Posted on Friday, May 18 2018 @ 11:25:26 CEST by Thomas De Maesschalck
Biostar launched a new series of iMiner systems for the cryptocurrency mining crowd. These systems feature AMD Radeon RX 500 series video cards in black server-like cases. They're fully equipped and ready to mine, the A564X12 features twelve Radeon RX 560 cards, the A578X6 has six Radeon RX 570 cards, and the A578XD flagship has eight Radeon RX 570 cards. The former two feature Biostar's TB250-BTC Pro motherboard, while the latter uses the TB250-BTC D+.
Each rig has the Intel Celeron G3930 dual-core processor and 4GB DDR4-2400. The flagship has a 120GB SSD, 1600W PSU and seven 120mm fans. The other two systems have a 90GB SSD, 1300W PSU, and five 120mm fans. For Ethereum mining, the iMiner rigs promise 148MH/s, 165MH/s, and 220MH/s, respectively.
Probably the funniest thing about this news is the way Biostar is marketing these rigs. AnandTech spotted the ads and notes Biostar is hoping to flog these systems to millenials who want to make an extra buck.
Biostar has this week released new iMiner systems specifically designed for crypto coin mining. The systems are stuffed with AMD RX 500-series video cards in a "space black" server-like chassis. Biostar positions the iMiner Multi-GPU Mining rigs as "fashionable and easy" along with dapper-looking young gentlemen in the promotional materials who look like they cashed in on the bitcoin bubble already.
Posted on Friday, May 18 2018 @ 11:08:38 CEST by Thomas De Maesschalck
EK introduces the EK-Vardar EVO 120ER RGB. The naming is pretty self-explanatory, this is a new Vardar fan with RGB LEDs. It's a PWM fan with double-ball bearing and it connects to a 4-pin 12V RGB header on your motherboard. The EK-Vardar EVO RGB fans spin at 500 to 2200RPM, offer up to 77CFM of airflow, up to 3.16mm H2O static pressure, and a noise level of 33.5dBA. The company sells these via its webshop for 21.90EUR a piece.
EK Water Blocks, the Slovenia-based premium computer liquid cooling gear manufacturer is releasing a new 120mm RGB version of its renowned EK-Vardar fan. Developed in-house and first introduced in 2014, the EK-Vardar fan is now also available as EK-Vardar EVO 120ER RGB. Beside the RGB lighting effect, the new fan also packs the Extended Range (ER) of PWM operation feature and the special Start-Stop function which brings a true 0dB experience for the absolute silent fan operation. The electronics of the new EVO 120ER RGB fan has been specially tweaked so that the fan stops spinning if the PWM signal falls below a certain duty cycle value. Not only will you be able to enjoy a totally silent PC, but this feature is also very useful for preventing radiator dust buildup in idling PCs and prolongs the lifecycle of the fan itself.
Posted on Friday, May 18 2018 @ 10:40:58 CEST by Thomas De Maesschalck
The pricing of NAND flash memory has been falling since the end of 2017, with price drops slowing to a level of around 10 percent in Q2 2018. However, industry sources believe strong demand from the mobile phone and PC markets may prevent NAND prices from falling further in Q3 2018. An expected uptick in demand may make the current oversupply situation less severe.
It's a typical boom-bust issue. Lower prices drive up demand, which in turn results in undersupply.
The fall in NAND flash prices has been accelerating the adoption of SSDs among PCs, the sources noted. ODMs have moved to promote models equipped with PCIe NVMe SSDs with volume production of their new devices slated to kick off in the third quarter, the sources said. Client SSDs are expected to transition from SATA to PCIe NVMe in 2018.
SSD prices had fallen about 50% from November 2017 to the end of the first quarter, Phison Electronics chairman Khein Seng Pua remarked recently. Prices are likely to rebound before or after the Computex 2018 trade show, and the supply of NAND flash will become tight again, according to Pua.
