Search the Community

Budget (including currency): 1000 USD / ~4000 PLN Country: Poland Games, programs or workloads that it will be used for: When it comes to games I will play GTA 5, Elite Dangerous, Halo Infinite, Cyber Punk 2077. The biggest tool on the hardware (especially GPU) will be probably my development job in machine learning – training models, evaluating them, building apps around them. Other details: These are the parts that I plan to build. RTX will be used thus lowering its cost to around 372 USD. Any suggestions GPU RTX ASUS TUF 3070 Ti CPU AMD Ryzen 5 5600X MB ASRock B450M-HDV R4.0 RAM Kingston fury renegade ddr4 32GB 3600 mhz POWER MSI MAG A550BN 550W CASE Genesis Irid 503 V2 (NPC-1558) SSD Lexar NM710 2TB M.2 2280 PCI-E x4 Gen4 NVMe

Hello I am a Phd student in biomedical engineering focusing on robotics. I need a PC for running solid works, MATLAB, visual studio code, Ansys, ROS and maybe someother simulation and machine learning softwares. I am looking at two options from Dell technologies Precision 3260 Compact Workstation Precision 3660 Tower Workstation I was wondering if these options are okay or is it better to have a Xeon Processor? or are there any recommendations from DELL. Thank you

Budget (including currency): Indian Rupees, about 1 Lakh Rupees (1200 US dollars) Country: India Games, programs or workloads that it will be used for: Mainly for Machine learning and Deep learning. I am a Uni student, hence the reason why I want a Mini-itx build. Being able to just pack it and take it on a suitcase is very necessary for me. I will be gaming too...thinks like RDR2 and many more (I don't really care about the type of games). Since I want to work on Large Language Models and Reinforcement Learning, VRAM will be important. But because I am limited by my budget and the space in the case, can't go for something like dual GPU config. It's a given but...I want the GPU to be CUDA supported. This locks me to the Nvidia realm...can't be helped. I know that the ARC cards have been recently supported in Tensorflow and PyTorch too, but I can't invest in it hoping that they will get better down the road. AMD is out of the question for ML and DL applications. I am not expecting to work on LLAMA 2 70b parameters or anything remotely crazy like that. But being able to work on the most basic LLM will greatly help me. I am also thinking of learning Unity for tinkering with Reinforcement Learning. I will be using windows for gaming. And thinking of using Ubuntu or Arch for programming (Not sure how they are for Machine learning tho...and I have close to no experience in Linux) Other details (existing parts lists, whether any peripherals are needed, what you're upgrading from, when you're going to buy, what resolution and refresh rate you want to play at, etc): Don't include Monitor in the parts list. Since I will be going to UNI, I will probably buy a portable monitor of separate budget (I have 3 months of holidays...monitor is not my concern right now, I already have some ultrawides at home). Already have a good keyboard (My lovely baby...RK61). I am currently using a MacBook Air M1 with 8GB RAM with Venv-Metal for Tensorflow. It is decent, but I am limited by the speed, storage and RAM. I will be buying in about a month. I know that Mini-Itx will have heating issues and this form factor is more expensive than the standard ones. But this is something I have to live with, I want the system to be portable enough that it should not be a hassle for me. But my room does have Air Conditioner that is always at 20 Degrees or something. That might help. I also have a RX560 at home, my father uses this for his daily workflow (Stock market). So messing with that system will lead me directly out of the house ;).

Hi all. I've been trying to figure out an approach for sourcing a gpu setup for my general purpose home server, I want to tinker w/ machine learning in my spare time. Going to be headed to college soon, so budget is limited, the main components will be from my previous PC build. As I understand things VRAM is extremely important in training ML models, but I also understand that you can split workloads across multiple GPUs? Part of me really wants to use a ARC setup, especially at the price point, and the nice video encoding functionality. Would I be better off grabbing 2 arc 380's(very budget friendly), an arc 380 and an arc 750, a single arc 770 @ 16gb, or trying to snag an older high-end rtx 20-series? Also, does anyone know if it's possible to train off of both Arc and Rtx at once?

