Any PC Builders Out There?

Funny enough, the rumor is Microcenter sold more 7900XTX cards today than 5080 cards. Part of it is probably that folks who couldn't get a 5080 for $1300 or more saw the XFX 9700XTXs sitting for under $1K and went home happy, but the other part I think is that the 5080 disappointed even the hardcore nVidia gamers.

 

Holy crap finally beat the scalpers and bots to get a 5080.



Since I had to get a combo if anyone wants the PSU since I have one already $85 and it's yours.

 

Ya we have combos here that help keep the scalpers back, but I dont think it will work.

Glad you got one, I've only owned one Zotac but it was a good brand.

 

Newegg actually does this to offload dead inventory. Things people don't want to buy that have sat too long on the shelf. It came up when Newegg got caught by GN scamming returns.

 

Also I read a pretty interesting theory as to why the 50 series uplift is so small, especially since nVidia advertising would suggest much larger gains from the new hardware design.

So I'm not a massive nuts and bolts guy when it comes to hardware architecture. While I understand the sub components, I am no expert on why certain configurations would give increased performance. In other words, while it's easy to say more memory bandwidth is more better, I'm not able to give a technical explanation to why a pipeline change in the chip would have a bigger impact.

Anyway, the 5000 series made one absolutely massive change that nVidia hints around in marketing materials. The GPU scheduler is no longer controlled from the CPU. The 5000 series puts a RISC-V coprocessor on board that handles hardware GPU scheduling. This is also why, even if drivers existed, the 5000 series will not work on many Linux builds and Windows 7. Hardware coprocessor support is exclusive, for now, to Windows 10 and 11, and some Linux builds.

This coprocessor offers a potentially large decrease in latency between the scheduler and the beginning of the GPU pipeline, which nVidia calls the Gigathread. However, this only works when the CPU doesn't need anything from the GPU. If the CPU does try to access anything from the GPU, it has to go through the coprocessor, which has a massive latency hit.

So what? Well, every game is written a bit differently and often written so that the base game plays on all cards but some features are built on top. If the base game is not coded to take into account the seperation between the CPU and GPU, it causes a latency hit that kneecaps the card.

Why did nVidia use a coprocessor if it has disadvantages? First up, the whole neural face generation thing and most of the new AI driven background features would be impossible with software or CPU GPU scheduling, at least with the current GPU layout nVidia is using. The coprocessor is a component of the nVidia secret sauce, it's a major component that unlocked the GPU drastically more for AI compute when utilized perfectly. Secondly, it takes the scheduling load off the CPU, which could eventually give higher frame rates as the CPU doesn't have to waste bandwidth on the scheduler.

Whats my take? Okay this is influenced by the article and comments, but nVidia screwed up big time. The 50 series has all of the major architecture improvements in the world, but they're nerfed by the coprocessor and drivers or API's that aren't written for it. This is all on nVidia. Yes, new games will get better gains as they use the AI features more, but nVidia has created a situation where everything has to line up perfectly to gain anything with the new architecture on existing and short term future games. For a company that brags up it's driver support and how closely it works with Microsoft on new features, it feels like nVidia rushed the cards out without refining the features. Sure, maybe the market adjusts their software, nVidia releases new drivers, and Microsoft makes changes to DirectX to take advantage of the coprocessor, but that does nothing when the cards are out right now.

It also creates a situation where game companies will have to design their games either around a coprocessor or without. The article does mention that there's no reason AMD couldn't use a similar design, but that AMD might figure out a way to do even better by utilizing a seperate NPU on the GPU to handle the AI tasks, which AMD has patients on.

 

Also, follow up, this could be why nVidia is very hesitant to talk about NPU performance. Using the GPU right now for AI tasks without proper coprocessor support beyond the basics DirectX has had baked into it for a while means that NPU comparison performance is all over the map. It might not even be something that can be fixed, since the cards are built on the idea that the priority is latency reduction for specific AI tasks the driver controls. It may be that things like custom LLMs or 3D AI generation, which have been shown to crash on the 5000 series more often than expected, may not be possible without a specific programming rewrite to work with the coprocessor.

What a cluster.

