- Samsung calls to change processors microarchitecture - Toshiba OCZ TL100 - another new line of SSD-drives from Toshiba - Massive graphics NVIDIA Volta solutions limited to the type of memory GDDR6 - Opinions about positioning AMD Vega 20 split -

	xtreview is your : Video card - cpu - memory - Hard drive - power supply unit source

POSTER: computer news   ||   SAMSUNG CALLS TO CHANGE PROCESSORS MICROARCHITECTURE	DATE:2016-09-29
In the EE Times website pages were fragments from a fresh interview with the main strategy of the company Samsung Electronics - Jon Sohn , from which you can learn a lot about the company's plans.  About different discuss below, but for now focus on the main thing -  Samsung skeptical attitude to the future development of traditional computing architectures. As recognized by the Samsung chief strategist, his company is working on that can transform data processing centers.  Computers, according to Sona, should work like the human brain, representing a mixture of memory and compute modules, presented in the form of massive and parallelized network.  "Evaluators" should be located as close as possible to the place of storage, which will increase efficiency and reduce consumption calculations.  In other words, the new processor architecture should focus on memory, said a spokesman for the company, leading to DRAM- market and NAND-chips ... What nonsense when considered critical memory bandwidth, and x86-based processors have limited bus for communication with memory - roughly the thought was expressed by Samsung representative  .  What is important, the company itself is going to fix the situation.  This will be done in two ways: through internal projects, what is the Samsung chose not to spread, and by means of investing in start-up companies, for example, in British GraphCore.  Last working on a co-processor, accelerating machine learning. This project will help Samsung to challenge similar NVIDIA projects .  The company does not see anything impossible in what may eventually appear Samsung own server with a proprietary processor and an array of 32-terabyte SSD. Another company's desire to have the expectation of low power consuming and processor performance helmets for virtual and augmented reality.  In the current design helmets and glasses too heavy to gain popularity among the general public.  It requires a processor with an active consumption of less than one watt and better - at the level of 300 mW.  From the company has a desire to improve the characteristics of OLED-screens helmets.  It takes at least double the resolution, but for the beginning of production of small-format OLED with a resolution of 10K investment in production at the level of $ 5-10 billion is needed again. Another promising direction for itself, Samsung believes health.  Among other projects it is necessary to single out one, led the way, this infamous personality, Mary Lou Jepsen (Mary Lou Jepson).  That she was the main developer of reflective displays for student OLPC netbooks.  Then she created her company Pixel Qi, and recently moved to work in the Facebook group.  The new project, Mary Lou - personal MRI.  Worn System uses optical emitters and sensors that can be produced on the lines for the production of conventional LCD-panels.  So Samsung interest in this project is clear.  They promise that the working prototype of the device will be released next year.   Select Rating : Currently 10.00/10 1 2 3 4 5 6 7 8 9 10 45974 Rating: 10.0/10 (4 votes cast) Samsung calls to change processors microarchitecture	Add Comment ( 1 ) SAMSUNG CALLS TO CHANGE PROCESSORS MICROARCHITECTURE

POSTER: computer news   ||    TOSHIBA OCZ TL100 - ANOTHER NEW LINE OF SSD-DRIVES FROM TOSHIBA	DATE:2016-09-28
As soon as we come to terms with the emergence of Toshiba the A100 and Toshiba the OCZ VX500 ,  Toshiba brought out another novelty.   Formal positioning - the lowest in the range of Toshiba OCZ (even lower than the Toshiba OCZ TR150 - recently renamed OCZ Trion 150).  . Form Factor 2.5 "7mm, SATA3 interface and volumes of 120 and 240 GB read speed is stated up to 550 MB / s write - up to 530 MB / s Performance building block transactions with random access reaches 85,000 and 80,000 IOPS read and write. respectively. The term of the guarantee trademark is three years. The resource is set to 30, and 60 TB of written data (an average of about 27 and 54 GB per day). Select Rating : Currently 10.00/10 1 2 3 4 5 6 7 8 9 10 45973 Rating: 10.0/10 (9 votes cast) Toshiba OCZ TL100 - another new line of SSD-drives from Toshiba	Add Comment ( 1 ) TOSHIBA OCZ TL100 - ANOTHER NEW LINE OF SSD-DRIVES FROM TOSHIBA

POSTER: computer news   ||    MASSIVE GRAPHICS NVIDIA VOLTA SOLUTIONS LIMITED TO THE TYPE OF MEMORY GDDR6	DATE:2016-09-28
So, it is assumed that the first representative of the architecture Volta will be presented at NVIDIA GTC 2017 in May next year.  Just note that the head of the company will act in a similar industry event today, but we will speak about artificial intelligence and robotics, so the surprises from the scope of graphic solutions are unlikely. About graphics solutions Volta is known not so much.  According to one version, they will be releasing TSMC 16-nm technology, spent more on Pascal.  According to another version, the release of Volta to 14-nm technology could do Samsung .  Moreover, the Korean contractor can master the production of updated graphics Pascal second generation solutions, all using the same 14-nm technology.  The last will be released in 2017, and, according to Chinese sources, almost without exception acquire memory type GDDR5X, except for the cheapest GP107-type solutions. The first-born of the family will use the Volta HBM2 type of memory, but such decisions will only be used to accelerate the computing segment.  The GPU Pascal GP100 already equipped with memory of type HBM2, but for the consumer market it is too expensive.  The first representative of the Volta architecture can be equipped with 32 or 48 GB of memory HBM2. What will get less expensive video cards generation Volta?  Memory type GDDR6, which develops  Samsung .  Theoretically, it can operate at frequencies up to 14-16 GHz QDR, adequately replacing GDDR5X in 2018.  However, the main advantage GDDR6 will not be high performance and low power consumption.  This will be, in particular, to double the standard amount of memory in the consumer graphics cards without sacrificing heat.  Volta Video Card "middling" will be equipped with 16 GB of memory type GDDR6 and 256-bit bus and higher-performance graphics card with 384-bit bus to get 24-48 GB of memory. Select Rating : Currently 10.00/10 1 2 3 4 5 6 7 8 9 10 45972 Rating: 10.0/10 (9 votes cast) Massive graphics NVIDIA Volta solutions limited to the type of memory GDDR6	Add Comment ( 1 ) MASSIVE GRAPHICS NVIDIA VOLTA SOLUTIONS LIMITED TO THE TYPE OF MEMORY GDDR6

POSTER: computer news   ||    OPINIONS ABOUT POSITIONING AMD VEGA 20 SPLIT	DATE:2016-09-28
Familiarity with AMD plans to release for production of graphics solutions Vega and Navi took place a week ago.   According to the resource VideoCardz, Vega 20 graphics solution  is 7 nm successor of Vega 10, and will be released in the second half of 2018.  It attributed the presence of 32 GB of memory type HBM2, support for PCI Express 4.0 and TDP level not more than 150 watts. Fudzilla predicted in late 2018 or early 2019, with the first-born will be a 7-nm Navi Navi 10. It will be equipped with memory type HBM3. Select Rating : Currently 10.00/10 1 2 3 4 5 6 7 8 9 10 45971 Rating: 10.0/10 (11 votes cast) Opinions about positioning AMD Vega 20 split	Add Comment ( 1 ) OPINIONS ABOUT POSITIONING AMD VEGA 20 SPLIT

	xtreview is your : Video card - cpu - memory - Hard drive - power supply unit source

computer news computer parts review OLD Forum Downloads New Forum Login Join Articles terms Hardware blog Sitemap myspace proxy Get Freebies