Hi, I found a basic notebook online to write code to implement CNN from scratch. Can some machine learning specialist help to fill the notebook with the relevant bits and pieces. cnn.ipynb

Budget (including currency): Approx INR 100,000 (can go upto INR 110,000, only if required) Country: India Games, programs or workloads that it will be used for: Machine Learning/Deep Learning, gaming is not very important Other details: Hello everyone, I hope everyone's safe and doing well. I am from India (which is important because components here are expensive than USA) and I am planning on building a single GPU based, Machine Learning/Deep Learning (ML/DL) Budget Rig soon. I'm just starting into this field and I'll be upgrading my build as I gain experience. So far these are the specs I have chosen - CPU - Ryzen 5 3600X CPU Cooler - Wraith Spire Stock Cooler MOTHERBOARD - ASUS TUF GAMING X570 PLUS (Wi-Fi) ATX AM4 GPU - Asus GeForce RTX 2070 8 GB DUAL Advanced RAM - G.Skill Ripjaws V 16 GB (2x8GB) 3200 MHz SSD - Crucial P1 500 GB M.2 NVMe HDD - 3.5" 2TB Seagate Barracuda Compute 5400 RPM POWER SUPPLY - Corsair RM750 (2019) Fully Modular 80+ Gold (750 Watts) CABINET/CASE - Antec Dark Phantom DP501 ATX Mid Tower Case I know for a fact that this build will be able to game if I need it to, but gaming is not my priority. Also, I DON'T CARE ABOUT RGB (Don't hate me for this xD). I want my machine to perform well for ML/DL workload. This will be my first build, I am relatively new to this idea of doing custom builds, and I have learnt whatever I know through Linus Tech Tips build videos, so please forgive me if I'm asking a lot of questions. I have listed my doubts below below - CPU Is Ryen 5 3600X too much or should I go ahead with Ryzen 5 3600 and save some money to spend elsewhere (SSD, RAM, etc.). Also can i expect this build to last for 3-4 years (future-proof) with either of these 2 CPUs? CPU COOLER Is it okay to use the Wraith Spire Stock cooler instead of some other dedicated CPU Cooler? I want my build to last long so if there's any harm in using the stock cooler, I'll put in some extra money and get one. GPU Is it better to get a RTX 2060 Super instead and spend the money on other components like RAM and Storage for this particular workload? Would i be giving up a lot of performance in doing so? MOTHERBOARD The sole reason for buying an X570 Motherboard is because I can have the option to upgrade to a 4th Gen Ryzen CPU in the future. I want to know what other cheaper alternative I have if I decide to stick to the Ryzen 5 3600X/3600 CPUs, because then I wouldn't have to necessarily stick to a X570 Motherboard RAM - Which one's better for a 3200 MHz RAM? Corsair LPX Vengeance or G.SKill Ripjaws V? - For ML/DL workload, is it better to go for 32 GB (2x16GB) 2666 MHz RAM? Or should I just buy the 16 GB dual channel 3200 MHz that I mentioned above instead? - Can I put 24 GB Ram if I want to, with a kit of 2x8GB 3200 MHz + 1x8GB 3200 MHz (16 GB Dual channel + 8 GB single channel configuration)? - Is it advisable to just go with 16 GB 3200 MHz RAM in the beginning and add more RAM in the future? SSD & HDD - Does choosing a 5400 RPM drive compromise a significant amount of performance as compared to the 7200 RPM? - Or should I get a 256 GB SSD and spend money on a 7200 RPM HDD instead? - Is it necessary to go for a M.2 NVMe drive, or can I also go for a 2.5" SATA based SSD? PSU Is 750 Watt too much for a power supply? I might do some upgrades in the future, like a better SSD and more storage, and possibly, a new CPU (Ryzen 4th Gen) CABINET/CASE Should I go for Cooler Master Masterbox K501L RGB (ATX) Mid Tower Cabinet instead? That will be a bit cheaper for me, and I could spend some money elsewhere. Also is there any other better case that I can buy and something that is priced the same? Any other recommendations? I am willing to wait for a few months if components are expected to become a bit cheaper in the near future. Thanks a lot for the help!