 

Without having an article linked, its hard to give my opinion about that, so I can only go off what youve shared.

Its possible, this architecture is actually more forward thinking and laying groundwork for what the future of gaming holds and while it may appear shitty now this could very quickly change for us in the coming years. I mean directx13 has got to be released somewhat soon. This could just be transitional architecture much in the way the 20 series was. Nvidia, has gained a lot of traction in the direction of how games were built and are going to be built.

I memba when, ray tracing came out and how upset people were over the fact nvidia released it but it wasnt really "usuable" so nobody understood why there was a massive shift with gtx to rtx but again the 20series was laying ground work that consumer level gamers werent able to achieve as most were buying 60series cards. As i recall even the 80series struggled with maintaining high frames and running ray tracing. Even DLSS 1.0 wasnt awesome.

I just know, that its also a shitshow when nvidia releases gpus, the only gpu people were happy with was the 10series. After that I seemed like the 20series wasnt enough (had too much tech that was deemed "useless" not justifying the cost increase), the 30 series wasnt enough, the 40 series wasnt enough, and well how history tends to repeat itself, the 50 series wont be enough, however i think that only holds true if you upgrade generationally.

If you own a 40series this isnt an upgrade for you (clearly.) And even prior if you owned a 30 series, the 40 series wasnt an upgrade. I owned a 10series wanting a 20 but so many reviews all said the same thing, if you own a 10series this isnt really an upgrade unless you were after the ray tracing tech and even so that would come at a framerate cost so you had to decide what was important to you.

So i dunno. Maybe amd can do it better, historically they can't at leadt on the top end. But look, these are two different companies that have completely different approaches to achieve their goals.

It would be mighty curious to watch a company like AMD overtake nvidia as then they could hold a large stake in the industry and maybe that shouldn't be wanted, theyve already put Intel in their place (maybe that lasts maybe it doesnt) but then they hold the GPU place as well? I hate to say it but then you have a company that holds consumer CPUs, GPUs and consoles, i dont see that boding well for consumers. And so maybe, we all see it as nvidia has no competition but maybe they are the competition, even if only to overcome it keeps AMD relatively honest. I dunno, I digress.

 

I'm hopeful all of this creates a big space for us "budget" guys to score some way-better-than-average cards for way less soon Don't really NEED a new GPU, but have definitely toyed with the idea of what I might want next if I ever do decide to go triple screen 1440 for the drift sim

 

AMD should have something for you.

 

I don't want an AMD monopoly on GPUs, but a duopoly would at least create more competition. Right now, as you mentioned, nVidia has so much marketshare that they can dictate the features included in the creation of new APIs and game engines. Unlike ray tracing though, nVidia is now hyper focused on proprietary software and hardware changes that can't be used by the competition since they aren't open source or industry standards. nVidia is creating a walled garden with their marketshare lead, and we should all be weary of it. If we ever reach the point that certain games are flat out incompatible current generation AMD or Intel GPUs due to nVidia specific features, the PC becomes a console and will lose it's appeal for many. Maybe that's the end game, maybe nVidia wants a market where there is no competition so that their game streaming efforts can create a false competitive market that's all powered by nVidia hardware. If cards to run games natively at the desired resolutions are too expensive, game streaming could take over with nVidia in control of the hardware to power it.

If you look back, the 30 series was extremely well received as it brought ray tracing to mostly acceptable frame rates and had a large raster improvement over the previous generation. It also brought wider support for GPU encoding and other hardware accelerated features. The problem was, shortly after launch, the 2nd crypto mining boom happened. nVidia cashed in and shorted individual gamers while selling crates of GPUs to mining farms. That's what people hated about the 30 series. Prices went way over MSRP and nVidia played dirty with gamers while professing to not be.

I can't find the exact article on the Blackwell issues again, but this is a similar article about the same changes from Reddit:


Disclaimer (Please read this before up- or downvoting): I'm reporting this early based on the limited info available to spread awareness and encourage more testing by qualified testers and analysis by qualified experts. That's why I've marked as a rumour even though it's not a rumour nor remotely close to settled facts. We're nowhere near knowing the true culprit(s) for Blackwell's performance inconsistency and this only neccessitates additional work and analysis so we can draw valid conclusions.
Not trying to push an agenda here or arm copers, just highlighting a potentially serious issue with the Blackwell's AMP logic unit that warrants investigation. If the AMP issue is real and can be fixed in software, then it'll require an army of NVIDIA software engineers to help rewrite application and game specific code and/or rework the NVIDIA driver stack.

