You should receive a performance that matches your needs and expectations.
You should receive a performance that matches your needs and expectations.
Here’s a revised version of your text:
The specs are solid—5950x display, big BIOS update (v4902), aiming for minimal latency. It runs at 4.4GHz across all cores, with 1.220V-1.235V regulation, 100.00MHz clock speed, 1.1V SOC, and 1.33V on DRAM—digital tuning heavily, a bit undervolted but likely in the ideal range. No power-saving tricks are used; it stays at 4.4GHz all the time, RAM scrambling turned off, and it hits 1T with many tweaks. Despite what some forum users claim about FPS and SMT, there’s no clear link. When SMT is off, FPS doesn’t jump much, though a few people say gains of around 10%. Possibly the hardware isn’t up to the job, or maybe it’s just my setup. Benchmarks show scores dropping noticeably, so it’s not the dark horse here.
ASUS ROG Crosshair v8.0 is running on Wi-Fi, not the best performer. My motherboard has four 16GB Corsair RGB Pro 16GB RGB NVMe Pro cards, all in a 4x RAID setup—two for speed, two for cold storage. I have around 180TB total across about 18 drives, split between RAID 5 and RAID 1 for balance. The system stays cool under load, with temperatures hovering just above 4°C.
I’ve used a Samsung 980 Pro NVMe drives, OS storage, and games—both at 1TB each. I’ve got a mix of SSDs: two 2TB NVMe drives and six 1TB NVMe drives for gaming. The display is a 32:9 240Hz panel with 5120x1440 resolution, using the latest drivers. Windows 10 Pro v22H2 is running at 1945.544 MHz, which is solid.
I built this back when DDR4 was new and there were known issues. With the price jump to DDR5, sticking with DDR4 seemed smart financially. I didn’t skimp on hardware—if I upgraded later, I’d probably go all-in on DDR5. I also kept a lot of spare parts in my closet, so my 3 laptops running data centers have multiple SSDs and M.2 drives.
The server is a 2U rack unit packed with drives. It’s mostly for cold storage since it holds a lot of drives inside and outside. My current specs are around 28,886 MC and 1,667 SC. I usually let the system warm up to about 53°C before testing. The machine can run for hours without issues, sometimes through the day.
I don’t expect a perfect score unless tested on a typical daily setup. If I go ultra-light, maybe I’ll get around 35,000—just not a real-world benchmark. Overclocking and high temps would be risky. No bloatware except Armory Crate, which is debatable. Should I remove it?
For games like Forza Motorsport or Alan Wake 2, I see gains of about 20fps with other ray tracing options. Spiderman 2 can hit 60fps max, and Cyberpunk 2077 averages 45-50fps. Most games stay solid, rarely drop below 80fps unless I’m doing heavy background work.
I’ve tested many benchmarks across different titles, but results vary. Online play is rare, so most scores come from single-player or local play. If you want a more accurate picture, share your own numbers and game details.
I’m not sure if my setup meets average performance expectations, especially compared to other titles I’ve tested. If you’re considering a Windows 11 upgrade, let me know—I’m happy with Windows 10. Also, enterprise editions feel outdated compared to the current UI; switching might be worth it. But if you’re into Windows 11, I’d love to hear your thoughts!
Lastly, if you have a 4090 in NVIDIA control panel, does it support DSR (virtual resolution)? From what I’ve seen, most users say no—those claiming otherwise usually use older cards.
I hope everyone has a chance to read this carefully. I attempted but gave up. Stay safe and have a great try. Decoding this will be tough!