Hello guys, I am an Engineering student. I am currently thinking whether i can use my RTX 2060 with RX 580 (currently not have) to work in one system. The system is used for 3D CAD (Solid Works) and for playing games, Machine Learning and other stuffs. So i am thinking whether it is possible to use the RX 580 for the solidworks and RTX 2060 for other stuffs beside solidworks? If it is possible, do you think it is needed? any other recommendation for my usage? Thank you for your reply! Looking forward for your reply!

Hey guys, I have been a long time fan of Linus Tech Tips, and I really enjoy the more technical work that Linus does with Anthony. So, to add on to potential ideas, they already had, I would really like it if they test out our little project that me and my friends made. We spent a lot of time on it and it is really taking shape. The test version will work for Windows, but the final product will run on a raspberry pi. This is free to use, and it should really help drivers. The program is a driving assistant that you can control with the google assistant. It will detect if you fall asleep non-obtrusively using a webcam and machine learning. It also has a HUD that can be enabled to map out obstacles in front of the car. All the processing is done offline so you don't need to worry about privacy like other products. The final product will be cheap since all you need to buy is a raspberry pi and 2 webcams, and the software will be free. It would be great if other users on the forum can also test this since we are just high schoolers, and don't have a large pool of people to help beta-test this product. Please report any issues to the GitHub repo. Thanks again, and Anthony, please make Linus do a video on this, I really want to see him try and figure this out. Here is a link to the GitHub repository: Driving Pal Thanks!

Hello all, me and my friend are currently in college and are getting into computer science and more specifically deep learning. We have been programming several neural networks but the problem is that they take forever to train on my crappy old laptop. I'm looking in to building a new computer specifically for training these networks. Since the graphics card is what does all the heavy work, I'll probably go with a 1080. My question is, what CPU and RAM setup would I need to go along with this build? Thanks

I want to build a PC around $1000. Now, this PC shall have 2 main tasks: • Supervised/unsupervised learning (as of now. Will move onto more complicated tasks later.) • SolidWorks Now I need a little help with the build. Should I opt in for a Quadro with a greater budget, or put greater priority into the CPU? Can someone please list the complete build for me?

?

Budget (including currency): ₹ 4,00,000 Country: India Games, programs or workloads that it will be used for: Training Neural nets, Bioinformatics and NAS Other details (existing parts lists, whether any peripherals are needed, what you're upgrading from, when you're going to buy, what resolution and refresh rate you want to play at, etc): I work in a research lab and are investing in a brand new system for our varying workload. We plan to use Linux as the main OS on 2 x 500 GB SSD on RAID-1 and have 4 x 4TB Drives on RAID-5 for storage. Our lab is setup such that we also need other researchers to have access to the data. The Workstation shall process the data and also allow other users on the network to access that data. We will be training neural nets and will be running bioinformatics simulation too. Here is the planned system configuration: 12 Core Xeon W-2265, 128 GB ECC, RTX 4000 (8GB) and a 10 Gigabit NIC with a 4K monitor. Will support RAM and GPU expansion slots for future upgrade. After revisiting LTT videos, I felt unRAID solution of hardware passthrough would set me up nicely with Processing and NAS. I'm here for suggesting on my setup. Reason for RAID is only because we want some kind of safety when we're running the simulation or neural nets for long hours lasting days. Data shall be backed up in some intervals, but the protocol is not set. Does my use case even warrant against the cost of unRAID license?! Windows will also be used, but initially planned to be run as a VM inside Linux. Going for ZFS on Linux with 'samba' for SMB share would be a better solution for my use case? I'm not sure. HELP ME! Should I go for RAID-6? Please help me understand. And do I really need a H/W RAID card for my comparatively small setup. I would appreciate any kind of suggestions over my choices and any help me to avoid obvious pitfalls. Thank you very much for your time.

