Page 973 of 990 FirstFirst ... 473873923948963968969970971972973974975976977978983 ... LastLast
Results 9,721 to 9,730 of 9891

Thread: Anandtech News

  1. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9721

    Anandtech: Intel's Xeon Cascade Lake vs. NVIDIA Turing: An Analysis in AI

    It seems like the new motto for Silicon Valley for the last few years has been “Data is the new oil,” and for good reason. The number of companies employing machine learning-based AI technologies has exploded, and even a few years after all of this has kicked off in earnest, those numbers continue to grow. This form of AI is no longer just an academic thesis or curious research project, but instead machine learning has become an important part of the enterprise market, and the impact on enterprise hardware – both purchasing and development – would be difficult to overstate. This is the era of AI.
    Today we’re taking a look at what’s perhaps the heart of Intel’s hardware in the AI space, Intel’s second-generation Xeon Scalable processors, better known as "Cascade Lake". Introduced a bit earlier this year, these new processors are still based on the same core Skylake architecture as the first-generation products, but incorporate a number of instructions and other modifications to speed up AI performance.


    More...

  2. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9722

    Anandtech: MediaTek Announces New Helio G90 Series SoCs: Gaming Focused Mid-Range

    Today MediaTek announces a new series in its product line-up: The new G-series starting off with the G90 and higher binned G90T. The new chips seemingly are a marketing exercise for MediaTek as it tries to battle again Qualcomm’s newest Snapdragon 7xx devices. The new G90 on paper very much looks like an update to the P90 which was announced late last year – updating the CPU and GPU IP whilst also slightly improving the camera capabilities of the chip.
    MediaTek Current P- & G-Series
    SoC Helio P90 Helio G90
    (Helio G90T)
    CPU 2x Cortex A75 @ 2.2GHz
    6x Cortex A55 @ 2.0GHz
    2x Cortex A76 @ 2.0GHz
    (2.05GHz)
    6x Cortex A55 @ 2.0GHz
    GPU PowerVR GM 9446 @ 970MHz Mali G76 MP4 @ 720MHz
    (800MHz)
    APU / NPU / AI Proc. / Neural IP 2x +140GMACs
    (Tensilica DSP)

    + In-house Inference Engine
    1127GMACs total
    2x APU

    +1TOPs total perf
    Memory 2x 16bit LPDDR4X @ 1866MHz LPDDR4X @ 2133MHz
    ISP/Camera 1x 48MP
    or
    2x 24+16MP
    1x 48MP (64MP)
    or
    2x 24+16MP
    Encode/
    Decode
    2160p30 H.264 & HEVC 2160p30 H.264 & HEVC
    Integrated Modem Category 12/13

    DL = 600Mbps
    3x20MHz CA, 256-QAM, 4x4 MIMO

    UL = 150Mbps
    2x20MHz CA,64-QAM
    Mfc. Process 12FFC
    The new G90’s main feature update is the switch from Cortex A75 cores to new Cortex A76 based IP. The new cores are clocked in lower at 2.0GHz, which is 10% lower than the 2.2GHz of the P90. The higher binned variant, the G90T, ups the frequency slightly higher by 50MHz at up to 2.05GHz. The big cores are accompanied by the same core config as on the P90- 6 additional Cortex A55 cores running at up to 2.0GHz.
    Another big change in the IP setup is that MediaTek is dropping Imagination’s 9XM core sin favour of a Mali G76. The new GPU comes in a MP4 configuration (MediaTek also likes to specify three execution engines per core), running at up to 720MHz in the G90 and 800MHz in the G90T. MediaTek promises 26% faster performance than its direct competitors, the competition here being Qualcomm’s Snapdragon 730.
    Part of the gaming theme of the SoC, MediaTek promotes its new “HyperEngine” game technology. It looks here that this is a response to Huawei’s GPU Turbo or Qualcomm’s own variant – it’s a plethora of software optimisations that promise to improve the experience of the phone. The most important aspect for MediaTek here seems to be the promise of 60% shorter rendering latency.
    In terms of the APU performance we don’t have immediate details of the G90’s, but it looks like things have remained relatively unchanged compared to the P90, with possibly Tensilica DSPs augmenting MediaTek’s own inference engine IP which is here stated to be able to operate at up to 1TOPs.
    On the camera department, the regular G90 remains the same as the P90 with up to 48MP single-sensor ability or 24+16MP multi-sensor capture. The G90T clocks the ISP higher and promises compatibility with the newest 64MP sensors that have been announced by vendors such as Samsung.
    The new chip continues to be manufactured on the 12nm FFC process and thus MediaTek should likely be able to price the chip extremely competitively against Qualcomm who has more advanced process node chips in this range.
    Related Reading:




    More...

