Computer Upgrade

Comments

EricLNZ wrote on 4/11/2023, 8:54 PM

Amazon are in Australia so trying to buy from Amazon US may not work?

RogerS wrote on 4/11/2023, 10:13 PM

I import from Amazon US and B&H in the US when the prices are better. Amazon.co.jp has been in business for two decades but I don't have to use it. With small orders (one thing at a time) there usually aren't import duties.

Custom PC (2022) Intel i5-13600K with UHD 770 iGPU with latest driver, MSI z690 Tomahawk motherboard, 64GB Corsair DDR5 5200 ram, NVIDIA 2080 Super (8GB) with latest studio driver, 2TB Hynix P41 SSD and 2TB Samsung 980 Pro cache drive, Windows 11 Pro 64 bit

ASUS Zenbook Intel i9-13900H with Intel graphics iGPU with latest ASUS driver, NVIDIA 4060 (8GB) with latest studio driver, 48GB system ram, Windows 11 Home, 1TB Samsung SSD.

VEGAS Pro 21.208
VEGAS Pro 22.122

Try the
VEGAS 4K "sample project" benchmark (works with VP 16+): https://forms.gle/ypyrrbUghEiaf2aC7
VEGAS Pro 20 "Ad" benchmark (works with VP 20+): https://forms.gle/eErJTR87K2bbJc4Q7

Rednroll wrote on 4/12/2023, 4:25 PM

So, in your personal view, do you believe that the upgrade from the Ryzen 9 3900X to Ryzen 9 5950X, was a significant upgrade for you?

I know you weren't asking me since I don't have personal experience but I often find these things can be difficult to quantify. Personally, I really like to use the website Userbenchmark.com to try and put some quantifiable numbers behind making these types of decisions. They have a "compare" link where you can compare 2 CPUs, GPUs, SSDs, etc and it will display the typical cost and provide a performance percentage increase, as well as a overall ranking of every CPU on the market.

These numbers are from collated data from many users running benchmark tests with those components present in their system.

Here is a link to that website where I selected those 2 CPUs you were asking about.

https://cpu.userbenchmark.com/Compare/AMD-Ryzen-9-3900X-vs-AMD-Ryzen-9-5950X/4044vs4086

That tells me the 5950X cost about $165 more and for that $165 you can expect to get an average performance increase of +19%.

Ranking out of 1405 CPUs:

5950X=46th and 3900X=127th

You can then dig further down into other comparisons of how that average 19% performance increase was determined as well. With that it's then up for you to determine based on your budget. Is that 19% increase worth the extra $165?

As you mentioned earlier in the discussion, the price of these GPUs today are outrageous. When I made my recent laptop purchase I had to make a choice between the RTX-4070, RTX-4080(+$500) or RTX-4090(+$1000). Having no personal experience in really understanding what I'm getting in return for each of those +$500 price increases, I turned to the UserBenchmark.com website to assist me in making that decision.

That told me for the additional $500 I would spend over the 4070 for the 4080, I could expect an average performance increase of +55%.

https://gpu.userbenchmark.com/Compare/Nvidia-RTX-4070-Laptop-vs-Nvidia-RTX-4080-Laptop/m2033663vsm2037743

So then I compared the 4080 to the 4090. That comparison showed me I could expect an average performance increase of about +18%.

https://gpu.userbenchmark.com/Compare/Nvidia-RTX-4080-Laptop-vs-Nvidia-RTX-4090-Laptop/m2037743vsm2036852

So it seemed to me a +55% increase in performance was pretty significant and I was able to justify to myself, spending that additional $500 for the 4080 may be worth it. Spending another $500 on top of that for the 4090 just to have the top available option for another +18% increase in performance, didn't excite me as much and I therefore decided to go with the RTX-4080.

As mentioned earlier, there's typically a sweet spot with electronics of the best value of price vs performance. That website in many ways helps me better determine and quantify where that price to performance sweet spot is at when i start choosing between components for a system.

I'm an engineer, I like numbers and working with statistical data. That website provides me statistical data summaries and isn't based off of someone's personal opinion or marketing hype.

MelvinGonsalvez wrote on 4/12/2023, 5:26 PM

I was delayed in installing my upgrade to Vegas Pro 365 with VP 20 but am finally doing so at this moment. I recently upgraded my dual RTX 2080 Tis running in NVLINK to two RTX 6000 ADA video cards but unfortunately nVidia doesn't include that type of SLI connectivity for these high-end cards anymore so I'm running them individually on my Win 10 PC. Davinci Resolve recognizes both independent cards and tasks them for the render process but I don't know offhand if VP 20 does (or plans to soon) address both cards. These ADA cards have 48GB VRAM each but they get scaled back to 44GB or so each when ECC memory is enabled. Can't wait to see the performance upgrade with these running V20 as I only ran v17 previously as it was the most stable for me.