Budget (including currency): INR 180,000 Country: India Games, programs or workloads that it will be used for: Training Machine Learning & AI Algorithms Other details (existing parts lists, whether any peripherals are needed, what you're upgrading from, when you're going to buy, what resolution and refresh rate you want to play at, etc): Completely new PC build. Already have a good enough Keyboard, Mouse and Monitor that serves the purpose. Hello Linus Tech Tips community, I am planning to build a PC to train my Machine Learning and AI algorithms and some occasional gaming. I used to train those off of my Omen Laptop (which actually does a decent job) with a i7-9750H CPU, GTX 1650Ti graphics card (Cuda cores are a important factor for me) and a measly 8GB DDR4 RAM earlier. But now I am at a point where I need some more juice out of my system. Also the thermals are a mess on heavy and sustained runs. Some of my iterations might take more than 24 hours to run. But I do not want to pay AWS or Azure the charges they are quoting for my use case. So I have decided to build myself a dedicated PC for this purpose. Apart from doing programming and training algorithms, I like to game occasionally. I mostly play racing games like F1, Moto GP and NFS. After about 2-3 weeks of research and checking availability in my country, I have narrowed down to these Components PCPartPicker Part List Type Item Price CPU AMD Ryzen 7 3700X 3.6 GHz 8-Core Processor $274.49 @ Amazon CPU Cooler Corsair H100i RGB PLATINUM 75 CFM Liquid CPU Cooler $139.99 @ Best Buy Motherboard Asus TUF GAMING X570-PLUS (WI-FI) ATX AM4 Motherboard $189.99 @ Best Buy Memory G.Skill Trident Z Neo 32 GB (2 x 16 GB) DDR4-3600 CL16 Memory $178.99 @ Newegg Memory G.Skill Trident Z Neo 32 GB (2 x 16 GB) DDR4-3600 CL16 Memory $178.99 @ Newegg Storage Crucial P1 500 GB M.2-2280 NVME Solid State Drive $62.99 @ B&H Storage Crucial MX500 1 TB 2.5" Solid State Drive $110.20 @ Newegg Storage Seagate Barracuda Compute 2 TB 3.5" 7200RPM Internal Hard Drive $54.99 @ Newegg Video Card Gigabyte GeForce RTX 2070 SUPER 8 GB GAMING OC Video Card Case Corsair iCUE 220T RGB Airflow ATX Mid Tower Case $99.99 @ Best Buy Power Supply Corsair RMx (2018) 850 W 80+ Gold Certified Fully Modular ATX Power Supply $164.99 @ Best Buy Prices include shipping, taxes, rebates, and discounts Total $1455.61 Generated by PCPartPicker 2020-06-18 09:31 EDT-0400 I am also considering future upgradability in case I need a dual GPU or a more powerful CPU someday. I am also open to about +10% flexibility on my budget. What do you experienced guys think? Any suggestions for any better or efficient components? Thanks in advance for all your help.

Hi, I am interested in getting a laptop and I am confused for which one should I buy. I am starting to learn machine learning and data science and I want to know whether I need a good gpu or not. I am looking for a laptop which can do machine learning and data science within the range of £700 (863 USD) to £1000 (about 1200 USD). I prefer a lower price but can go up to £1000, I am not a gamer and do not want any gamer look to the laptop and want it to have a clean look. Can you help me? This needs to be available in the UK and I want it too last a few years.