Posted on Friday, May 18 2018 @ 10:21:56 CEST by Thomas De Maesschalck
Intel's 10nm node is causing the firm a major headache as issues with the process resulted in multi-year delays and allowed AMD to catch up again. The 10nm issues were one of the talking points at Intel's 2018 Stockholders' Meeting.
At the event, Intel CEO Brian Krzanich said that he doesn't expect the 10nm issues would carry over into 7nm. In particular, Krzanich points out that Intel will set a less aggressive goal for 7nm than it did with 10nm, and he also hinted that EUV may make things easier:
And the simple answer is most of the problems that we encountered on 10-nanometer will not impact 7-nanometer and let me explain some of those. One 10-nanometer, we believe will be the last technology that we try and develop without EUV. EUV is the next generation lithography tool and if you understand semiconductor manufacturing lithography or the printing of the lines is the critical step in producing the product at the next smaller geometry. And we've been stuck without EUV for the longest period in the semiconductor industry history on a single type of lithography tool.
And 7-nanometer will be the first one that transitions to the new lithography tool, which then opens up our ability to print features that are much, much smaller, much more easily. So that's step one that's different between 10- and 7-nanometers. The other thing as we went back, if you remember for the earnings call, I said that one of the things we didn't ten 10-nanometers was took a much more aggressive scaling factor. We said we - instead of our typical 2.4, the industry actually goes between 1.5 and 2.
We went all the way up to 2.7 and that 2.7 is, you can just think of it as one over 2.7 as kind of being how much smaller you're trying to make it, caused a lot more interaction with the geometries as we try and make them so much smaller. At 7-nanometers, we go back to our 2.4 as an example. So we made changes like that to make 7-nanometer much more like our traditional technology innovations that we've done and hence we believe that most of the issues that occurred at 10-nanometers would not carry over into 7-nanometers. And we continue to monitor our progress on 7-nanometers and continue to be very positive about where we're at with 7-nanometers.
Guess we'll have to wait and see if Intel's manufacturing group can deliver on its promises this time!
Posted on Friday, May 18 2018 @ 09:55:28 CEST by Thomas De Maesschalck
Intel turns 50 this year and the firm is running a series that looks back on its past. In this edition, Intel reminiscences about the 1101, a piece of static random access memory that was the first high-volume metal-oxide semiconductor memory and the first chip to use silicon gates. This chip was made in 1969, it was not a commercial success but paved the way for the future:
Intel's 1101 static random access memory (SRAM) was the first high-volume metal-oxide semiconductor (MOS) memory and the first chip to use silicon gates.
The device was the result of a challenging development process. The conceptual groundwork for metal-oxide semiconductor memory had been established before Intel's founding, but no one was even close to a commercially viable mass production method.
The task of developing one was daunting.
Tom Rowe, Intel's first process engineer, recalled that the largest obstacle was getting a high die yield out of a wafer (Intel's goal was 20 die per wafer), a problem he feared might be intractable: "Every time we'd fix a problem, we'd uncover another one. … For all we knew the silicon gate process was no good. We'd make process change after process change, and many design changes, but it was still yielding only about two die per wafer, which meant it was a commercial disaster."
Nevertheless, the development team pushed forward with singular focus.
"Joel Karp and I redesigned a good portion of the product while the first moon landing was going on," designer Les Vadasz said. "We listened to 'one small step for man' on the radio while we scrambled to rework the chip."
Gordon Moore and Tim Rowe, meanwhile, began experimenting with chemical "dips" for the silicon to see if they could improve the materials bonding process.
The company's determination paid off in dramatic fashion one day when one of those chemical experiments unexpectedly produced a wafer that produced 25 die instead of the usual two, bringing the MOS well past its target yield in an instant. Everyone was astonished — when Vadasz learned what had happened, he began jumping up and down yelling "It's a super dip!" over and over.
"We proved that MOS silicon gate could make it that day," Rowe said.