  3. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9723

    Anandtech: Examining Intel's Ice Lake Processors: Taking a Bite of the Sunny Cove Mic

    Intel has been building up this year to its eventual release of its first widely available consumer 10nm Core processor, codenamed "Ice Lake". The new SoC has an improved CPU core, a lot more die area dedicated to graphics, and is designed to be found in premium notebooks from major partners by the end of 2019, just in time for Christmas. With the new CPU core, Sunny Cove, Intel is promoting a clock-for-clock 18% performance improvement over the original Skylake design, and its Gen11 graphics is the first 1 teraFLOP single SoC graphics design. Intel spent some time with us to talk about what’s new in Ice Lake, as well as the product's direction.


    More...

  4. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9724

    Anandtech: Honor Set to Enter TV Market with New Honghu 8K CPU

    Not being shy to take some gambles, this one appears to be its biggest yet: Honor is going to be developing both TVs and display scalers for the consumer market. Today at a briefing Honor disclosed some minor details about its new Honghu 818 chipset, which it states will be the driving force behind its push into consumer televisions.
    Gallery: Honor Set to Enter TV Market with New Honghu 8K CPU



    More...

  5. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9725

    Anandtech: GIGABYTE Aorus Liquid-Cooled GeForce RTX 2080 Super Launched

    GIGABYTE’s Aorus brand has introduced two liquid-cooled GeForce RTX 2080 Super graphics cards. The Aorus GeForce RTX 2080 Super Waterforce 8G features a hybrid closed-loop cooling system that is ready to go out of box. The Aorus GeForce RTX 2080 Super Waterforce WB 8G comes with a pre-installed water block and is designed to work with custom-built liquid cooling systems.
    Both graphics cards are based on NVIDIA’s TU104 GPU with 3072 cores clocked at 1615 MHz in base mode and up to 1860 MHz in boost mode. Compared to reference NVIDIA graphics boards powered by the same chip, GIGABYTE’s Aorus GeForce RTX 2080 Super Waterforce 8G feature seven video outputs, up from five. Both graphics cards feature liquid or hybrid cooling systems that cool down the GPU, memory, and VRMs.
    The Aorus GeForce RTX 2080 Super Waterforce 8G is a dual slot video card that is equipped with a hybrid cooling system featuring a 240-mm heat sink and two fans. Being aimed at enthusiasts, the board and the cooler feature addressable RGB LEDs that can be controlled using appropriate software from GIGABYTE.
    The Aorus GeForce RTX 2080 Super Waterforce WB 8 comes with a water block only and is meant for PCs with custom liquid cooling solutions. Looking extremely stylish, the water block has addressable RGB LEDs.
    GIGABYTE’s retailer partners will start to sell the new graphics cards in the near future. A clear advantage of the new units is their four-year warranty, which is especially valuable in Europe where a two-year warranty for electronics is a usual thing.
    Related Reading


    Source: GIGABYTE


    More...

  6. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9726

    Anandtech: Honor 20 Pro Launched in UK, +70% Prime Day 2019 Performance

    When Honor announced its new Honor 20 Pro in May, one of the confusing things was that despite the global launch being in the UK, the device wasn’t ready. Review samples were recalled, and uncharacteristically for the company, they couldn’t tell us when it would be ready. Part of this was down to how the US Commerce Department had listed Honor’s parent company, Huawei, on a trade blacklist that would require US companies to obtain licenses to work with them, which had knock on effects with Google and similar businesses. Fast forward a few months, and now the Honor 20 Pro is officially being released in the UK.



    More...