Well I ran it through Chat GPT & got this: Quick Specs Overview: Processor : AMD Ryzen 9 5950X BIOS : Massive overhaul (v4902), latency as low as possible across the board Clock & Voltage : 4.4 GHz all-core, 1.220V-1.235V, 100.00 bclk, 1.1V SOC, 1.33V on DRAM System Details: Digi+ Power Delivery : Heavily tuned, somewhat undervolted (in the sweet spot), with no power-saving features. 4.4 GHz all-core, RAM scrambling disabled, 1T timing, and many other tweaks. SMT (Simultaneous Multi-Threading) : Despite what others say on forums, I see no correlation between FPS and SMT. With SMT off, I don't notice any significant FPS increase, while some claim up to a 10% gain. Maybe their GPUs are the limiting factor, but with my BIOS setup and hardware, I don’t see any noticeable change, especially in benchmarks. In fact, my scores drop significantly with SMT off. Hardware: Motherboard : ASUS ROG Crosshair VIII X570 (not the Dark Hero version) RAM : 4x 16GB Corsair Vengeance RGB Pro (64GB total) at 3600MHz. If I had known that G.Skill RAM is the best, I would have gone with that instead. Maybe on my next build (if this one ever dies, which I doubt). Storage : 2x Samsung 980 Pro NVMe SSDs (1TB each) for OS and game storage 6x Samsung QVO or EVO SSDs (1TB each), used primarily for game storage, in various RAID configurations (RAID 0 for performance, RAID 1 for cold storage) A separate server in the basement houses a ton of drives, totaling about 180TB in RAID 5 across 18 drives for actual cold storage. Graphics Card : ASUS ROG Strix 4090 “OC” at 2835 MHz GPU / 10501 MHz memory. I had it overclocked to 3000 MHz / 11000 MHz, but for baseline purposes, it generally runs at stock frequencies. Cooling : Custom watercooling setup, overkill but functional! The system holds just over 4L of water. Display: Monitor : Samsung G9 49” 32:9 display, 240Hz, 5120x1440 resolution Software : Running Windows 10 Pro, version 22H2 (build 1945.5440), with the latest drivers and firmware installed. Performance Observations: All games are run at native resolution, with in-game settings cranked to max. Windows is not in "ultra-light" mode, as I’m running a typical daily usage scenario (ICue, various background apps, and some services disabled). I believe scores taken from a “light mode” setup don’t reflect real-world performance. Overclocking and temperatures can be affected by such a setup, potentially causing issues under normal usage. Benchmarks (3DMark, Heaven, etc.): R32 Average Score : 28,886 MC 1667 SC These scores were taken after the machine warmed up to ~53°C, not immediately after a cold boot. I ran benchmarks throughout the day, ranging from a single pass to extended runs of 5, 15, 30, or even 120+ minutes. These scores were also recorded with typical background applications running, like iCUE and a few others. Game Performance: Forza Motorsport (2023) : With RTX Global Illumination (RTGO), I hover between 65-100 FPS, with a 20 FPS increase when using the other ray tracing option. Spider-Man 2 : Never dips below 60 FPS, with highs between 120-140 FPS and an average around 100 FPS. Alan Wake 2 (latest update with DLSS3 and RTX 5000 optimizations): Frames vary: 45 FPS during intense NPCs, lighting, and shadows, 65-75 FPS otherwise. Rarely does it exceed 80 FPS. Senua’s Saga: Hellblade 2 : 45-50 FPS lows, 60 FPS average, rarely over 75 FPS. Cyberpunk 2077 : 45 FPS lows, 100-110 FPS highs, 70 FPS average. Other Games : Varies greatly, but generally 120-200 FPS on most titles. I don’t play online multiplayer, so all games are single-player or non-competitive. Performance Considerations: Stuttering : I don’t experience stuttering in any games unless I'm doing background tasks like RAR compression. Even then, the impact on gameplay is minor and not unplayable. This has been verified with various benchmarking tools and in-game benchmarking modes. Other Notes: I’m aware this is a DDR4 system. When I built this, DDR5 was new and had bugs, plus the price hike for new AM5/DDR5 setups didn’t seem worth it. If I were to upgrade anything, it would require replacing everything and moving to DDR5, which I don’t need right now. Server Setup : I have three laptops running in a data center setup, each with 2-3 SSDs plus M.2 drives. My server is a real 2U rack-mounted unit with a ton of drives crammed inside (and outside), totaling around 180TB in RAID 5. Benchmark Scores : I won't post detailed 3DMark or Heaven scores here for accuracy. I try to keep benchmarks consistent and run them under normal conditions (not ultra-light setups). Future Plans: I’m planning to reinstall Windows 10 soon, but I’m content with it. I don’t feel the need to upgrade to Windows 11. The interface in Windows 11 doesn’t appeal to me, and I prefer the consistency between Windows 10 and my data center environments. If Windows 11’s interface improves in future enterprise versions, I might consider the switch. Questions: Windows 11 : Should I upgrade to Windows 11? I don’t mind the new interface, but I'm happy with Windows 10 and its compatibility with my data center tasks. DSR on 4090 : Does anyone else with an RTX 4090 have the DSR (Dynamic Super Resolution) option in NVIDIA Control Panel? I’ve seen mixed reports online, and NVIDIA’s website mentions that DSR is not supported for 4xxx or 5xxx series cards. Does anyone have experience with this? Feel free to share your benchmarks, machine specs, and FPS numbers for comparison. I’m curious if my performance aligns with the average for my setup. Also, if you notice anything wrong with my configuration, please let me know! Thanks, everyone!
Access it directly through the NVIDIA control panel. According to my knowledge, all RTX cards include it.
It's perfectly fine using DDR4. You don't have to force yourself to avoid AM5.
Windows 10 will end in October. If Microsoft continues providing a free upgrade to Windows 11, you should consider it. But recently, reports suggest they are tightening the system requirements for Windows 11.
not listed in my preferences, and i've tested various drivers over the past couple of years while owning it—monitors, displayport, and even for laughs, hdmi