The successful MOS SRAM process proved the wisdom of Intel's unusual approach to research and development during its first year. The company had focused on developing three different technologies at once — that way, the most viable new avenue of technology could present itself and the company would be ready. One avenue was bipolar SRAM, which led to Intel's first product, the 3101, but proved too easy to develop, so imitators followed suit and siphoned Intel's profits. A second avenue was multichip memory, which proved too difficult and was abandoned. The third avenue was MOS, which struck just the right balance of technical innovation and viability. Moore would later say Intel worked its way towards MOS through the "Goldilocks technology strategy."
The process technology developed for the 1101 became the industry standard and a key revenue source for Intel. And while Intel's first product, the 3101, had already established the company's ability to improve on established technologies and manufacturing processes, the 1101 established the company's ability to develop industry-changing new ones.
Posted on Thursday, May 17 2018 @ 20:33:35 CEST by Thomas De Maesschalck
Alphacool introduces the Eisblock GPX, a new RGB LED-infused graphics card waterblock. The company has models for AMD's Radeon RX Vega and NVIDIA GeForce GTX Titan X and GeForce GTX 1080 Ti video cards. These cost 149.79EUR but there's also an RGB LED-free version with plexi window for 119.79EUR.
The world-renowned manufacturer, Alphacool, offers the Eisblock GPX a very noble and high-quality full cover water cooler for your graphics card. The Eisblock GPX is available in two versions; An eye-catching version with a plexiglass window with RGB lighting to really show off or a simple but elegant version made of acetal.
Both coolers are based on an identical copper heat sink, which is completely nickel-plated. The surface area where the water flows is very large, so that all components that give off heat are actively cooled with water which increases performance. The 35 fins of the cooler are only 0.6 mm thick and cover even large GPUs completely. The large cooling surface is also ideal for graphics cards with HBM memory.
Backplates for both cooler variants are included, but they differ significantly. The acetal version uses a classic backplate coated with fins for better cooling. The plexi version of the Eisblock GPX water cooler uses a smooth backplate, which together with the cooler encloses the entire PCB. This makes the cooler and graphics card look like a single unit rather than a front and back plate. The SLI connector on Nvidia graphics cards is protected by a screwed cap, which can be removed if necessary.
The optical highlight of the Plexi variant, besides the very large window, of course also the complete and integrated RGB lighting. This is compatible with ASUS Aurora Sync, Biostar VIVID LED DJ, Gigabyte RGB Fusion or MSI Mystic. But pay attention to the connection on your mainboard. You may need a Male-to-Female adapter for some motherboards.
With the Alphacool EisBlock GPX full cover water cooler for your graphics card you get a high-quality cooler with excellent cooling performance. You have the choice between a simple but noble black acetal version or a striking, wonderfully illuminated and chic plexi version with a large viewing window.
Posted on Thursday, May 17 2018 @ 18:44:04 CEST by Thomas De Maesschalck
Tom's Hardware reports NVIDIA will finally start rolling out its next-gen video cards in July. An anonymous industry source told the site that the successor to the GTX 1080 will be released that month.
Initial cards will be NVIDIA Founder's Edition models, cards from third parties are expected to follow in August or September. The site also writes that laptop versions are anticipated later in the year.
The new GeForce cards are reportedly based on Turing.
These sources indicate that Nvidia will be delivering the GPU and memory over to partners on or around June 15th. We expect the company to deliver Founder's Edition Cards to retail sometime in July. New Quadro cards based on Turing could debut at Siggraph in August.
Posted on Thursday, May 17 2018 @ 16:27:54 CEST by Thomas De Maesschalck
Toshiba announced it finally received approval from China's anti-trust regulators for the sale of Toshiba Memory Corporation. A lot has been written about this subject over the past year or so, the main gist is that Toshiba saw itself forced to sell its NAND flash memory unit due to issues caused by other business units.
