Ryzen 9 7950X3D

- 9 months and I did not find any code about DMMSY's method - 7950X3D has 96MB L3 cache and the the graph is very sparse and tree-like - The code is opensource and everyone is welcome to contribute. [...]

[...] That's on a 7950X3D with 64 GB RAM and a Samsung 990 Pro SSD. Maybe it performs worse on slower hardware. [...]

I genuinely mean that, fly.io (although as unreliable as it might seem) is already around 5x to 10x cheaper depending on use case, depending on some services it's actually <infinity times cheaper because it's actually completely free when you self host! GCP pricing is absolutely wicked when they charge $120/month for 4vcore 16gb ram, can get around 23 times more performance and 192gb ram for $350/month with Xtbps-ish ddos protection. I have 2 2x7742 1tb ram each, 3 9950x3ds 192gb ecc, 2 7950x3d's all at <$600/month obv the original hardware cost in the realm of $60k - the epyc cpu's were bought used for around $1k each so not a bad deal, same with ram overall the true cost is <20k. [...] This also includes 100tb+ hdd of storage and 40tb of nvme storage all connected with 100gbps switch pair for redundancy for a cheap cheap price of $500 for each switch. [...]

Nice, I'm running a 7900XTX/7950X3D so I may give it a try - I don't game anything like as much as I used to though anyway so if the odd game did cause issues it's no biggie.

[...] My main PC is a 7950X3D which has the same core count/threads as the Strix unit, and the Strix benches within margin of error as the 7950X3D. Which is to say the performance is the same. [...]

[...] The AMD Ryzen 9 7950X3D I previously tested on has 96 MB of L3 cache on the first die, and has lower L3 latency than Lion Cove in Intel’s Arrow Lake platform. Beyond L3, AMD achieves better load-to-use latency even with slower DDR5-5600 36-36-36-89 memory. [...]

Hi HN, I recently finished building a deterministic AI-based firewall called Jave Ethical Defender. [...] During real-world testing, it successfully neutralized a 5,000+ node botnet with 100/100-rated IPs from AbuseIPDB using under 10% CPU on a 7950X3D. [...] I’ve published documentation, architecture, screenshots, and the actual performance report here: (Core engine is gated for responsible access requests to avoid misuse.) It’s built to be lightweight, non-subscription, and eventually open to self-hosters and SMBs who want real AI-first perimeter defense. [...]

[...] A card with a fraction of the FLOPS of cutting-edge graphics cards (and ideally proportionally less power consumption), but with 64-128GB VRAM-equivalent, would be a gamechanger for letting people experiment with large multi-modal models, and seriously incentivize researchers to build the next generation of tensor abstraction libraries for both CUDA and ROCm/HIP. [...] Even DRAM on the AMD Ryzen 9 7950X3D shows less latency. [...]

Pricey appliance-oriented development. [...] And at least go with 10 GbE PoE+ for WiFi 7 APs. Non-cloud DVR security cameras are fine with ordinary PoE and 1 GbE (because cloud only WiFi cameras are inherently insecure). [...] 2 UPSes with 2 2U expander battery packs each; I probably should move to some sort of LiFePO4 UPS at some point. [...] Also have 2 dedicated 7950X3D boxes with 2x 100 GbE each for network load testing and development. [...]

I recently bought a machine with an RTX 4090 and a 7950X3D. [...]

If you look at modern games like AC: Valhalla and compare Wine with Windows [1], you'll see that the Linux scheduler is superior, resulting in a much better worst-case performance. (But of course the Wine emulation results in a slightly lower best-case performance.) Or some games that have been tested with Proton by the developers, like Far Cry 6, are just faster on Linux in every case. [...] And the Ryzen 9 7950X3D is very popular as Hetzner's AX102 Dedicated Server, for example. [...] Windows licenses are expensive if you need 2000+ CPU cores...) The result is a superior scheduler co-optimized with the CPU, so I'm not surprised that Linux came out ahead. [...]

[...] (Like USA & AMD 7950X3D / AX102 server) 2. [...]

I do have a 7950X3D. It’s improved a lot, I still use Lasso but strictly speaking I don’t really need to.

[...] The 7900X3D and 7950X3D became infamous for having less practical performance compared to their lower-tier brother, because the scheduler couldn't grasp which cores to schedule threads on. [...] AMD might make great hardware, but they consistently seem to give nary a damn about software.

[...] All the ease-of-use and ergonomics of C++, and the raw bare metal performance of Python. [...] This is on a Ryzen 9 7950X3D. [...]

LLMs seem to be a bit more accessible than some other ML models though, because on a good CPU, even LLaMA2 70b is borderline usable (bit under a token/second LLaMA2 70b on an AMD Ryzen 7950X3D, using 40 GiB of RAM.) Combined with RAM being relatively cheap, seems to me like this is the most accessible option for most folks. While an AMD Ryzen 7950X3D or Intel Core i9 13900K are relatively expensive parts, they're not that bad (you could probably price out two entire rigs for less than the cost of a single RTX 4090) and as a bonus, you get pretty excellent performance for code compilation, rendering, and whatever other CPU-bound tasks you might have. If you're like me and you already have been buying expensive CPUs to speed up code compilation, the fact that you can just run llama.cpp to mess around is merely a bonus.

M3 Max 3151 / 20463 7950X3D 3173 / 22341 As a developer my Go builds take 5 seconds. I invested in very fast NVMes and coming from 5 minute Java builds this is great. Though what I want is <1sec builds. [...]

