Resolve Edit PC: Is this a balanced spec for £1500?

I'm waiting on my pc components to arrive but I tried to put together a balanced system for Resolve. Here is what I ordered. I'm just wondering if I'll have buyers remorse on the GPU.
CPU Ryzen 3900x
Mb MSI MPG X570 GAMING PRO CARBON WIFI
Memory Ballastic 32 gig 3200mh cl 16
GPU EVGA 1080TI SC Black 11GB
Power supply EVGA 700 Gold
SSD's (2) XPG SX8200 Pro 1TB M2 NVME
I have a full tower case and I'm using the Noctua D15 cpu cooler
I have (2) four tb segate hard drives for backup

I was ordering the 2070 super but found a nice deal on the 1080TI I heard Resolve loves V-Ram but I'm not sure if this was the best choice.

Thanks for sharing your system spec. It certainly helps me with my decision making.

After doing a little more research I changed a few components. The Mb, I changed to the AsRock Taichi. I did this as Micro Center is running a promo if you buy the CPU and MB there they discount $50. While there we measured the CPU cooler I'm planning on using to see if there was room to mount it and not block the first port for the GPU. It looks like it will be very tight but should still work. The tech there said if not the second port would be fine and I would not notice any performance drop. The Taichi has more features that should work better for Resolve. It has (3) M2 ports all using PCIe 4. So as others have suggested using a ssd for a scratch disk and another for Resolve should speed things up. AsRock has a helpful web site to find compatible RAM. Unlike most other 570 boards the Taichi does not Daisey Chain their memory; so you want to fill all four slots for the best stability and hopefully best performance; this was a note on the AsRock memory capability page for this board. One memory kit they suggested seems a nice balance price to performance it is the CORSAIR Vengeance LPX 32GB 3600MH kit ( 4- 8Gb DIMMs ), this also will fit under the Noctua D15 CPU cooler as it is low profile. The Taichi also has a Thunderbolt connecter on the board that I believe you have to use their proprietary card. In short this card has a few bells and whistles that should work better for content creators. I'm glad the original items were on backorder so I could change before they shipped. I'm not an expert by any stretch but if your interested check out the AsRock website and several YouTube reviews for yourself on the AsRock board. Or wait a little longer till the dust settles and more reviews and details arrive along with updated drivers.

Hi All

I plan to build my first Ryzen based edit PC in the coming weeks.

Having read the configuration guide, these forums, Puget Systems and too many YouTube videos, I'm still unsure on the best way to maximise my available budget.

I aim to edit HD and UHD footage from various Olympus and Panasonic m43 cameras to produce videos of no more than 30 minutes duration. Completed edits will be outputted to DVD, USB sticks and YouTube in HD and UHD.

Given a total budget of £1500, what tweaks would you recommend to this spec which comes to £1443 on PC Part Picker?

CPU: Ryzen 2700X
CPU Cooler: Noctua NH-D15 SE-AM4
Motherboard: ASRock X470 Taichi
Memory: Corsair Vengeance LPX 32 GB (2 x 16 GB)
Graphics: Sapphire Vega 56 8GB Pulse
Storage: Corsair - MP510 960GB System drive
Storage: Corsair - MP510 960GB Scratch drive
Storage: An existing 4TB drive
Case: Fractal Design Meshify C
Power Supply: EVGASuperNOVA T2 1000 W 80+ Titanium

Thank you for your assistance.

Shaun

Shaun, you may be able to save some money by changing the CPU to the newer Ryzen 3800x and using the stock cooler that comes with it. Also the power supply could easily downsize a bit as a 700W would be more than enough. I like our choice of motherboard but with what you save on the cooler and power supply you could get the 570 Taichi. Good luck with your build!

Shaun, you may be able to save some money by changing the CPU to the newer Ryzen 3800x and using the stock cooler that comes with it. Also the power supply could easily downsize a bit as a 700W would be more than enough. I like our choice of motherboard but with what you save on the cooler and power supply you could get the 570 Taichi. Good luck with your build!

Thank you Al. You've been a big help as has everyone else who has contributed to this thread.