At the moment, Toshiba is no longer in desperate need of cash, so some investors were pushing the firm to cancel the deal. Gaining approval from Chinese regulators took so long that Toshiba is no longer contractually obliged to sell its memory unit to the Bain-led consortium. However, Toshiba isn't exploring alternatives and confirmed the deal is expected to close on June 1, 2018:
Corporation (TOKYO: 6502) has announced that it today confirmed
receipt of all required
anti-trust approvals in respect of the sale of Toshiba Memory
Corporation (hereinafter TMC).
On September 28 2017, Toshiba announced that it had entered into a share purchase
agreement (hereinafter SPA) with K.K. Pangea (hereinafter Pangea), a special purpose
company formed and controlled by a consortium led by Bain Capital Private Equity, in
respect of the sale of all
of the shares of TMC, and that it aimed to close the transaction
as soon as possible. The parties today confirmed that all required
anti-trust approvals have
been granted, and that all conditions for the closing of the transaction are now satisfied.
The parties will now take necessary procedures to close the transaction, which is currently
expected to occur on June 1, 2018.
Toshiba Group’s consolidated forecast for FY2018, ending March 31, 2019, as announced
on May 15, 2018 in “Toshiba Announces Consolidated Results for Fiscal Year 2017, to
March 31, 2018”, and in “FY2017 Consolidated Business Results”, assume the
completion of the sale of the Memory business by the end of 1Q of FY2018.
Toshiba will promptly announce any matters related to this transaction that require
Posted on Thursday, May 17 2018 @ 13:41:00 CEST by Thomas De Maesschalck
Jonsbo adds another RGB LED-infused processor cooler to its lineup. The new CR-201 RGB HSF features four 6mm diameter direct-touch copper heatpipes that dissipate the heat to an aluminium fin array. Basically, a typical tower heatsink but with some special design touches. The heatsink and heatpipes are black, and the fin stack features an acrylic top-plate with RGB LEDs.
Airflow is provided by a 120mm PWM fan with RGB LEDs. It spins at 800 to 1500RPM, pushes out up to 52.1CFM, and has a noise level of maximum 30.8dBA.
The cooler measures 129mm x 160mm x 94mm (W x H x D) and weighs 971g. It supports Intel LGA115x and AMD Socket AM4 processors, and is rated at up to 135W TDP. The pricing is unknown.
Posted on Thursday, May 17 2018 @ 13:34:23 CEST by Thomas De Maesschalck
TEAMGROUP introduced new memory modules for AMD's Ryzen 2000 series processors. The firm is now shipping T-FORCE DARK PRO memory kits with a frequency of up to 3466MHz. We did not receive further specifications, nor info about pricing or availabilty.
TEAMGROUP, the world’s leading memory brand, officially announces the launch of an advanced version of T-FORCE DARK PRO memory kits with new specifications for the 2nd Generation AMD Ryzen™ processors. With the rollout of new AMD processors, TEAMGROUP releases an upgrade specification of T-FORCE DARK PRO DDR4 gaming ram up to 3466 MHz. Its high performance and stability prove to meet the requirements of the gaming CPU and provide the best experiences for overclockers and gamers.
T-FORCE DARK PRO is built with high efficiency forged aluminum heat spreader. A unique personal style can be created by matching red and gray nameplates to its low-profile iron gray. T-FORCE DARK PRO also provides absolute stability and reliability with laboratory-scale burn-in test and compatibility verification for major motherboard brands worldwide. From the moment it was released, the popularity of T-FORCE DARK PRO has been on the top demand of DDR4 gaming market. Now T-FORCE is pleased to announce the new specifications DDR4 3466 MHz which push the envelope to the ultra-high performance which will satisfy all the gamers’ need for top-notch hardware gaming experience.
This all new T-FORCE DARK PRO DDR4-3466 MHz is a high performance memory powered by selected high quality original IC chips, which brings the performance to a new height for the 2nd generation AMD Ryzen™ and become the non-stop gaming conquest to raise the user experience throughout.