[...] I haven't tested Eco mode myself, but from what i've read performance drops by 10-20% depending on the configuration (105w vs 65w). [...] Add: and 7950X3D would do even better. [...]

[...] But I don't think either option is really feasible for people paying Hetzner-like prices for hosting; Hetzner's newest Xeon offering is seemingly Cascade Lake, and the only EPYC offered is a single-socket Rome 7502P with 128GB DDR4 for 142 euros, which seems very hard to justify, given they also offer a 7950X3D with 128GB DDR5 for 25 euros less. [...] Not sure how secure that really is, though; I haven't really looked into the ARM offerings as much as I should have, mostly since, if you don't want Apple, I'm not aware of a good middle-ground between a cheap SBC and a $3,000+ Ampere server, outside of jerry-rigging some second-hand Gigabyte Cavium ThunderX2 nodes off eBay.

Ditto - 7950X3D and 6900XT; game stutters like crazy at times. [...]

Yes, when I was buying this PC beginning of the year I was consdering 7900X, 7950X and 7950X3D. At the end chose the 7950X, super happy about it. Absolute beast.

ryzen 9 7950x3d, 4070 ti super, & 64gb of ram 🙇‍♀️ and for all the details i got the matching cooler (thermaltake th820) and lcd case panel, a couple 2tb SSDs, and gigabyte 650 aorus elite ax for my mobo!!

[...] Performance seems to be fairly poor despite my system being pretty beefy (Ryzen 9 7950x3d). I see the performance monitor and notice that the render loop seems to fairly regularly exceed the latency for 60fps, despite this being a 'simple' task by modern standards. [...]

[...] I’ve been watching performance reviews of some video game things recently and to my ears it’s just total gobbledygook now. The 13900KS, 14900K, 7900X, 7950X3D, all sorts of random letters and numbers. [...] Intel has been using i3, i5, i7, and i9 forever. [...]

[...] It would presumably be more pronounced on the 7950x3d since only one has access to the extra cache. [...]

[...] Performance does need some work; when I manage to get it to actually ignore vsync (hence 160 FPS), my frame times are all over the place - with a 7950X3D and 6900XT.

7950X3D has 96MB L3 cache and the the graph is very sparse and tree-like. [...] (add -rs in the url) I'll update the readme with more info... [...]

I have a 7950x3d. [...] I guess I'm subconsciously drawn to chips with hardware loop buffers disabled by software.

These aren't build failures. While they are not compilation failures, I do sort-of consider unit test failures to be build failures if the unit tests are typically ran as part of the build. [...] Even their build process changes over time. Chromium and Electron are both fairly easy builds. Not because they are simple to build, but rather because the build process is extremely automated and documented. (The real hard part is balancing concurrency with OOM: Electron defaults to -j200 and on my 7950X3D it will happily eat all of the 128 GiB of RAM. [...] I would've thought you'd just go to the previous version, which is presumably more stable and supported too.) Well, I want the bug to be fixed. Even if I am not going to submit the actual patch that fixes it, providing bisect details in my bug report will vastly increase the odds that it gets fixed. However, with the magic of bisect I often can make working patches for complex bugs and then patch them. If it's software I use on my system, I can then add a temporary patch in my Nix configuration and continue to run the latest version. [...] I don't like to rely on hoping someone else will figure out my bugs if I can help it. [...] If I run into a bug on something I maintain a package for, I will often send a PR fixing it and then pull that patch into my package temporarily. [...]

[...] Though nowadays I would probably use a thin client Raspberry Pi running Moonlight to do it cheaply. [...] I picked up a 5930k for my build at the time for its 40 PCI-E lanes. [...] Also with the best CPUs for gaming nowadays being AMD's X3D series because of its additional L3 cache, I wonder about the performance hit with 2 different VMs fighting for cache. [...] The 7950X3D had 3D cache only on half of its cores, so games generally performed better pinned to only those cores. So with only 2-3 VMs/PC, and you still needing a GPU for each VM which are the most expensive part anyway, I'd pay a bit more to do it without VMs. [...]

7950X3D, 96G, 18TBx4, 4TB NVMe x2 my GPUs are gtx1080, rx570 and the 7950x3d, FSP 1000W ATX3 platinum I use proxmox as my OS. [...]

[...] a Ryzen 8700G, the top of the APU stack, only has 8 cores while Ryzen 7950X3D will go up to 16. [...] AMD traditionally has made a lot of enthusiast and system builder sales but I wonder if that will change in the future. [...] where CPU performance doesn't matter but you need a display output and some graphics acceleration. [...] [1] AM5 Ryzen CPUs do have display outputs now, but AFAIK there's no hardware acceleration on it, you need a dedicated GPU anyway. [...]

357 - Chrome with extensions 511 - Chrome without any extensions 433 - Clean firefox install 500 - Clean edge install On an AMD Ryzen 7950X3D / 5600MHz RAM Interestingly, I got substantially different results depending on whether I used chrome with my typical extensions or not. Looking at the flamegraph, I don't seem to see the full picture of why there's such a substantial difference. Bitwarden/React Developer tools (the 2 primary extensions I use) don't seem to make enough of an impact to account for the roughly 40% performance increase seen without them. [...] I'm not incredibly familiar with browser performance profiling, so there's probably more to the story.

[...] It'd be nice to have new hardware that didn't cost more than a decked out Ryzen 7950X3D system.