  7. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9727

    Anandtech: TSMC Announces Performance-Enhanced 7nm & 5nm Process Technologies

    TSMC has quietly introduced a performance-enhanced version of its 7 nm DUV (N7) and 5 nm EUV (N5) manufacturing process. The company’s N7P and N5P technologies are designed for customers that need to make then 7 nm designs run faster, or consume slightly lower amount of power.
    TSMC’s N7P uses the same design rules as the company’s N7, but features front-end-of-line (FEOL) and middle-end-of-line (MOL) optimizations that enable to either boost performance by 7% at the same power, or lower power consumption by 10% at the same clocks. The process technology is already available to TSMC customers, the contract maker of chips revealed at the 2019 VLSI Symposium in Japan, yet the company does not seem to advertise it broadly.
    N7P uses proven deep ultraviolet (DUV) lithography and does not offer any transistor density improvements over N7. Those TSMC clients that need a ~ 18~20% higher transistor density are expected to use N7+ and N6 process technologies that use extreme ultraviolet (EUV) lithography for several layers.
    While both N7 and N6 will be ‘long’ nodes that will be used for years to come, TSMC’s next major node with substantial density, power, and performance improvements is N5 (5 nm). The latter will also be offered in a performance-enhanced version called N5P. This technology will also feature FEOL and MOL optimizations in order to make the chips run 7% faster at the same power, or reduce consumption by 15% at the same clocks.
    Related Reading:


    Source: WikiChip.Org


    More...

  8. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9728

    Anandtech: AMD Releases New Chipset Drivers For Ryzen 3000: More Relaxed CPPC2 Upscal

    It’s been nearly three weeks since AMD’s launch of the new Ryzen 3000 series CPUs and our extensive coverage of the new parts. Among one of the things that didn’t quite go as smoothly is AMD’s BIOS and software situation where as things were still very much in flux following the launch.
    One issue that was repeatedly brought up by the community over the past weeks was the new CPU’s idle behaviour both in terms of temperature as well as voltages. In particular, the new parts seemingly looked like they rarely idled at lower performance states and instead looked to remain at high frequencies even when not doing much.
    While initially appearing as an issue, it really wasn’t one and rather just a side-effect of AMD’s new CPPC2 fast frequency ramp-up behaviour. Monitoring applications that are badly programmed tend to have a too heavy of a monitoring loop that causes load on the CPU – triggering a frequency ramp-up as the CPU is seeing a larger load. Given the new CPU’s sub-1ms ramp-up this meant that it was very hard to actually catch the machine at the lower frequencies – even though it most likely did idle correctly.
    AMD has now addressed this concern and tweaked the CPPC2 behaviour in the new Ryzen power plans with the release of a new chipset driver package.
    As AMD states in their community brief on the issue, part of the new behaviour change is that the new scheduler settings will now have a much more relaxed ramp-up time compared to the previous versions. In particular, when the chip will be at its base frequency and idling voltage, it will now take a significantly longer load for the chip to ramp up to its boost frequencies.
    In our quick A/B testing between the two driver versions, we can see that prior to the update the CPU would ramp up in around 840 microseconds to its boost clocks, whilst on the new power plan in this data-set took it a longer 17.5 milliseconds.
    The new behaviour thus should make the CPU ramp-up much less susceptible to smaller transient loads. The new boost duration is still very much adequate and extremely fast – sustained CPU workloads will see largely imperceptible difference, while intermittent workloads such as games also won’t be affected as once the CPU gets over the initial base frequency ramp threshold it maintains the sub-1ms frequency change behaviour.
    I also took a look at the Windows power plans if they changed, and it seemed that they indeed did. While on the old version the CPU would idle at ~2.2GHz, the new driver idles at 3GHz. Seemingly the frequency up-scaling has also been slightly slowed down as in my quick testing I saw frequency ramp up half as quickly.
    AMD has also addressed concerns about the reported high temperatures of the chip. The company explains that generally the value that most applications are reading out is the maximum of several sensors on the chip. Essentially this acts as the junction temperature of the chip – whilst most of the die would actually be a different/lower temperature.
    A new version of Ryzen Master now includes a different temperature readout algorithm that is meant to better represent the “overall” temperature of the die rather than the absolute maximum a sensor reports. AMD says this is a better representation of the temperature of the CPU. Besides averaging across different sensors, it also averages readouts over a small time-window. In my testing the most affected scenarios are idle and low-load scenarios and the new temperature behaviour isn’t nearly as erratic and spiky.
    Related Reading:





    More...

