Captain_Yuri said: AMD Ryzen 7000 “Zen 4 3D” CPU Rumors: 2nd Gen V-Cache, Early Zen 4 & Zen 4 V-Cache Performance Figures https://wccftech.com/amd-ryzen-7000-x3d-zen-4-3d-v-cache-cpu-rumors-specs-performance-launch/ Personally idk if I want to continue to post MLID rumours because of how wrong he actually has been about Zen 4 IPC, Raptor Lake and Arc. I think maybe once we see zen 4 details next week, then we will see how off he really is and then decide on it. If the chart is legit though, skipping zen 4 and waiting for zen 4X3D would be the wise choice unless you want to spend a ton of money only to be slower than 5800X3D. |
Given that the article itself says:
The chart points out four specific benchmarks though it is not mentioned what these workloads are nor the specific config but considering X3D parts show a major performance benefit, it is likely that these are games and not actual workloads. All CPUs tested here were tested on the same power level and it is also mentioned that the Zen 4 chips weren't running at their final boost clocks which we know are going to be surprisingly high in the 5.5 GHz+ range (5.7 GHz on the flagship).
Add also that we don't know what memmory was used in any of the systems, and I'd say that those benches are maybe not useless, but certainly not a fair representation of Zen4's performance.
Captain_Yuri said: NVIDIA GeForce RTX 4090 24 GB Graphics Card In Production Since August, Leaked Documents Claim https://wccftech.com/nvidia-geforce-rtx-4090-24-gb-graphics-card-august-production-leaked-documents/ I will say with Ampere launch, we did have that whole countdown thing. However this year, Nvidia has been very silent. I think while we will still see Lovelace being announced in September with October/November launch, Nvidia will continue to be quiet in order to try and clear as much 3000 series inventory as possible. |
I'd have been very surprised if those card weren't already on production, given that the demand at launch will be far greater than what they'll be able to produce in any given month.
Same goes for AMD, who is likely already mass producing their Navi 31 chips for their launch later this year.
Captain_Yuri said: Pakistan is implementing GPU tax based on memory capacity https://videocardz.com/newz/pakistan-is-implementing-gpu-tax-based-on-memory-capacity "So, how does this work. When someone buys a GeForce RTX 3060 Ti with 8GB memory, unless stated otherwise, the assessed value for this card will be 192 USD. This value will be converted to Pakistani Rupee (assuming 215 exchange rate for dollar) which will put a 42k Rs. value and then 36% duty and tax will be added with a value of around 15K Rs. A total cost of the card will be 57k Rs. In general, memory size is a good indicator of the graphics card’s performance segment, but in numerous instances it will lead to some incorrectly assessed values. Think RTX 3070 8GB with lower price than RTX 3060 with 12GB VRAM, or Radeon RX 6500XT with 8GB with the same assumed value as RTX 3070." Nvidia's weird memory configurations with Ampere really fucked this up lul |
I can understand the logic behind that, but I think a better and more fair way tto tax them would be by the power they use. More power = more performance = higher price point to begin with.
Please excuse my bad English.
Currently gaming on a PC with an i5-4670k@stock (for now), 16Gb RAM 1600 MHz and a GTX 1070
Steam / Live / NNID : jonxiquet Add me if you want, but I'm a single player gamer.