Detailing The Performance Inconsistencies
Blackwell's overall performance consistency and application support is extremely lackluster compared RTX 40 series and resembles an Intel ARC launch more than a usual rock solid NVIDIA launch where relatively uniform uplifts are observed across the board and applications and every usually just works. Application performance sees the wildest inconsistencies but they do extend to gaming as well. The lackluster performance at 1080p and 1440p is an issue plaguing both the 5090 and 5080, that gets somewhat resolved at 4K.

With 50 series Delta Force and Counter-Strike 2 experience FPS regression from 4080S to 5080 as shown in Hardware Unboxed's 5080 review. In the same review TLOU and Spiderman manage atrocious 0% gains for 5080 vs 4080S at 1440p. The result of that is that when using upscalers the performance gains of 50 vs 40 series tank. And remember upscaling is almost mandatory for heavy RT games a key selling for NVIDIA graphics cards. But the worst example yet is probably TechPowerUp's Elden Ring ray tracing performance at 1080p where the 5080 trails even the 4070 Super, available >here<. In TechPowerUp's review Elden Ring at native 1080p both 5090 and 5080 fail with 5-8 FPS regression vs the 4090 and 4080S.

But the most odd thing so far has been that RT performance uplift was consistently worse than raster uplift in nearly every single 5080 review. This is clearly shown in the TechPowerUp review, available >here< where in the majority of games 5080 and 5090 saw larger penalties from turning on RT than equivalent 40 series cards.

Then there's the obvious lack of support for a ton of professional and AI applications where reviewers had to wait for an update adding Blackwell support, but that obviously didn't happen, not even a week later when the 5080 launched. IDK if this is just me but I don't recall this level of incompatability with professional applications for any of the previous launches (20-40 series), isn't it unprecedented for an NVIDIA generation?
And when applications work their performance is sometimes broken resulting in a 5090 loosing to even a RTX 4080. Just watch some of the professional workload centric reviews and you'll see how bad it is.

The most insane performance degradation I've seen is outside of professional workloads is the testing by Guru3D with 3D Mark ray tracing, available >here<. When testing the 5080 in 3D Mark Hybrid ray tracing benchmark they observed a 21% lead over the 4080S. But when they ran the full path tracing benchmark the 5080 was now 31% slower than the 4080S and could only match a 4070 Super. The 5090 has the same issue although to a lesser degree showing a lead vs the 4090 of 45% (Hybrid RT) vs 24% (Full PT).

The Possible Culprit
Thanks to NVIDIA's Blackwell GPU Architecture Whitepaper we might have a likely culprit, although it's likely it's not the only one, although probably still the most significant. The new AI Management Processor is in fact much more than just an AI workload scheduler:

"The AI Management Processor (AMP) is a fully programmable context scheduler on the GPU designed to offload scheduling of GPU contexts from the system CPU. AMP enhances the scheduling of GPU contexts in Windows to more efficiently manage different workloads running on the GPU. A GPU context encapsulates all the state information the GPU needs to execute one or more tasks."

"The AI Management Processor is implemented using a dedicated RISC-V processor located at the front of the GPU pipeline, and it provides faster scheduling of GPU contexts with lower latency than prior CPU-driven methods. The Blackwell AMP scheduling architecture matches the Microsoft architectural model that describes a configurable scheduling core on the GPU through Windows Hardware-Accelerated GPU Scheduling (HAGS), introduced in Windows 10 (May 2020 Update)."

AMP is a direct on-die RISC-V CPU context scheduler with extremely low latency and high bandwidth access to the Gigathread Engine. It sits in front of Gigathread engine and offloads context scheduling from the CPU and taps into Hardware Accelerated GPU Scheduling (HAGS) supported in Windows 10 and 11. This tight cointegration is crucial for MFG, neural rendering and LLM integration into video games, and beneficial to multitasking, content creation, and existing gaming experiences. It doesn't just magically work as intended and requires a proper code implementation and can be a double edged sword (more on that later).