Hi, I want to get started on supervised machine learning using neural networks. I know a fair bit about how neural networks work. I also have some coding experience (mainly C++, C#, python). Building my own basic neural network in C++ would be tedious, but would really make me understand everything in depth. The other option I consider is getting started using the tensorflow library for python. This would probably be more fun because the first good be results can be reached way more quickly. Which option is better? My second problem is that I am looking for an interesting, yet not to hard, little project to start with. Any suggestions? regards, Xilef

I have trained a neural network in Keras but when running an sklearn confusion matrix on it after training I found that it always outputs 0. [[32849 0] [ 7215 0]] What have I done wrong? Code here

I was looking for some advice for my current server build. Any help appreciated. PS: I took this template from a reddit forum lol What will you be doing with this PC? Be as specific as possible, and include specific games or programs you will be using. * I am a machine learning developer looking to build a virtualization server. I need high computational resources on the fly. I am also an occasional gamer, so I want the option to run games. Additionally want to possible run a plex server on this build. * I need to use multiple OS's constantly for what I do, so I want to be able to run many operating systems concurrently. I want to connect to my machine with a monitor and connect externally via the internet with GUI or SSH. What is your maximum budget before rebates/shipping/taxes? Let's assume that money is not a concern When do you plan on building/buying the PC? Note: beyond a week or two from today means any build you receive will be out of date when you want to buy. Let's assume right now. What, exactly, do you need included in the budget? (Tower/OS/monitor/keyboard/mouse/etc) Just the PC components, I am not concerned with the case (fun fact I might be building it into a desk), or peripherals Which country (and state/province) will you be purchasing the parts in? If you're in US, do you have access to a Microcenter location? Canada If reusing any parts (including monitor(s)/keyboard/mouse/etc), what parts will you be reusing? Brands and models are appreciated. Nothing to reuse Will you be overclocking? If yes, are you interested in overclocking right away, or down the line? CPU and/or GPU?** Overclocking is a yes Are there any specific features or items you want/need in the build? (ex: SSD, large amount of storage or a RAID setup, CUDA or OpenCL support, etc) CUDA, and possibly RAID Do you have any specific case preferences (Size like ITX/microATX/mid-tower/full-tower, styles, colors, window or not, LED lighting, etc), or a particular color theme preference for the components? No case concerns, as I might be building it into a desk Do you need a copy of Windows included in the budget? If you do need one included, do you have a preference? Nope Other Info: Let's assume that money is not a concern, and I am looking to incorporate an 18 core CPU, 64GB ram, and 4 graphics cards maybe P4000s or even RTX 2080s. I want to share the 4 graphics cards between the OS's if that's possible. What's the best value hardware for this scenario, and possible virtualization software setup. Needless to say, I want to minimize video tearing and I/O complications.

Summary Open AI released the public beta for their new chat tool powered by their GPT technology. GPT is series of increasingly amazing deep learning models, that since GPT-3 are closed source. At their simplest, they try to continue a phrase that's given to them, according to which words are more statistically likely to be chosen next, given the words before and after and a statistical model. Embedded inside the parameters of the model is the statistical correlation between words, and, as the model gets bigger and bigger, with a wider and wider "memory", trained on more and more data, this statistical model can represent increasingly complex knowledge. Luke and Linus tried it on stage during the WAN show, with impressive results. The model is always very confident, even when it's inventing facts out of thin air. ChatGPT is banned from answering on Stack Overflow. People have been using ChatGPT to answer in mass questions on Stack Overflow without checking if correct. Quotes My thoughts The model is mighty impressive. Very fast to respond. It's confident both when it's right and when it's wrong, and the chat/thread functionality allows to refine the output incrementallty. One of the conversation that I tried: I would have very much liked to have such a useful chat bot when I was at the university. I'm doubtful it is very accurate on ambiguous/niche tasks, but on common task that are likely to be covered by a lot of text in the training data, the performance is nothing short of incredible. Sources News https://openai.com/blog/chatgpt/ https://meta.stackoverflow.com/questions/421831/temporary-policy-chatgpt-is-banned Talk with the actual bot https://chat.openai.com/chat Ars Tecnica article https://arstechnica.com/information-technology/2022/12/openai-invites-everyone-to-test-new-ai-powered-chatbot-with-amusing-results/ Yannik (creator that covers ML news)