Sassylola wrote on 4/12/2023, 5:57 PM

https://www.tomshardware.com/news/rtx-4090-beats-rtx-6000-ada-in-content-creation-performance

My System Home Built

Intel 13900K Latest Intel Chipset. Latest Intel Management Engine Installed. NO OC (PL 1 Set to 253W PL 2 set to 253W)

Arctic Liquid Freezer ll 360mm AIO

Gigabyte Aero G Z790 MB Latest BIOS

ZOTAC GAMING GeForce RTX 4090 AMP Extreme AIRO Latest Nvidea Studio Drivers Installed. With GPU OC results were about the same as GPU set to default settings. I have kept GPU at default settings

64 GB G Skill DDR 5 6000mhz Ram 2X32

Corsair 1000W RMx Power Supply

SK Hynix P41 NVMe 1TB Operating Drive. ( Boot Drive )

WD 2TB SN 850X SSD 1st Render Drive NVMe

SK Hynix P 41 2 TB NVMe Storage Drive

2 TB Sk Hynix P 41 SSD 2nd Render Drive For final renders NVMe

Win 11 Pro all currant updates applied

Fractal Design R6 Case

Samsung - M7 Series 43" Smart Tizen 4K UHD Monitor

Vegas Pro Suite 22 Ver 122

1 Audigyfx Sound Card

SSK USB C External NVMe drive Enclosure with 500GB Samsung 970 EVO Plus for backups.

PROBOI 10G Hard drive Enclosure. USB C. 4 4TB Spinning Drives for Backups.

 

 

RogerS wrote on 4/12/2023, 10:45 PM

VEGAS doesn't address both cards and doubt distributed rendering is in the works any time soon. Hopefully there won't be conflicts with having two of the same card installed. Do try a benchmark or two (see my signature) and we'll see how it performs!

Custom PC (2022) Intel i5-13600K with UHD 770 iGPU with latest driver, MSI z690 Tomahawk motherboard, 64GB Corsair DDR5 5200 ram, NVIDIA 2080 Super (8GB) with latest studio driver, 2TB Hynix P41 SSD and 2TB Samsung 980 Pro cache drive, Windows 11 Pro 64 bit

ASUS Zenbook Intel i9-13900H with Intel graphics iGPU with latest ASUS driver, NVIDIA 4060 (8GB) with latest studio driver, 48GB system ram, Windows 11 Home, 1TB Samsung SSD.

VEGAS Pro 21.208
VEGAS Pro 22.122

Try the
VEGAS 4K "sample project" benchmark (works with VP 16+): https://forms.gle/ypyrrbUghEiaf2aC7
VEGAS Pro 20 "Ad" benchmark (works with VP 20+): https://forms.gle/eErJTR87K2bbJc4Q7

Wolfgang S. wrote on 4/13/2023, 1:25 AM

The only point is, that an i-GPU can be used for playback of long-GOP footage, while an RTX40xx can be used as overall GPU in the preferenes. But Roger is right, beside that Vegas does not utilize two GPUs.

Desktop: PC AMD 3960X, 24x3,8 Mhz * GTX 3080 Ti (12 GB)* Blackmagic Extreme 4K 12G * QNAP Max8 10 Gb Lan * Resolve Studio 18 * Edius X* Blackmagic Pocket 6K/6K Pro, EVA1, FS7

Laptop: ProArt Studiobook 16 OLED * internal HDR preview * i9 12900H with i-GPU Iris XE * 32 GB Ram) * Geforce RTX 3070 TI 8GB * internal HDR preview on the laptop monitor * Blackmagic Ultrastudio 4K mini

HDR monitor: ProArt Monitor PA32 UCG-K 1600 nits, Atomos Sumo

Others: Edius NX (Canopus NX)-card in an old XP-System. Edius 4.6 and other systems

Reyfox wrote on 4/13/2023, 3:42 AM

userbenchmark.com tests gaming benchmarks, which are different than productivity benchmarking. And the price listings are way off for the 5950x which can be had for $500, and they have been accused of extreme biasing towards Intel and Nvidia. This can be Googled and there is a lot of information.

I'd go to Anantech and other tech sites, and video sites such as Gamers Nexus, or PCWorld. Yes, these sites compare CPU's from that generation and past, the PCWorld video adds previous gen CPU's. They give a more well rounded review including productivity and gaming.

Last changed by Reyfox on 4/13/2023, 3:45 AM, changed a total of 2 times.

Newbie😁

Vegas Pro 22 (VP18-21 also installed)

Win 11 Pro always updated

AMD Ryzen 9 5950X 16 cores / 32 threads

32GB DDR4 3200

Sapphire RX6700XT 12GB Driver: 22.5.1, testing 24.7.1

Gigabyte X570 Elite Motherboard

RogerS wrote on 4/13/2023, 3:58 AM

Techgage is a good resource as they actually use VEGAS for productivity tests. https://techgage.com/

The next round of tests on the RTX 4070 should go beyond gaming.