Doing the largest redesign of the GPU wide frontend (not GPC level) since Fermi introduced the Gigathread Engine in 2010 without significant game and/or driver code rewrites is asking for trouble. On 40 series and prior the CPU communicated directly with the Gigathread Engine. But on 50 series, assuming no code rewrites, the CPU has to communicate to the Gigathread Engine through the AMP which adds significant latency and scheduling overhead or either partially or completely breaks scheduling. This results in severe performance degradation without code rewrites as seen with Elden Ring RT and 3D Mark Path tracing. It's also not surprising that when implementing a change this profound some applications just straight up refuse to work.

There's a Twitter post by "Osvaldo Pinali Doederlein" on AMP, where people are discussing why AMP could be causing Blackwell's inconsistent performance and how big the performance impact of Blackwell's new CPU -> AMP -> Gigathread scheduler paradigm is without code rewrites. Psst it's likely -10 to -20%.

The 5090 also seems to have worsened memory and SRAM latencies and L1 bandwidth regression vs the 4090 as reported by "harukaze5719" on Twitter. This is unrelated but could in some instances explain some of the 5090’s performance degradation (vs mean uplift).

(Conclusion): It's too early pass definitively judgment on Blackwell's performance issues, what's causing them, how likely the issues are to be either software and/or hardware related, and if they can even be fixed. With that said there's clearly a lot wrong and the issue spans across many games and many different types of compute and AI applications.
Getting 5080 gaming uplifts anywhere from sub -20% (Elden ring RT 1080p) to +40% (CB2077 HW Canucks 5080 Review) is unprecendented for an NVIDIA launch. NVIDIA Blackwell has severe performance inconsistencies and is reminiscent of ARC Battlemage. This shouldn't be dismissed as node stagnation or something to be expected. No there's clearly something wrong a fundamental level, which could be either a hardware flaw, broken software or a combination of both.
Hopefully NVIDIA can iron out the issues and improve Blackwell's overall performance and consistency over time, and improve it enough to deter AMD from doing something really stupid with RDNA 4's pricing.

 

Ill take a read at that when i got a bit more time to digest it.

To your points, yes the 30 series was a mess in general and yes there was more of an uproar over the pricing then anything. My point is the 20 series was transitional architecture which allowed them to make a second and third card afterward to improve. I mean now were at a point where Ray tracing is a just another option. Thats crazy to think about. But the 30 series was a crazy time, i think additionally pc gaming sort of took off quite a bit with all the work from home bullshit. People had money and time.

Nvidia doesnt owe gamers anything, capitalism and all. If you wanna blame a corporation for what they are doing, Microsoft was the real OG, later followed by Apple. Bring your consumers in and trap them in your products and services, it works and its gross.

And when you think about where gaming has been going yes i agree its eventually going to be on servers. Steam changed all of that, we dont even own those games. I think g2g is one of the few places that you can own the game and move to and from places.

Pc gaming is ultimately impractical anyway, it kind of always has been, ever since consoles were released as good as theyve been. pc gaming is just an after thought. So its just a luxury that were able to afford for the time being. I mean if I just used a ps5 i could open an entire corner of my room for space.

 

The 20 series also sucked at raster performance Gen on Gen. If the 20 series had of given significant raster uplift, the blowback from the poor RT performance would of been non existent. I'll also add that the 10 series was S teir. The 9 series included controversy and meager performance, and the 10 series was a huge improvement without any real price spike. It also killed AMD in a way. The Fury cards had no problem keeping up with the 980 and 980Ti, but there was no answer from AMD to the 1080 and 1080Ti.

I never said I'm mad about nVidia or hate capitalism? That being said, trust and monopoly busting used to be a big US thing. I know that's pretty much over, and without getting political, it's okay for people to wish for the days when capitalism wasn't an answer for why monopolies are okay. Personally, I have zero power over any of that and I refuse to feel shamed or guilty over it.