  9. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9729

    Anandtech: AMD Quarterly Earnings Report Q2 FY 2019

    AMD Quarterly Earnings Report Q2 FY 2019
    AMD announced their second quarter earnings for the 2019 fiscal year, and the company’s revenue was $1.53 billion for the quarter. This is down 13% from the same quarter last year. Gross margin improved from 37% to 41% year-over-year. Operating income was $59 million, down from $153 million a year ago, and net income was down $81 million to $35 million. This resulted in earnings-per-share of $0.03.
    AMD Q2 2019 Financial Results (GAAP)
    Q2'2019 Q1'2019 Q2'2018
    Revenue $1531M $1272M $1756M
    Gross Margin 41% 41% 37%
    Operating Income $59M $38M $153M
    Net Income $35M $16M $116M
    Earnings Per Share $0.03 $0.01 $0.12
    Although AMD was in the black for yet another quarter, this is certainly a dip that AMD does not expect to last. Their forecast for Q3 2019 is a 9% year-over-year increase in revenue to $1.8 billion, and they’ve recently launched new products that could help them achieve those goals.
    AMD Q2 2019 Computing and Graphics
    Q2'2019 Q1'2019 Q2'2018
    Revenue $940M $831M $1086M
    Operating Income $22M $16M $117M
    Looking back at Q2 though, Computing and Graphics revenue was down 13% to $940 million, and AMD attributes this drop to lower graphics channel sales. This drop was slightly offset though by higher client CPU and datacenter GPU sales. Also good for AMD and their investors is that their average selling price for client processors has increased thanks to more Ryzen sales, and GPU average selling price has also increased thanks to datacenter GPU sales. The Computing and Graphics segment had an operating income of $22 million for the quarter, compared to $117 million a year ago.
    AMD Q2 2019 Enterprise, Embedded and Semi-Custom
    Q2'2019 Q1'2019 Q2'2018
    Revenue $591M $441M $670M
    Operating Income $89M $68M $69M
    AMD’s other major segment is their Enterprise, Embedded, and Semi-Custom, and this product group also saw revenues fall 12% to $591 million for the quarter. AMD attributes this drop to lower semi-custom product revenue, which you can more or less read as console sales, and that makes sense since the current generation consoles are reaching the end of their life, but both Microsoft and Sony have both committed to AMD platforms for their next generation consoles, so expect this segment’s fortunes to get a bit better soon. Operating income was $89 million for this group, which was up from $69 million last year. The higher operating income is thanks to higher EPYC processor sales, which is also a great sign for this segment.
    Although this quarter’s revenue certainly saw a dip, AMD did just launch their latest third generation Ryzen this month, which wouldn’t be reported in their Q2 earnings which ended June 29th. As we saw in our review, this is a great step forward for AMD’s processor designs, and they have also launched their Navi based GPUs in July, so it makes some sense to see a dip prior to a major product launch. We’ll keep our eye on their results for Q3, but as previously mentioned they are expecting this to be a short-term drop, and with their new product lineup, that seems like a safe bet.
    Source: AMD Investor Relations


    More...

  10. RSS Bot FEED's Avatar
    Join Date
    09-07-07
    Posts
    32,363
    Post Thanks / Like
    #9730

    Anandtech: 32 GB Unbuffered DIMMs Listed from Seven Brands: DDR4-2400 to DDR4-3000

    Now that both Samsung and Micron are shipping their 16 Gb DDR4 memory chips to third parties, we're seeing wider availability of 32 GB unbuffered memory modules (UDIMMs). To date, six brands have either introduced, or started to sell their 32 GB unbuffered DIMMs, and in the coming months more manufacturers are expected to follow.
    The Basics

    Before we proceed, let us recap the basics here. Because of the way memory sub-systems work, high-capacity memory modules (in our case, 32 GB and higher) for workstations and servers are built differently than regular DIMMs for client PCs (which are called unbuffered DIMMs, or UDIMMs). Registered DIMMs (RDIMMs) carry a register chip that buffers the address and command signals, whereas the Load Reduced DIMMs (LRDIMMs) replace the register with an Isolation Memory Buffer that buffers the command, address, and data signals. While both the register chip and the IMB allow hardware vendors to build high-capacity memory modules and memory subsystems, neither RDIMMs nor LRDIMMs work with regular client platforms. Therefore, if you use a contemporary desktop and need a lot of memory for some reason, you'll need 32 GB UDIMMs.
    You can read more about contemporary types DIMMs an appropriate article covering different types of contemporary memory modules
    At the time of writing, 32 GB UDIMMs are supported by client platforms based on AMD’s 400 and 500-series chipsets as well as Intel’s 300-series chipsets.
    The List

