• 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's Infinity Fabric

I´m looking at pairing my 1080ti with something decent, say a ryzen 2600 on an msi carbon pro motherboard. I am mulling over the cl16 3200 memory from ocuk at £128 or thinking of going the whole hog on cl14 samsung b die pro stuff at £199. Thing is, and i think i know the answer to this question, am i likely to see any benefit in fps at 2k/4k 60hz? I'm guessing not?
 
single_channel_vs_dual_channel.jpg


https://forums.overclockers.co.uk/t...l-vs-dual-channel-best-ram-settings.18831204/


For one thing that compares bandwidth.

The fabric speed has barely changed. Single or double channel memory has the same infinity fabric performance within the cpu and 3466 is a marginal bump of 8% where as going from 2666Mhz to 3200Mhz is a 20% speed bump. If the infinity fabric were say bottlenecked at 3000Mhz then having them all above 3200Mhz proves nothing.

The difference between 3200Mhz single and dual channel starts and finishes between the memory controller and the memory. CPUs aren't particularly bandwidth limited, more often latency limited and increasing the infinity fabric speed will reduce latency when moving data from cache to cpu, from one cpu to another, etc.

The only thing those benchmarks show is that memory bandwidth doesn't do a lot for performance when gaming on a discrete gpu, it doesn't tell you much of anything about infinity fabric performance.
 
I could and it probably would overclock a bit but I can't be arsed to mess around with it, the difference between 2133 and 2400 is less than 2133 to 3000 obviously. There maybe small gains purely down to RAM speed but it wouldn't be in the hundreds.

Rather than peeps just speculating what the differences may or not be, below are 3 Cinebench runs done less than half an hour ago. The clock speed is the same on all 3 runs, the only difference is ramspeed and ram timmings. I used Cinebench because the original question in the first post was about how ramspeed affects the cpu, as usual the thread has wondered off into gamming. As we all know, Cinebench runs all cores at the max, so any difference in the scores can only be down to ramspeed and timmings.

2kh57a.jpg

288s6d4.jpg

110bslc.jpg
 
Rather than peeps just speculating what the differences may or not be, below are 3 Cinebench runs done less than half an hour ago. The clock speed is the same on all 3 runs, the only difference is ramspeed and ram timmings. I used Cinebench because the original question in the first post was about how ramspeed affects the cpu, as usual the thread has wondered off into gamming. As we all know, Cinebench runs all cores at the max, so any difference in the scores can only be down to ramspeed and timmings.

2kh57a.jpg

288s6d4.jpg

110bslc.jpg

To be fair to the others, the OP got the answer but then brought up gaming himself lol.
 
To be fair to the others, the OP got the answer but then brought up gaming himself lol.

OP has zero credibility when running 32GB of ram for "gaming" and complaining he bought "Slow ram" OP needs to learn how to spec PC components correctly to achieve the best performance per £ spent as lets be honest, buying 32GB of slow ram for gaming is...well...ludicrous and screams you have no idea what you are doing.
 
OP has zero credibility when running 32GB of ram for "gaming" and complaining he bought "Slow ram" OP needs to learn how to spec PC components correctly to achieve the best performance per £ spent as lets be honest, buying 32GB of slow ram for gaming is...well...ludicrous and screams you have no idea what you are doing.

Lol, I think that may be end of thread ;)
 
Didn't buy it for gaming, bought it because I wanted the e-peen of having 32GB. By the time I need faster RAM for gaming I'll have moved on to DDR5 no doubt and a better graphics card. My old i5-2500k with 16GB RAM and Radeon 7950 lasted me 5-7 years.

Only upgraded because I wanted to double my CPU and graphics score in 3D Mark Timespy which I did and also doubled my RAM and made it a bit faster too 1333 vs 2133 now. My old system was perfectly OK just got the upgrade itch and upgrade prices weren't bad.
 
OP has zero credibility when running 32GB of ram for "gaming" and complaining he bought "Slow ram" OP needs to learn how to spec PC components correctly to achieve the best performance per £ spent as lets be honest, buying 32GB of slow ram for gaming is...well...ludicrous and screams you have no idea what you are doing.

100% agree. As my benchies showed above, there is a difference when running faster ram. In my case i run 32GB because most of what i do nowdays needs loads of ram. But i do like benching still and having 32GB of fast ram still gets me very good bench results, as a look on any of the artificial benchies here https://forums.overclockers.co.uk/threads/links-to-benches-owners-threads.18774374/ will prove.

Peeps still, after years and years of proof on HWbot and god knows how many benching sites, have no comprehension of how useful and faster tight timmings and fast ramspeed is in everyday use................let alone benching.
 
I hardly play games find them a bit boring to be honest hence why I only have an RX 580 instead of a much better card. I just like benching my machine with 3D Mark. Mostly just use my PC to browse the web, it's a bit of a waste but like I say enjoy benching it with 3D Mark
 
I hardly play games find them a bit boring to be honest hence why I only have an RX 580 instead of a much better card. I just like benching my machine with 3D Mark. Mostly just use my PC to browse the web, it's a bit of a waste but like I say enjoy benching it with 3D Mark

I did actually laugh out loud at this. If I was one of the Billy Goats Gruff there's no way I'd be crossing a bridge tonight!
 
I did actually laugh out loud at this. If I was one of the Billy Goats Gruff there's no way I'd be crossing a bridge tonight!

It's the truth, I try and play BF4 and give up after about 10 minutes, that said I'll probably buy BF5 and play it for about 5 hours then give up and not load it again for months. That said I'm not a Luddite and to tight to upgrade my PC even if I know I don't really need it, that said I don't upgrade every gen for 5% more performance.
 
I'm not a troll if that's what you are implying. I simply enjoy discussing PC stuff on this forum, I'll probably dissapear again for years soon when I lose interest in this forum again.
 
I hardly play games find them a bit boring to be honest hence why I only have an RX 580 instead of a much better card. I just like benching my machine with 3D Mark. Mostly just use my PC to browse the web, it's a bit of a waste but like I say enjoy benching it with 3D Mark

32GB of ram and you mainly surf the web. :eek:

OP, sell the 32GB ram that you clearly don't need for benchmarking, a little gaming and mostly web browser and get 16GB of 3200mhz 8pack CR14 memory in there to optimise your setup.
Sod the cost,just do it :D. It's currently discounted to £199.99 from £218.99.

I was thinking he could just sell 16GB and keep the rest. Unless he wants much fast ram for some reason. :confused:
 
32GB of ram and you mainly surf the web. :eek:

Well I noticed the other day with Edge loaded with a number of tabs open I was using 6.2GB, then I loaded a VM and was using 7.6GB. Having 32GB means I can load BF4 as well without closing programs but to be fair I used to be able to do that with 16GB.
 
Well I noticed the other day with Edge loaded with a number of tabs open I was using 6.2GB, then I loaded a VM and was using 7.6GB. Having 32GB means I can load BF4 as well without closing programs but to be fair I used to be able to do that with 16GB.

As you have been here since 2007 let's assume you are not a troll......................ok, i accept you are not a troll................that's done with then. So the only other outcome that really springs to mind is that you are stupid, because no sane person could come to any other conclusion.
 
As you have been here since 2007 let's assume you are not a troll......................ok, i accept you are not a troll................that's done with then. So the only other outcome that really springs to mind is that you are stupid, because no sane person could come to any other conclusion.

Yeah I'm not the sharpest tool in the shed, below average IQ. I'm probably a bit insane as well. But at least I don't take myself to seriously.
 
Back
Top Bottom