Steam is, well, unique. It's also basically a monopoly, but Gabe isn't interested in leveraging that position like most companies. I'll stop there, I'm not upselling the Steam model so much as saying I don't really have a problem with it personally. If Steam started doing more sketchy things, I probably would be more inclined to knock them. Software ownership and preservation is a real problem, and there's plenty of examples of the system being abused.

PC gaming, because of Windows continuity, has always offered a far more diverse experience than consoles. I fear the death of indie games and run away random successes if the cloud takes over. Gaming in general is in a weird place. Sony isn't making enough even though they dominate the high end console market. I wouldn't be shocked if the PS6 is their last console if they don't turn a big profit on it. Microsoft was always in the console market to protect DirectX and Windows, and they seem to be pushing the cloud as their next step to do that. They haven't even reacted to things like Proton DB, which signals their strategy is changing beyond even what's been said by Phil. Nintendo, is something. Utter dominance. The Switch really hammered home that. They took what was basically a chip built for streaming devices and leveraged their exclusive titles into a truely dominant position. The Switch 2 looks to be a generation or two backwards in hardware, but I expect it to absolutely dominate people's homes unless the pricing isn't right. Nintendo might be the worst company though, from a consumer point of view. They patient troll. They aggressively use lawsuits. They actively control the availability of titles on their services to allow for the resale of games for full price over and over. It's kind of nuts, but they do it all so well that no one cares.

 

Microcenter has very limited stock of 5080's and 5090's at the moment. WTF happened to $999 for the 5080 and $1999 for the 5090?

I'm seeing $1300-$1500+ for the 5080 and $2500+ for the 5090s.

 

Depends on what you believe, but Tech Tubers are saying that nVidia left the AIB partners no room to price cards at MSRP and make a profit. The only way to make even break even, supposedly, is to sell a majority of models well over MSRP.

It fits. It's also well known since EVGA left the market that nVidia is horrible to their AIB partners. The FE now exists to set a minimum price for advertising purposes.

 

Ok.... Two more quick stupid questions and I should be out of your hair

I ended up installing the SSD, And it was pretty much instantly working And it seems like it made the computer a thousand times faster, but the graphics/picture quality seems like it's something from the '90s. I know the computer is old as hell, but before the SSD install these picture was fine. I'm assuming it has something to do with a driver not being installed, or someone on Reddit mentioned a BIOS issue. ?

After I put in the SSD, I did absolutely nothing but turn it on so maybe I didn't do it right?

Second question: just because I like tinkering, I took the advice here and changed out the thermal paste on the CPU, But it seemed like when I tightened the cooler back down, it was 10 times tighter than before, And I'm curious if that's bad for the CPU.. I'm assuming it can't be good, to be squished like that, But I have no idea, and I was also pretty careful to not put the paste on too thick.

 

Having done a new install (assuming) I bet you just need the latest drivers for your laptop. You'll need separate drivers for your graphics (assuming it has its own GPU).

Can't comment on the thermal paste part but unless something was out of alignment, it shouldn't have felt "tighter".

 

CPUs use ZIF (zero insertion force) connections for a reason. They rely on a combination of the latch mechanism and the cooler tension to provide enough force for every connection to work. Your cooler should, by design, not be able to put more force on the socket than necessary. Even if it could, it wouldn't likely cause damage to the CPU or socket itself, but the MB could bend too much and break contact between the socket and the MB. If everything is working, you're probably okay.

Also, the cooler should use a either springs on the mounting mechanism or precisely spaced hardware to put tension on the cooler against the CPU. As long as you have no extra parts around after reassembly, you should be good.

 

Ordered some parts to start a new server/NAS build, my Poweredge R420 is not powerful enough to encode the videos from plex, and I’m going to need some more storage before too long.


This will be the case I’m going with.


And I think the storage plan is to get 6 of these, and set them up in ZFS2 configuration for Plex storage. And maybe use an M.2 drive for a cache drive.

 

Had some time today to being the custom case build out with a Fractal North XL

I wanted to go with a more natural theme (wooden) and since the North series already has wood wooden accent on the front I wanted to add more to it. I found some super thin walnut veneer and got to work tonight. This is what I've done so far.

 

Looking good! I feel like that needs incandescent case lighting, lol