I am just getting started with this machine! Please subscribe and like to see the next videos. I have this machine for over a month, so let me know if there is anything you would like to see me try!
How do you get around the new tech that is Nvidia RTX cards? I am running into so many problems because the new cards can't run older projects using cuda 10 and tensorflow 1.x. The new RTX cards seem to only run Cuda 11.1 and tensorflow 2.x
Am looking at getting a P920, have a quick question as can't see the answer in the manual Does it come with the 8-pin power cables to connect to the graphics cards, or do they have to be ordered separately, and if so are they standard or Lenovo proprietary ?
I do not currently have the system, I returned it many months back. The system did have standard 8-pin connectors TO the graphics card. The other end of the cable, that connected into the power supply, MIGHT be more proprietary. I do not recall if there were extra connections, I know it was apparently possible to put a 3rd GPU on that system somehow. Mine was only equipped with 2. The entire power supply system in that machine is very modular, things snap in and out quite easily. BUT, that also means it is very Lenovo specific/proprietary.
I just got one of these things on Ebay with Both of the monitors - for less then 800 USD - BUTT the P920 is no longer top of the line. Still one of most well built systems I've ever owned. I do wish Intel had better single threading performance. I paid 400 USD for your 20k computer - 2 years later. Was it worth it!?
what is a good (low to moderate cost) pc for someone wanting build / train deep neural networks? how to choose the CPU, number of cores, etc. AMD or Intel? i7, i9 or xeon? How to choose the amount of ram. how to choose which nvidia card? Is the number of cuda cores the best metric? Please illuminate this confusing subject with your thoughts! (Thank you for the videos!)
@@seanhdr4k629 well that would more depend upon the model or with nvidia cards you need to google how many cuda cores they have. More cores doesn’t imply a better model, more cores allow you to do more training in the same amount of time.
@@monkyspnk777 There are many factors to be considered when measuring GPU performance for deep learning training. For example, memory bandwidth, Not only CUDA cores, memory. I already knew about things you mentioned. I was asking for general opinion or remarks in real life. I wasn't being serious.
Jeff, is there a rig where you can, say, start with one GPU card and then gradually add more over time? That sort of approach might be better for my budget.
Perhaps you found what you needed already. Most High End workstations from Dell, Lenovo and HP would work. They have multiple PCIe x 16 slots and tons of RAM sockets. Most of them that came out in the past 10 years or so supports dual CPUs as well. Look into Dell 7910/20/60, HP Z840, HP Z8/Z6 G4
Please do a video where you refurbish an old machine for the rest of the world who can’t afford a new one. Bonus points if you rebuild a dual processor 5,1 macpro, I have a source for nvidia gpu that are chipped for macs.
The machine was decently loud, running full power, I could not really use anything but a lav mic. It also really made my office hot if I had the door closed, and it was the dead of winter!
But can it run Minecraft though? Haha seriously though awesome video, I've been a big fan of Lenovo ever since I've used one of their Thinkpads for work and their ThinkStations look incredible.
I have Ryzen, Core, Xeon, and Epyc systems. You don't understand the limitations of Ryzen. The point of running Epyc or Xeon is for compute scalability, memory scalability, and expansion capability. Ryzen has very few PCIe lanes, and even with HD-DIMMs, it caps at 192GB RAM. Whereas, Epycs and Xeons support terabytes of RAM, and have close to or over 100 PCIe lanes available. Xeon systems make great VM servers as the DCPMM memory can now be had for reasonably 29 cents per gigabyte. There are also vendor-specific CPUs such as the 8259CL that require a voltage regulator mod but are quite capable in 2S at $260 for two. Your comment was untrue 3 years ago and it's untrue now.
I am just getting started with this machine! Please subscribe and like to see the next videos. I have this machine for over a month, so let me know if there is anything you would like to see me try!
How do you get around the new tech that is Nvidia RTX cards? I am running into so many problems because the new cards can't run older projects using cuda 10 and tensorflow 1.x. The new RTX cards seem to only run Cuda 11.1 and tensorflow 2.x
Nice! Too bad you can't keep it. :(
Would love to have one
Sic - this is 7.238.000,00 Zimbabwe Dollars! ;-)
I' waiting for january to get a $1000 RTX 3080TI 20GB to complete my Ryzen machine learning rig.
Nice GPU, I will have to make the hop to Ampere at some point.
@@HeatonResearch The Nvidia RTX A6000 for 45 hundred us dollars is an approprate card for a professor: ua-cam.com/video/kw78MUnOqIs/v-deo.html
Exciting. Always loved the hp/lbm/lenovo workstation cases.
A three year old machine... that is still jaw dropping!!
Does the top GPU stay cool, they look pretty close together & those GPU's only having one fan, is there sufficient air flow for the top GPU?