    Twitter user momomo_us, who is from Japan, has managed to get a list of 32 GB unbuffered memory modules that are either available now or are about to hit the shelves there. The list is valid for Japan, yet we do know that ADATA and G.Skill are about to launch their 32 GB UDIMMs in the near future too, so our list includes data on these memory sticks as well.
    There are a number of remarks to be made about 32 GB memory modules. Samsung’s mass-produced 16 Gb DDR4 memory chips are rated for 2133 MT/s, 2400 MT/s, and 2666 MT/s data transfer rates, yet even the company itself sells 32 GB DDR4-2933 memory modules. Meanwhile, its partners go all the way to DDR4-3000, albeit at 1.35 Volts. Meanwhile, the only modules that semiofficially feature Micron’s 16 Gb DDR4 chips are rated at 2400 or 2666 MT/s, at 1.2 Volts.
    List of 32 GB Unbuffered Memory Modules Announced
    Note: Data is not official
    Brand Data Rate
    (MT/s)
    Latency Voltage DRAM Vendor PN More Info
    ADATA 2666 CL19 1.2 V Micron (?) AD4U2666732G19-B -
    Asgard 2666 CL16 1.2 V ? ? -
    Asgard 3000 CL16 1.35 V ? ? -
    Corsair 2400 CL16 1.2 V Micron ? -
    Corsair 2666 CL16 1.2 V Samsung ? -
    Corsair 3000 CL16 1.35 V Samsung ? -
    Crucial/Micron 2666 CL19 1.2 V Micron CT32G4DFD8266.16FB1 -
    G.Skill 4000 CL18 ? Samsung F4-4000C18-32GVR -
    Gloway 2400 CL17 1.2 V ? ? -
    Gloway 3000 CL16 1.35 V ? ? -
    Samsung 2666 CL19 1.2 V Samsung M378A4G43MB1-CTD -
    Samsung 2933 ? ? Samsung M471A4G43AB1-CVF -
    The Modules

    Now that we know the specs, let us talk about the modules themselves:

    ADATA
    What ADATA has shown so far were 32 GB DDR4-2666 CL19 DIMMs at 1.2 V. Considering the clock rate, these modules hardly need a heat spreader, yet knowing the company, we cannot exclude a possibility of enthusiast-class 32 GB UDIMMs with heat spreaders.
    Asgard
    Asgard’s Loki T2 and W2 memory modules are designed for enthusiasts, so they come with heat spreaders. The 32 GB DDR4-3000 CL16 modules need 1.35 V voltage and therefore need an enthusiast-class platform. Meanwhile, their 32 GB DDR4-2666 CL16 modules use industry-standard 1.2 V voltages.
    Corsair

    Corsair’s 32 GB Vengeance LPX unbuffered DIMMs come with DDR4-2400, DDR4-2666, and DDR4-3000 speeds. Depending on speed bins, these modules reportedly use memory chips from Micron or Samsung and require 1.2 V or 1.35 V. Corsair’s UDIMMs traditionally rely on the company’s custom 10-layer PCB designed to ensure quality signaling when operating at higher clocks, and are equipped with black heat spreaders.
    Crucial/Micron

    Micron’s Crucial brand introduced its DDR4-2666 CL19 32 GB UDIMMs back at Computex and these modules are expected to show up on the market shortly. Crucial’s 32 GB UDIMMs do not feature any heat sinks, but use industry-standard voltage, which makes them compatible with a wide variety of PCs.
    G.Skill

    G.Skill demonstrated its 32 GB UDIMMs at Computex, yet these modules are not available just yet. The company traditionally addresses enthusiasts with its products, which is why the memory sticks are rated for DDR4-4000 CL18 speed and are equipped with red heat spreaders. Given the unique combination of performance and capacity, expect G.Skill’s 32 GB unbuffered DIMMs to cost more than competing products.
    Gloway

    Gloway’s 32 GB UDIMMs are rated to operate at DDR4-2400 CL17 and DDR4-3000 CL16 speed bins, according to the listing. Depending on performance, the modules need 1.2 V or 1.35 V and come with heat spreaders.
    Samsung

    Samsung was first to start to producing 16 Gb DDR4 memory chips and was naturally the first to launch 32 GB UDIMMs. At present, the company offers DDR4-2666 CL19 and DDR4-2933 32 GB unbuffered DIMMs that come without any heat spreaders (based on the pictures of the modules from AVADirect).
    Related Reading:


    Sources: ADATA, Corsair, Crucial, G.Skill, Samsung, Twitter/momomo_us


    More...

Thread Information

Users Browsing this Thread

There are currently 8 users browsing this thread. (0 members and 8 guests)

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Title