Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

porina

Member
  • Content Count

    10,225
  • Joined

  • Last visited

Awards

About porina

  • Title
    Slightly Salty, Fairly Fishy

Contact Methods

  • Twitter
    https://twitter.com/porinapew

Profile Information

  • Location
    UK
  • Occupation
    former electronic/acoustic engineer

System

  • CPU
    i7-6700k
  • Motherboard
    Asus Maximus VIII Hero
  • RAM
    G.Skill Ripjaws V 3200 2x8GB
  • GPU
    Gigabyte 1650
  • Case
    In Win 303 NVIDIA
  • Storage
    Samsung SM951 512GB, WD Blue 1TB
  • PSU
    Corsair HX750i
  • Display(s)
    HP LP2475w
  • Cooling
    Noctua D12
  • Keyboard
    Logitech G213
  • Mouse
    Logitech G403
  • Sound
    Cheap Amazon speakers that aren't bad at all
  • Operating System
    Windows 10 home
  • Laptop
    Asus FX503VD (7300HQ, 1050, 16GB ram)
  • Phone
    Pixel 3a

Recent Profile Visitors

30,843 profile views
  1. ECC registered requires specific support to work at all, and you're not going to find that on consumer level boards.
  2. If true, then disappointing from a gamer perspective. I do wonder, in what form do these sales take place. Presuming nvidia are selling the 30 series FE to miners, since if AIBs were involved we might hear more separately. I'd guess more likely 3070 and 3080 since the high VRAM and cost of the 3090 is unlikely to make that attractive, unless the miners get different much pricing. Will the miners have to pay over retail rates to get up the queue? Or is there essentially some discount for a bulk buy? It could go either way. You might say, why would nividia discount a card they know t
  3. Seems suspect the SSD would apparently write an order of magnitude more data to flash than received from host. I think this might be a misreading of a value somewhere. In terms of wear on the NAND I think the more interesting values are max erase count, and wear leveling count. The normalised values I think scale from 100 (new) to 0 (worn out), so you're not even half way through its wear life yet.
  4. That article is pretty badly written and it is unclear to me what the extent of the support is. Regardless, without official AMD support I'm not going to expect it to go anywhere fast.
  5. Lucky you. I've not seen any 3070 or 3080 on general sale from a reputable place (non-scalper) since launch. I have seen 3090's though.
  6. 4th gen consumer was Haswell. There are Haswell CPUs supporting DDR4, but you need to go X99 chipset, and beware the CPUs have a 5000 model number.
  7. This is what got my attention. There is software I'm interested in that already runs on both nvidia and AMD hardware but not Intel. If this gives a way for the nvidia code to run on Intel, it could open up use cases on iGPUs, not that I'm expecting much performance from them compared to dGPUs.
  8. Apparently this project translates CUDA to oneAPI Level 0, which I think only Intel are supporting for now at least. I'm not fully up to speed with oneAPI but my understanding is that it is a significantly Intel created open standard. Unless AMD also starts to support oneAPI, I don't think this software will be any use on AMD. What state and performance is that in? Got a name so I can look it up?
  9. This appeared on my radar when Raja retweeted it. https://github.com/vosen/ZLUDA The creator of it so far has only been working on getting it to work on Geekbench, were it shows an overall performance increase compared to OpenCL, although they do caution that GeekBench is an ideal case for the translation work going on. It is described as "proof of concept" level so it may be some time before it can be put to serious use. It will be more interesting to see what relative performance is like for less optimal code cases. Apparently it works on the iGPU since Skylake era.
  10. I'm not talking about capacity, I'm talking channels. This product is clearly aimed at use cases where the already low memory bandwidth is not a concern (Cinebench users rejoice). IMO even the full 8 channels isn't really enough to feed it under more stressful workloads.
  11. CAS latency is only one of many timings though, and the large array of others also have some influence in overall results. Still, depends on the software being run, some scale better with overall latency, some scale better with overall bandwidth. I'm not up to speed on what modern games do when it comes to that behaviour. Differences are minor and it isn't going to make or break a system.
  12. It hurts to see the crippled ram layout, but I guess it is sufficient for light usage cases. I wonder if they could have given it the full ram support if they went SODIMM instead?
  13. I bought the 2080Ti after 3070 was launched. For a comparable cost and performance, it will suffice, since my long term goal is a 3080 or whatever higher VRAM model they bring out in that area. Stock is continually coming in, but it is going out just as fast to existing pre-orders, so you don't see any appear on normal store pages. Example from one major UK seller who is listing numbers of units each week: https://www.scan.co.uk/shops/nvidia/rtx-30-series-faqs
  14. I'm sure you can get a Ryzen 5000 and Radeon 6000... just before the Ryzen 6000 and Radeon 7000 launch, and the cycle repeats once again... It will be interesting to see how this plays out. Gonna slum it with my last gen recently purchased used 2080Ti for now... I've pretty much given up any expectation of getting what I want of latest gen this year.
×