Am looking at getting a P920, have a quick question as can't see the answer in the manual
Does it come with the 8-pin power cables to connect to the graphics cards, or do they have to be ordered separately, and if so are they standard or Lenovo proprietary ?
I do not currently have the system, I returned it many months back. The system did have standard 8-pin connectors TO the graphics card. The other end of the cable, that connected into the power supply, MIGHT be more proprietary. I do not recall if there were extra connections, I know it was apparently possible to put a 3rd GPU on that system somehow. Mine was only equipped with 2. The entire power supply system in that machine is very modular, things snap in and out quite easily. BUT, that also means it is very Lenovo specific/proprietary.
I just got one of these things on Ebay with Both of the monitors - for less then 800 USD - BUTT the P920 is no longer top of the line. Still one of most well built systems I've ever owned. I do wish Intel had better single threading performance. I paid 400 USD for your 20k computer - 2 years later. Was it worth it!?
Jeff, do you happen to know what is the monthly electricity cost of this machine if it runs while computing 24/7 for a month??
Great video Jeff - looking forward to what you do with that beast! Do share your source code for your killer projects :)
oh yes, for sure.
@@HeatonResearch SUBSCRIBED!
@@shadowshadow2724 Thanks! and welcome to the channel.
what is a good (low to moderate cost) pc for someone wanting build / train deep neural networks? how to choose the CPU, number of cores, etc. AMD or Intel? i7, i9 or xeon? How to choose the amount of ram. how to choose which nvidia card? Is the number of cuda cores the best metric? Please illuminate this confusing subject with your thoughts! (Thank you for the videos!)
I would take 512G - 1T RAM as some of the processing / multi processing will require, Sir. Dual GPU is cool.
Time to unveil Proton 😁
When I go to the website for the P920, I only see windows pro as the offered operating system. How did this machine come with Ubuntu?
I want to know the true benefit of this machine over a decent gaming pc. Benchmarks please!
That is actually next week's video.
Is that Quodro card better than NVidia’s V100 or A100 cards?
Define what “better” means to you.
@@monkyspnk777 better training yield
@@seanhdr4k629 well that would more depend upon the model or with nvidia cards you need to google how many cuda cores they have. More cores doesn’t imply a better model, more cores allow you to do more training in the same amount of time.
@@monkyspnk777 There are many factors to be considered when measuring GPU performance for deep learning training. For example, memory bandwidth, Not only CUDA cores, memory. I already knew about things you mentioned. I was asking for general opinion or remarks in real life. I wasn't being serious.
Jeff, is there a rig where you can, say, start with one GPU card and then gradually add more over time? That sort of approach might be better for my budget.
Perhaps you found what you needed already. Most High End workstations from Dell, Lenovo and HP would work. They have multiple PCIe x 16 slots and tons of RAM sockets. Most of them that came out in the past 10 years or so supports dual CPUs as well. Look into Dell 7910/20/60, HP Z840, HP Z8/Z6 G4
Ohhh pity it doesn’t have the full 2TB of RAM 😄 But this machine would be a real asset to do great scientific research.
Hah, indeed. That last 512GB of RAM is REALLY expensive, as it means you need a higher grade Xeon CPU.
@@HeatonResearch That is true. I also noticed that on the configuration page,
@@HeatonResearch forget xeon. Threadripper is where all the cool kids are
Please do a video where you refurbish an old machine for the rest of the world who can’t afford a new one. Bonus points if you rebuild a dual processor 5,1 macpro, I have a source for nvidia gpu that are chipped for macs.
what‘ about the noise?
The machine was decently loud, running full power, I could not really use anything but a lav mic. It also really made my office hot if I had the door closed, and it was the dead of winter!
But can it run Minecraft though? Haha seriously though awesome video, I've been a big fan of Lenovo ever since I've used one of their Thinkpads for work and their ThinkStations look incredible.
People from 2050: my cell phone is more powerful than this box
more like 2030
Xenon is an element, XEON is a CPU name
AMD ryzen literally made dual xeon obsolete at the price point.
I have Ryzen, Core, Xeon, and Epyc systems. You don't understand the limitations of Ryzen. The point of running Epyc or Xeon is for compute scalability, memory scalability, and expansion capability. Ryzen has very few PCIe lanes, and even with HD-DIMMs, it caps at 192GB RAM. Whereas, Epycs and Xeons support terabytes of RAM, and have close to or over 100 PCIe lanes available. Xeon systems make great VM servers as the DCPMM memory can now be had for reasonably 29 cents per gigabyte. There are also vendor-specific CPUs such as the 8259CL that require a voltage regulator mod but are quite capable in 2S at $260 for two. Your comment was untrue 3 years ago and it's untrue now.
ฉันใช้มันอยู่เหมือนกัน
Хорошая машинка, жаль дорогая(
#theleakerlane
That's great but boo for the Intel chips. AMD all the way