill has a Ryzen certified 3488 kit coming soon bet that thing will make Ryzen sing.
http://www.overclock.net/t/1625187/the-ryzen-gaming-performance-gap-is-mostly-gone
That's good to hear. I can't wait until AMD fixes issues with high memory frequencies so this is possible for more folks.
The memory issue is a MB support issue. Though memory validation does play a part to a degree.
Juan if that pans out right, we should be seeing a price drop soon enough (If people actually stop being fanboys).
Edit: I don't understand it that much, and I don't even know if the benchmarks are correct.
Yes, that PcPer link was given above and discussed in this same page.
cdrkf :
@juan, AMD themselves have announced 1600x specifications of 3.6ghz base, 4ghz turbo. That is imo going to be the stand out gaming part for ryzen- given the same clocks it should match the r7s in games as 6 core / 12 threads is ample, but at a substantially lower price.
If you check the twitter discussion I gave the specs for the 1600X and the 1500X. If you check his response he said me there is no trace of those models and he speculates the timeline has changed.
I am not fully convinced that the R5 1600X will match the R7 1800X on games. Hardware.fr found that the six-core Broadwell was 4% behind the eight-core Broadwell despite having 6% higher clocks, and the 6C Ryzen has same clocks than the 8C Ryzen. If those percentages apply to Ryzen then the 1600X could be 10% behind the 1800X, but in the other hand the 1600X seems to have more L3 cache per core and that could reduce the gap. It is difficult to estimate accurate performance in this case.
That's a fair point- it's possible a 6c Ryzen might drop a bit of performance. Still, we're not talking much if it does drop a bit. The key (as always) is price, my take away from all this is whilst Ryzen doesn't really compete head to head with Skylake (which let's be honest few of us expected it to), it's close enough that priced well it makes a lot of sense. I do agree with many here in the fact that the main issue isn't the performance is terrible, but rather AMD shouldn't have pitched it as a Skylake competitor. To compare it in current titles against the 7700k was a very poor decision (and if they were going to do so- they should have had a few patched titles ready to back up their claims- I do think things like Ashes will potentially fly with a bit of work as it's the sort of title that can leverage a high core count cpu).
Reviewers compared it to the 7700K, because YT reviewers...
AMD compared it squarely to the 6900K, which is a pretty valid comparison.
AMD compared Ryzen to 7700k both in slides and demos...
Facebook
Twitter
Instagram