Insert_Nickname
Diamond Member
Got to share this as it's an ignored issue, single rank vs dual rank tests
It is not exactly news though. AMDs memory controllers have always liked dual rank DIMMs. I'm not surprised at those results.
Got to share this as it's an ignored issue, single rank vs dual rank tests
Well, if they buy an R7-1800X instead of a 6900K, they can use the extra 600-700 dollars on faster storage.
That'll result in a greater boost to their productivity than any marginal PCIe effects.
It is not exactly news though. AMDs memory controllers have always liked dual rank DIMMs. I'm not surprised at those results.
When you consider the extra hardware (NVMe/PCIeSSD drives etc) that you can afford to use with an R7 vs. a 6900K, then its a non contest.
In other words it needs to be fast, period, and so the spending is there, period. You'll buy fast storage regardless of whether you're working on a Ryzen or an i7.
I don't really agree with that. You need storage that is fast enough to not be a bottleneck. Some bottlenecks you can live with, others not. I think it's arguable that if your machine allows you to use X number of effects with two GPUs, but only with fast storage, and without fast storage you get no streaming capability at all at the required media resolution, then the storage throughput becomes an irrelevant factor because it's a requirement. In other words it needs to be fast, period, and so the spending is there, period. You'll buy fast storage regardless of whether you're working on a Ryzen or an i7.
In which case you'll buy more than 1 machine. So better cost/performance still works for R7.
you're really going out of your way to try and justify paying 2x the price for nearly identical performance....OK, so let's say you put that savings somewhere else? How about fancier plugins for Avid or Premiere? Well, you're going to say that you will be purchasing those anyway...
OK, so why should no pros actually care about saving $500? What is the reason?
The people I'm talking about creating content aren't going to benchmark using Grand Theft Auto at 1080.
I don't think you understand how it works. Either that or I don't.
If I'm color correcting 4k footage with multiple nodes in Resolve then I won't be using multiple machines, I'll use one. Because the data rate is high I'll need fast storage regardless of the effects processing. Thus, fast storage is a must, one computer is the norm, and GPUs determine processing capacity.... and GPUs rely on lane-count....
If I'm color correcting 4k footage with multiple nodes in Resolve then I won't be using multiple machines, I'll use one.
I don't really agree with that. You need storage that is fast enough to not be a bottleneck. Some bottlenecks you can live with, others not. I think it's arguable that if your machine allows you to use X number of effects with two GPUs, but only with fast storage, and without fast storage you get no streaming capability at all at the required media resolution, then the storage throughput becomes an irrelevant factor because it's a requirement. In other words it needs to be fast, period, and so the spending is there, period. You'll buy fast storage regardless of whether you're working on a Ryzen or an i7.
If it's the number of PCI-E lanes that bother you the most, then just wait for Naples to release in H2.People keep ignoring the lane count. It's not irrelevant. It doesn't matter if your CPU performs at 80% of the top Intel CPU for 50% of the cost if you need lanes that the Ryzen doesn't provide. A requirement is a requirement.
I don't think you understand how it works. Either that or I don't.
If I'm color correcting 4k footage with multiple nodes in Resolve then I won't be using multiple machines, I'll use one. Because the data rate is high I'll need fast storage regardless of the effects processing. Thus, fast storage is a must, one computer is the norm, and GPUs determine processing capacity.... and GPUs rely on lane-count....
They do a blind test, and find that ryzen does feel more smooth, though the differences seem small.Has this been poseted yet?
Then if you were building a new rig today, 1800x would be a great platform for you. Why wouldn't it?
What does PCIe lanes have to do with anything. I am confused. Unless again you are talking about tri-sli/xfire..? Or FirePro stuff..?
Why would an average home user dump that kind of money into an Intel HEDT platform, that is going EOL soon..? When AM4 is just released..?
Is it an interactive process or one that is sent off into a code?
Since it should be sent to code, then having more machines = more productivity.
Anyone serious about such endeavours should be continuously developing their own code for corrections - otherwise as you say - they are doing it wrong and only think they are professional content creators.
And these people you are talking about dont believe in showing actual data to back up their assertions? Where is the data that shows how badly 4X or 8X PCIe is penalized?
I wasnt sure you were trolling untill this post. Now im sure.
Obviously if money doesnt matter and you have an unlimited budget and cost is of no consequence then sure a top end intel HEDT is the best option, most PCIe lanes and most cores(at least untill the new HEDT AMD chips launches).
But for the 99.999999999999% of customers who do not have a unlimited budget Ryzen offers the same performance for WAY less money.
The cost difference of R7 to 6950X is just too hard to ignore, the price perf is so far in AMD's favor that you had to go to the cost is no object intel is best argument just proves the point that AMD is offering way more for the money than intel.
You clearly have no idea what I'm talking about if you're suggesting users write their own code for this.
Last week it was the AVX2 guy, this week the HEDT guy and the thread gets hijacked for 3-4 pages just because 1 guy can't let it go and 10 others are feeding him.