Custom PC (2022) Intel i5-13600K with UHD 770 iGPU with latest driver, MSI z690 Tomahawk motherboard, 64GB Corsair DDR5 5200 ram, NVIDIA 2080 Super (8GB) with latest studio driver, 2TB Hynix P41 SSD and 2TB Samsung 980 Pro cache drive, Windows 11 Pro 64 bit

ASUS Zenbook Intel i9-13900H with Intel graphics iGPU with latest ASUS driver, NVIDIA 4060 (8GB) with latest studio driver, 48GB system ram, Windows 11 Home, 1TB Samsung SSD.

VEGAS Pro 21.208
VEGAS Pro 22.122

Try the
VEGAS 4K "sample project" benchmark (works with VP 16+): https://forms.gle/ypyrrbUghEiaf2aC7
VEGAS Pro 20 "Ad" benchmark (works with VP 20+): https://forms.gle/eErJTR87K2bbJc4Q7

MH7 wrote on 4/13/2023, 10:31 AM

userbenchmark.com…have been accused of extreme biasing towards Intel and Nvidia. This can be Googled and there is a lot of information.

You know, that did come to my mind but I wasn’t sure if it was true or not. I have been using some other sites to find out performance numbers. You can often even look up Cinebench R23 numbers and they can be helpful.

Last changed by MH7 on 4/13/2023, 10:32 AM, changed a total of 1 times.

John 14:6 | Romans 10:9-10, 13, 10:17 | Ephesians 2:8-9
————————————————————————————————————

Aussie DaVinci Resolve 19 User as of 1st September 2024
(I may upgrade to DaVinci Resolve Studio at some point in the near future).

Aussie VEGAS Post 20 User as of 9th February 2023 — Build 411 (Upgraded from VEGAS Pro 18)

VEGAS Pro Help: VEGAS Pro FAQs and TROUBLESHOOTING GUIDES

My YouTube Channel: https://www.youtube.com/@TechWiredGeek

Video Cameras: Sony FDR-AX700 and iPhone 12 Pro Max (iOS 17)

============================================

My New Productivity Workstation/Gaming PC 2024

CPU: AMD R7 7800X3D

Motherboard: ASRock X670E Steel Legend (AM5)

RAM: Corsair Vengeance 32 GB (2 x 16 GB) DDR5-6000 CL30 Memory

Main SSD: Samsung 980 Pro 1 TB SSD
Storage SSD: Western Digital Black SN850X 2 TB SSD

GPU: Asus TUF GAMING OC Radeon RX 7800 XT (16 GB)

OS: Windows 11 (Build: 23H2)

Main Monitor: LG 27UD88-W 4K IPS

Secondary Monitor: LG 27UL850 4K HDR IPS

MelvinGonsalvez wrote on 4/18/2023, 10:54 AM

@ Rednroll:

nVidia had their RTX 6000 ADA cards originally listed at $6800 while other vendors had a substantial markup over that price but nVidia offered a discounted promotional program (through their board partners for early adopters of this card) that ended on 2/15/23 where a supposed "50%" discount was offered so I was able to pick up two of these cards for $4800 each plus tax.

MelvinGonsalvez wrote on 4/18/2023, 11:04 AM

@ Sassylola:

Yes, I saw that earlier comparison between the RTX 4090 & the RTX 6000 ADA on Tom's Hardware and those performance results troubled me at the time but I still decided to go for the latter, lower performing card since I learned it was engineered to run 24/7/365 with its much lower wattage (300W), slower GPU clock speed, and lower clocked VRAM. I hope to see much better results when larger project databases take advantage of its 48GB of GDDR6 memory.

MelvinGonsalvez wrote on 4/18/2023, 11:28 AM

@ RogerS:

My recent VP20 simple render test consisted of several UHD 30fps clips from my Sony videocamera that were sequentially lined up on a timeline with basic crossfades between them. With nVidia enhancements chosen for the UHD output file, I observed my i9 9920X CPU (12 core, hyperthreaded to 24) running at around 65% usage, 15GB of my MB's 128 GDDR4 memory being utilized (while running in Win10 64-bit Pro.), GPU 0 (then running at 72 deg. C) was hitting around 60% usage with 3.5GB/44.5GB EEC VRAM being used, and GPU 1 (then reading 50 deg. C) was pegging between 85% - 100% with 2.1GB/44.5 EEC VRAM being utilized.

I don't understand offhand how VP20 has been programmed to take advantage of these dual RTX 6000 cards running independently on the MB without being connected with NVLINK but I'm happy to see them both used. They do however show being tasked with more processing power per GPU when the same files are rendered in Davinci Resolve.

My great relief though in this basic testing to date has been to finally see realtime timeline-playback in my viewport of 4K files, even at a "Best-Full" setting. My previous dual RTX 2080 TIs with NVLINK could not do it and even struggled with playing back FHD files at some lower-quality viewing resolutions.