Pascal GPU series of NVIDIA`s GeForce 20 is self possessed to feature faster memory chips and higher clock speeds, GDDR6 and HBM2 is supported by Volta GPU. For the year 2017 and 2018 the company plans are quiet noteworthy and we should thank Ishimura for the leak and giving us a little glimpse to NVIDIA’s GPU road-map for the next coming years. The same person who before gave the information about the next generation Volta GV104, GV110 and GV102 GPU’s few months ago on the Chinese Baidu forums to talk about the upcoming GeForce gaming graphic cards coming later in the year of 2017 and 2018 respectively. One of the interesting fact is that NVIDIA would actually introduce a GTX 2080 Ti based on the GP102 GPU with powering the flagship Titan X instead of a GTX 1080 Ti as previously they have thanked about.
For 2017, the NVIDIA GeForce 20 Pascal Refresh Allegedly Includes GTX 2080 Ti, 2080 & 2070 Cards With Faster Clock Speed and GDDR5X Memory.
The GTX Titan X would be completely replaced by the new card, selling at a significantly lower price and delivering a better performance. Replacing the existing GP104 based GTX1080 is also in the plans and GTX 1070 cards with a higher clocked variants and bringing the GDDR5 memo to a GTX XX70 class card for the first time. Whether a potential GTX 2070 would feature more CUDA cores than the existing 1070, its not clear right now. Therefore the faster memory and the clock speed is believed to be enough to create a more appealing and competitive product than the existing one offered by the company. A faster “GTX 2080” will replace the GTX 1080. 10-15% more is expected by the GeForce GTX 2080 to deliver the performance compared to the card it’s been replaced with and then sells for $499 at the launch time. Making it a more cheaper product.
Geforce GTX 2050, 2060 and 2060 Ti are also in the line of production.
GP107 and GP106 based graphic cards are like the GTX 1060 and the newly released 1050, 1050 Ti are leaked to be released soon. The same feature is said to be in the new cards with the same GPU but with faster memory and higher clock speed. GTX 2060 and GTX 2060, the btand new GeForce graphic cards will replace and succeed the existing GeForce GTX 1060 3GB and GTX 1060 6GB offerings from the NVIDIA economy range.The rumored refreshed parts are yet to arrive at some point amidst one year from now, more towards the end of summer and start of school season.
The GTX 2060 and GTX 2060 Ti are both anticipated that would highlight 6GB of GDDR5 memory and the same 192-bit memory interface. They are likewise expected that would dispatch at considerably more moderate MSRPs contrasted with their predecessor, departed siblings. NVIDIA’s revived $200-$300 GeForce 20 series Graphic cards are relied upon to clash with AMD’s fresh out of the brand new Vega 11 GPU. Vega 11 is AMD’s altogether new mid-run GPU in light of the new V9.0 design engineering.
The GPU will highlight second generation High Bandwidth Memory and has been outlined from the beginning to be both more power effective and faster significantly than AMD’s present Polaris 10 based RX480 design card. Little is thought about Vega 11 other than it’s required to supplant Polaris 10 one year from now. It’s anticipated to convey more than 7 TFLOPS of single accuracy figure contrasted with the 480’s 5.8 TFLOPS at the same time coming in at a lower TDP than 150W. NVIDIA’s invigorated GP106 GTX 2060 Ti and 2060 would confront furious rivalry from AMD’s Vega 11 should they dispatch one year from now.
Lineup of the NVIDIA`s GeForce 20 Series Pascal Refresh:
HBM2 & GDDR6 Memory are Being Supported By NVIDIA Volta:
Ahead of 2017, in 2018 Pascal refresh is set to be replaced by Volta. A more refined version of TSMC’s 16nm FinFET process will be based upon the new GPU architecture. Just like Maxwell before it the new architecture will focus more on pushing the performance per watt envelope without being in the help of a die shrink, same process being used as from the origin. Volta architecture is basically not much of an information for us. They are planning to introduce the GPUs pairing with GDDR6 as well as HBM2 memory. It is expected to boost the memory bandwidth by 60% of GDDR6 chip of 16GB comparing to its present GDDR5X 10Gbps memory chips, on which the company is pledging in its GTX 1080 and Titan X cards.
In 2018, NVIDIA GeForce 30 Series Launching with Volta GV102, GV104, GV110 GPUs
For 2018, Volta GPUs are planning for
- NVIDIA GV110 GPU
- NVIDIA GV104 GPU
- NVIDIA GV102 GPU
GV104 speaks to Nvidia’s bread and spread Volta gaming GPU. The organization will focus on the well known ~$400 portion with a chop down variation version of the chip and the top of the line with a completely empowered variation. As per what we know the organization won’t veer much from what we’ve seen it verifiable do with its 104 class chips. GV104 will supplant GP104, the chip that powers the current GTX 1080 and GTX 1070 cards, and the exceptionally same chip that self control their revived variations in 2017. GV104 based representation cards, i.e. GTX 3080 and 3070 cards will highlight 256-bit memory interfaces with 16GB of GDDR6 memory running at 16Gb/s for a sum of 512GB/s of bandwidth capacity, 32GB/s more than what’s accessible to the 384-bit GTX Titan X.
GP102 will be replaced by GV102 chips, currently powers the GTX Titan X, the chips that is expected to power the GTX 2080 Ti. the fastest gaming chip of NVIDIA for GV102 has to offer in the Volta Stack. On the other hand GV110 is produced for the professional market. Deep learning, data center workload and think AI inferencing. It will be more expensive and larger than the GV102 that will significantly feature better double precision compute and HBM2 rather than the GDDR6 memory. The high end server will grade Volta card that are expected to feature with 32GB of 1TB/s and HBM2 of memo bandwidth launching in 2018.
Via Source : TechStunt