I plan to go all in on the 3900X with stock cooler and a small power supply. Yes, it blows my original budget, however, I believe it hits the point of diminishing returns for me. Hopefully, Amazon's Prime Day deals will prove helpful. If not, then c'est la vie! I'll buy, build and be happy.

Thank you Al. You've been a big help as has everyone else who has contributed to this thread.

I share this sentiment. Thank you all for your support. I will research what motherboard is the right one for me, and post my list in the coming days.

Shaun, you may be able to save some money by changing the CPU to the newer Ryzen 3800x and using the stock cooler that comes with it. Also the power supply could easily downsize a bit as a 700W would be more than enough. I like our choice of motherboard but with what you save on the cooler and power supply you could get the 570 Taichi. Good luck with your build!

Thank you Al. You've been a big help as has everyone else who has contributed to this thread.

I plan to go all in on the 3900X with stock cooler and a small power supply. Yes, it blows my original budget, however, I believe it hits the point of diminishing returns for me. Hopefully, Amazon's Prime Day deals will prove helpful. If not, then c'est la vie! I'll buy, build and be happy.

All compute power starts with a good PSU.

You imply that Scratch uses an other de-bayer algorithm then RED and Davinci Resolve do with RED material. Even if scratch can decode 100.000 fps in realtime, the GPU still has to de-bayer.

I'm implying no such thing. You're basing that assumption on the illogical assumption that the deBayer algorithm is the only variable involved, which is clearly not the case.
[/quote]

Resolve is also pretty fast with 8k.R3D half-res good when you have atleast 8 GB of VRAM. RX590 is fast enough for real-time 25 fps editing.

Yet not nearly as fast as Scratch. I like Resolve quite a bit, but Scratch dances circles around it.

From what I've gathered so far, the mainboard will be fine as long as it is x570 and the feature set covers personal needs. So whether I buy a cheaper x570 mainboard modell or something from the high end spectrum won't have that much of an impact, as long as I'm satisyied with the feature set?

From what I've gathered so far, the mainboard will be fine as long as it is x570 and the feature set covers personal needs. So whether I buy a cheaper x570 mainboard modell or something from the high end spectrum won't have that much of an impact, as long as I'm satisyied with the feature set?

Hi.

You are right. Many high end X570 motherboards are build for Gamers that like to overclock the CPU. For Resolve will it be better to choose a CPU with more cores, as in Resolve the CPU is used to run the app, disk I/O and compression and decompression of codecs. Resolve does all its image processing in the GPU on the graphics card. And higher CUDA/OpenCL performance are better.

And some codec, resolution and bit width can be hardware accelerated in either nVidea or AMD GPU's or in some Intel CPU's, but only for the STUDIO version of Resolve. Then can you choose a CPU with less Performance.

Regards Carsten.

and be aware than many OFX wil either not load -or- run very very slowly on CPU if you chose AMD video cards

Do you have more details on this?

You imply that Scratch uses an other de-bayer algorithm then RED and Davinci Resolve do with RED material. Even if scratch can decode 100.000 fps in realtime, the GPU still has to de-bayer.

I'm implying no such thing. You're basing that assumption on the illogical assumption that the deBayer algorithm is the only variable involved, which is clearly not the case.

Resolve is also pretty fast with 8k.R3D half-res good when you have atleast 8 GB of VRAM. RX590 is fast enough for real-time 25 fps editing.

Yet not nearly as fast as Scratch. I like Resolve quite a bit, but Scratch dances circles around it.[/quote]

With RAW material it starts with the de-bayer speed/quality, the rest follows.

Quick question

I understand 3600MHz C16 and 3200MHz C14 RAM hit the sweet spot for Ryzen 3000 processors. They are proving difficult to acquire at a reasonable price. Would 3200MHz CL16 RAM be too big a step down?

Kind regards,
Shaun

Hi.

The CAS Latency for DDR4-3200 CL 14 is 8,75 and the CAS Latency for DDR4-3200 CL 16 is 10,00. So the difference is 14,3%

Regards Carsten.

Hi.

The CAS Latency for DDR4-3200 CL 14 is 8,75 and the CAS Latency for DDR4-3200 CL 16 is 10,00. So the difference is 14,3%

Regards Carsten.

Thank you Carsten. Would this equate to a 14.3% drop in real world performance?

Would this equate to a 14.3% drop in real world performance?

Hi.

In Resolve the CPU is used to run the app, disk I/O and compression and decompression of codecs.
Resolve does all its image processing in the GPU on the graphics card.

This only affect CPU performance.

Regards Carsten.

Would this equate to a 14.3% drop in real world performance?

If you're asking whether slower DRAM will lead to a 14.3% drop in the real world performance of Resolve, the answer is almost certainly "no".

The likelier figure is 0%.

Hello,

The RAM is a row / column matrix:

-The CL is the latency to read / write on a given
-the tRCD is the latency to select a column
-the tRP is the latency to activate a line
-the tRAS is the minimum time between the activation of a line and the beginning of the activation procedure of the next line.

Clearly, these 4 timings are different measures, even if the Latency CAS (the first value) is the most important because is applied to each operation on the memory without exception.

If this build is well balanced, I'm goining with this:

CPU: AMD Ryzen 7 3700X € 349
Motherboard: ASUS TUF GAMING X570 € 237,90
Memory: G.Skill DIMM 32 GB DDR4-3600 Kit € 144,90
Storage: Corsair Force MP600 1 TB € 234,90 (System)
Video Card: RTX 2060 € 349
Case: Used
Power supply: Used

Any suggestions?

EDIT: That would be 1315.70€ total

Try to get a GPU with atleast 8 GB of VRAM.
Use the NVMe drive for scratch and get a cheap SSD (Sata or NVMe) as system drive (256...512 GB should be enough).
You might also need some storage drives.

Try to get a GPU with atleast 8 GB of VRAM.

I chose the 2060, because I want to upgrade later on, when pcie 4.0 cards are available. Will I get by with a 2060 working on 1080p material on the free version of davinci?

Use the NVMe drive for scratch and get a cheap SSD (Sata or NVMe) as system drive (256...512 GB should be enough).
You might also need some storage drives.

Okay, that's what I'm going to do. Thanks!

EDIT: Is a GPU even advantagous using the free verion of davinci?

I have no problems doing 1080p work with the older GTX1060 so I would expect you'll be just fine with the RTX2060 and I'm using the Studio version which places even more reliance on the GPU than the free version.

Try to get a GPU with atleast 8 GB of VRAM.

I chose the 2060, because I want to upgrade later on, when pcie 4.0 cards are available. Will I get by with a 2060 working on 1080p material on the free version of davinci?

Use the NVMe drive for scratch and get a cheap SSD (Sata or NVMe) as system drive (256...512 GB should be enough).
You might also need some storage drives.

Okay, that's what I'm going to do. Thanks!

EDIT: Is a GPU even advantagous using the free verion of davinci?

You can get a PowerColor - Radeon RX 590 8 GB RED DRAGON Video Card at mindfactory.de for € 180

Spend the money you saved on the studio version of Resolve (it's worth the money).

Try to get a GPU with atleast 8 GB of VRAM.

I chose the 2060, because I want to upgrade later on, when pcie 4.0 cards are available. Will I get by with a 2060 working on 1080p material on the free version of davinci?

Use the NVMe drive for scratch and get a cheap SSD (Sata or NVMe) as system drive (256...512 GB should be enough).
You might also need some storage drives.

Okay, that's what I'm going to do. Thanks!

EDIT: Is a GPU even advantagous using the free verion of davinci?

You can get a PowerColor - Radeon RX 590 8 GB RED DRAGON Video Card at mindfactory.de for € 180

Spend the money you saved on the studio version of Resolve (it's worth the money).

I want to learn the basics with the free version first. I plan on getting the studio version, as soon as I am familiar with the software. What would be the drawback with the radeon RX 590 8gb?

I want to learn the basics with the free version first. I plan on getting the studio version, as soon as I am familiar with the software. What would be the drawback with the radeon RX 590 8gb?

See for yourself
https://videocardz.com/amd/radeon-500/radeon-rx-590

https://videocardz.com/nvidia/geforce-20/geforce-rtx-2060

Cons: RX590 uses more power and has slower memory
Pros: RX590 has more compute power, more memory and is €170 cheaper.

I want to learn the basics with the free version first. I plan on getting the studio version, as soon as I am familiar with the software. What would be the drawback with the radeon RX 590 8gb?

See for yourself
https://videocardz.com/amd/radeon-500/radeon-rx-590

https://videocardz.com/nvidia/geforce-20/geforce-rtx-2060

Cons: RX590 uses more power and has slower memory
Pros: RX590 has more compute power, more memory and is €170 cheaper.

Yes, but am I not missing the advantages of the intel cards?

I want to learn the basics with the free version first. I plan on getting the studio version, as soon as I am familiar with the software. What would be the drawback with the radeon RX 590 8gb?

See for yourself
https://videocardz.com/amd/radeon-500/radeon-rx-590

https://videocardz.com/nvidia/geforce-20/geforce-rtx-2060

Cons: RX590 uses more power and has slower memory
Pros: RX590 has more compute power, more memory and is €170 cheaper.

Yes, but am I not missing the advantages of the intel cards?

Intel has no cards.

Pros: RTX2060 uses less power and has faster memory
Cons: RTX 2060 has less compute power, less memory and cost €170 more.

Thanks for the information! This is what my list now looks like:
CPU: AMD Ryzen 7 3700X € 349
Motherboard: GIGABYTE X570 AORUS ULTRA € 299
Memory: G.Skill DIMM 32 GB DDR4-3600 Kit € 144,90
Storage: Corsair Force MP600 1 TB € 234,90 (scratch)
LiteOn MU X1 512 GB, Solid State Drive € 72,90 (system)
+ Seagate ST8000DM004 8 TB(SATA 6 Gb/s, 3,5")€ 199,90 x2 (archive)
Video Card: ASRock Radeon RX 590 Phantom Gaming X 8G OC € 199,90
Case: Used
Power supply: Used

I'm planning on buying these parts in two weeks.

EDIT: Total is about 1700€. More than I planned on spending, but thats okay, if it works properly.

Cons: RTX 2060 has less compute power [than rx590]

The would-be buyer has already expressed impatience with my contributions to this thread, so I'm not addressing him, but how did you reach that conclusion?

Generic video benchmark score (higher is better):

Radeon rx590 --9,358
rtx 2060 - 12,938
RTX 2060 Super - 13,239

https://www.videocardbenchmark.net/high_end_gpus.html

Game-specific comparison:
+40% performance gain on average, in favor of 2060
https://gpu.userbenchmark.com/Compare/N ... 4034vs4033

Comparative Resolve performance: unknown.

LiteOn MU X1 512 GB, Solid State Drive € 72,90 (system)

Hi.

Yes you are right. The System SSD don't need to have the same speed or the same size as the scratch SSD:

But I don't understand your choice of SSD. It only have 3 years of warranty and I can't find any information on its expected Write Endurance. Not even on Lite-ON's webpage:

https://www.liteonssd.com/product/consumer/mu-x1

When In look at this Best SSDs 2019 link:

https://www.tomshardware.com/reviews/be ... ,3891.html

Do I find the top Adata XPG SX8200 Pro 512 GB with Warranty 5 years and a Write Endurance of 320TB/0.34 DWPD

You can buy it in Germany from 75,20 €. But I will suggest you find a shop you like.

Regards Carsten.

and be aware than many OFX wil either not load -or- run very very slowly on CPU if you chose AMD video cards

Do you have more details on this?

I've never heard of this before this post, and I've been running systems with both Nvidia and AMD cards. That said, I actually encountered this problem today while grading dailies on set, using my new 3700X / RX 5700 XT system.

I've found the 5700 XT to be a good performer in Resolve, especially for the money. However, I've been having problems with thumbnail rendering errors and some H264 material not displaying properly in both Resolve and Scratch. After talking with people at Assimilate and AMD, I'm placing my bets on this being an AMD driver issue, but one that is specific to the 5700 XT, because when I replaced it with an RX 580 to see if the problem would reoccur, it didn't; the 580 worked flawlessly. It's just the 5700 XT that has this problem. I thought it was just H264 footage, but today when I tried to apply the Glow effect on a clip, I got a bad anomaly in the frame which appeared not only in the preview, but also the render. See attachments below for examples of the problems I'm having right now.

Other than this issue, the 5700 XT is proving to have a lot of bang for the buck. Right now I'm working on set of a pilot shooting on Varicam in 4K DCI AVC Intra codec, and there are no problems whatsoever with playback, scrubbing as fast as I want to, and on final render with corrections out to 2K DCI files, I'm getting c. 73fps. Not too bad for a 400-dollar card.

so I'm not addressing him, but how did you reach that conclusion?

He provided the links. This is raw out of the box processing power, so the 590 technically has more. However, this does not take into account any overclocking or how effectively software utilizes that power.

https://videocardz.com/amd/radeon-500/radeon-rx-590
RX 590 : 6,769 GFLOPS (FP32)

https://videocardz.com/nvidia/geforce-2 ... e-rtx-2060
RTX 2060 : 5,242 GFLOPS (FP32)

The Radeon "technically" has more? But it performs about 40% worse, consistently, on a range of benchmarks?

I would have thought that a "pro" for the 2060 would be that it performs much better than the Radeon. But I guess not....

I've never heard of this before this post, and I've been running systems with both Nvidia and AMD cards. That said, I actually encountered this problem today while grading dailies on set, using my new 3700X / RX 5700 XT system.

I've found the 5700 XT to be a good performer in Resolve, especially for the money. However, I've been having problems with thumbnail rendering errors and some H264 material not displaying properly in both Resolve and Scratch.

Hi.

Can you Please confirm what Mario Kalogjera wrote in an another thread:

Same happens here with RX 580 and R16, no sign of this in R15 AFAIK.

I will not ask you to go back to Resolve ver 15. But can you please confirm that you actual are on Resolve 16 Beta?

Regards Carsten.

I would have thought that a "pro" for the 2060 would be that it performs much better than the Radeon. But I guess not....

Performs better on generalized benchmarks, and gaming benchmarks, and neither of those are anything like how resolve uses a gpu.......

For example 2080 ti vs Radeon 7

The 2080 ti wipes the mat with the Radeon 7 in generalized benchmarks.
https://gpu.userbenchmark.com/Compare/N ... 4027vs4035


In resolve 15 it very much depends on exactly what you are doing!
https://www.pugetsystems.com/labs/artic ... ance-1382/

Wow, so two RTX2080ti's are only 50% faster than a single RTX2060?

On a "bang for your buck" ratio, that makes the RTX2060 outstanding value!

From my own tests, overclocking rx 580 GPU brings almost no performance gain in resolve, but overclocking VRAM brings 2-3 fps in Candle benchmark. I am very inclined to conclude that 3D and graphics performance are of very little relevance when it comes to resolve, but memory bandwidth/performance and number of cores are, just like in mining scenarios.

However, there are reports that rx 5700 series, as it is now, being gaming-centric, has similar OCL performance to rx 580. That's why I'd like to see those Candle benchmark results.



Sent from my GM 5 Plus d using Tapatalk

Wow, so two RTX2080ti's are only 50% faster than a single RTX2060?

Hi.

As Dan Wrote:

'In resolve 15 it very much depends on exactly what you are doing!'. Here is an another slide from the same slidedeck, showing 'Basic + 4 Power Windows + 3 OpenFX + TNR'



But I will like to suggest you to go the the link yourself and read Pugets comments. Here is a link:

https://www.pugetsystems.com/labs/artic ... ance-1382/

Regards Carsten.

The would-be buyer has already expressed impatience with my contributions to this thread, so I'm not addressing him, but how did you reach that conclusion?

Just to be clear, I am thankful for your contribution.

I find the top Adata XPG SX8200 Pro 512 GB with Warranty 5 years and a Write Endurance of 320TB/0.34 DWPD

You can buy it in Germany from 75,20 €. But I will suggest you find a shop you like.

That sounds like a good option. I will go with that one. Not sure what choice in regards of the GPU would be optimal. I so far have considered the RTX2060, 5700 XT and RX590.

CPU: AMD Ryzen 7 3700X € 349
Motherboard: GIGABYTE X570 AORUS ULTRA € 299
Memory: G.Skill DIMM 32 GB DDR4-3600 Kit € 144,90
Storage: Corsair Force MP600 1 TB € 234,90 (scratch)
top Adata XPG SX8200 Pro 512 GB € 75,20 (system)
+ Seagate ST8000DM004 8 TB(SATA 6 Gb/s, 3,5")€ 199,90 x2 (archive)
Video Card: ASRock Radeon RX 590 Phantom Gaming X 8G OC € 199,90 (???)
Case: Used
Power supply: Used

That sounds like a good option. I will go with that one. Not sure what choice in regards of the GPU would be optimal. I so far have considered the RTX2060, 5700 XT and RX590.

If you are (apparently) prepared to spend €400 for a GPU, I'd say rather go for the RTX 2060 Super since it has 8 GB VRAM, and should have probably about the same performance and price as 2 x RX590 (and you get HEVC 10-bit encoding). RX 5700XT is still an enigma with Resolve.

That sounds like a good option. I will go with that one. Not sure what choice in regards of the GPU would be optimal. I so far have considered the RTX2060, 5700 XT and RX590.

Hi.

You gave the Forum two weeks to find out. I expect the Forum to give you two GPU's to choose between. And may be AMD will solve the MP4 problem with a new driver before the end of the two weeks.

But at the same time do I like to begin a discussion on if the AMD Ryzen 7 3700X to € 349 is the right CPU for you, and will it be an good idea for you to later upgrade to the Studio version of Resolve.

In Resolve the CPU is used to run the app, disk I/O and compression and decompression of codecs.
Resolve does all its image processing in the GPU on the graphics card. And higher CUDA/OpenCL performance are better.

And some codec, resolution and bit width can be hardware accelerated in either nVidea or AMD GPU's or in some Intel CPU's, but only for the STUDIO version of Resolve. Then can you choose a CPU with less Performance.

I wonder why none have asked you for the codec, resolution and bit width of your source videos?
May be you later will be able to take advantages of hardware acceleration. May be not.

I have two initial inputs to what CPU we want to recommend for this build. What do other forum members think?

From 10:24 they start to talk about DaVinci Resolve. Please stop the YouTube Video a 10:30 and look for the differences between the Ryzen 7 3700X + 2080Ti and the Ryzen 9 3900X + 2080TI.
I will like to hear, why there are this difference. Is it ONLY difference in encoding speed? And do you think they are acceptable?

And my second input is Here:



And what AMD explain:

'Concerning AIDA 64, we didn’t talk about it in the call, but one of the optimizations brought by Zen 2 is the reduction of the writing bandwidth from a CCD to the IOD from 32B/cycle to 16B/cycle when the writing bandwidth remains fully provisioned at 32B/cycle. Since workloads have little writing, the link does not need to be 32B wide. This choice of design makes it possible to optimize the surface area used but also the consumption to concentrate innovation efforts in other parts of the architecture. In other words, with the 3700X and only one chiplet the observed behavior is normal (and on 3900X with two chiplets you will logically observe higher theoretical writing results).'

From this German link translated by Google Translate:

https://translate.google.com/translate? ... 7-3700x%2F

How many of you think that the slower performance of the Ryzen 7 3700X in the Youtube, can be related to the difference in DDR4 memory bandwidth between Ryzen R9 3900X and R7 3700X?

Regards Carsten.

Candle benchmark

Hello

Do the test yourself rather than refer to external sources ...

You download the project (compatible V15) and the clip that goes with it. You just have to make a relink of the clip.

scr10_20140123 _UHD_v15.drp
https://drive.google.com/open?id=1C_UJR ... 3qr7-jQE56

Tutorial - Tracking_UHD.mov
https://drive.google.com/open?id=1jxgLc ... 2Jxwm3qNaL

After you go in the color page and you do the test.
CTRL + N to change the next test
CTRL + B to change the previous test.

It's a UHD version. Just a little more stressful than the HD

@Jean Claude: Ok. But can you provide the link to the comparison results? There are only results for the HD version around.

Here are results for the puny little RX 580 (no fps restriction in project settings):

09 Blur = 20 fps
18 Blur = 10 fps
30 blur = 6 fps
66 Blur = 3 fps

1 TNR = 30 fps
2 TNR = 15 fps
3 TNR = 8 fps
6 TNR = 5 fps

Performs better on generalized benchmarks, and gaming benchmarks, and neither of those are anything like how resolve uses a gpu.......

For example 2080 ti vs Radeon 7

Okay, I think I finally understand. Because a different AMD GPU, at 5x the price, performs slightly better in Puget Resolve benchmarks than generic benchmarks might suggest, that proves that the 590 is a better performer than the 2060, despite all evidence to the contrary.

This imaginative approach to equipment evaluation may not satisfy everyone, but it has one uncontested advantage: claims don't have to be proven.

Okay, I think I finally understand. Because a different AMD GPU, at 5x the price, performs slightly better in Puget Resolve benchmarks than generic benchmarks might suggest, that proves that the 590 is a better performer than the 2060, despite all evidence to the contrary.

Your reasoning skills seem to be lacking here. The comparison had nothing to do with suggesting the 590 is a better or worse card than the 2060. The comparison was to show that just looking at whatever benchmark you can get your hands on can be very misleading.

Wow, so two RTX2080ti's are only 50% faster than a single RTX2060?

Pugets, scoring system doesn't work like that.

You gave the Forum two weeks to find out. I expect the Forum to give you two GPU's to choose between. And may be AMD will solve the MP4 problem with a new driver before the end of the two weeks.

I can wait a bit longer if it helps clearing things up.

And some codec, resolution and bit width can be hardware accelerated in either nVidea or AMD GPU's or in some Intel CPU's, but only for the STUDIO version of Resolve. Then can you choose a CPU with less Performance.

Since I'm an absolute beginner, diving in head first with the Studio version, might not be necessary. I wonder if I can save some money with a cheaper GPU in order to buy Studio with a 4.0 PCIe GPU later on.

I wonder why none have asked you for the codec, resolution and bit width of your source videos?
May be you later will be able to take advantages of hardware acceleration. May be not.

I use a Canon D1 Mark IV, and the codec is H.264 AVC. More information can be found here:
https://www.imaging-resource.com/PRODS/ ... 4VIDEO.HTM
Audio will be recorded via condensor mic with channelstrip in cubase.

@Jean Claude: Ok. But can you provide the link to the comparison results? There are only results for the HD version around.

Here are results for the puny little RX 580 (no fps restriction in project settings):

09 Blur = 20 fps
18 Blur = 10 fps
30 blur = 6 fps
66 Blur = 3 fps

1 TNR = 30 fps
2 TNR = 15 fps
3 TNR = 8 fps
6 TNR = 5 fps

Hello,
For me: correct results with RX 580 for UHD. So not a puny little GPU.
It will surprise me that it is beaten by everyone with Asus Prime X370-Pro + Ryzen 1600X @ 4GHz, even with the new hardware ... Let's wait ..

@cremedouble: since you don't wish to go Studio, and thus none of the video encode strengths of NVidia cards over AMD apply to your situation, the RX 590 will be pretty nice for you considering the price, even more so because your source media is h.264 8-bit HD, that's also rather easy on the CPU you are planning to get.

.../...

I wonder why none have asked you for the codec, resolution and bit width of your source videos?
May be you later will be able to take advantages of hardware acceleration. May be not.

I use a Canon D1 Mark IV, and the codec is H.264 AVC. More information can be found here:
https://www.imaging-resource.com/PRODS/ ... 4VIDEO.HTM
Audio will be recorded via condensor mic with channelstrip in cubase.

without source clip example, difficult to judge. Place a link to download a small piece of clip.

without source clip example, difficult to judge. Place a link to download a small piece of clip.

There are examples in his link, at the end of the review.

without source clip example, difficult to judge. Place a link to download a small piece of clip.

There are examples in his link, at the end of the review.

Hello,
Where? I can not find in this long thread ... sorry.