Published: Wednesday, May 31, 2017 | By: Dennis
There is plenty of hot Computex news coming out of Taiwan and in following with my evaluation of the EVGA X299 Dark I figured it was time to look at the new GTX 1080 TI Kingpin edition.
For those of you who don’t know the Kingpin edition cards are specially designed “Classified” cards that were tweaked for LN2 Extreme overclocking. They work perfectly well on air however memory timings and the onboard VRM have been adjusted for high frequency overclocking and work better under those conditions. Of course to get the most from your Kingpin edition card you also had to find the special “unlocked” BIOS along with using an EVBot or the special voltage software used to control the power delivery. Neither were available direclty from EVGA.
Thing is with Pascal based chips the internal voltage controller limited how much you could do with the GPU and is one reason why we never saw a GTX 1080 Ti Classified, until now.
The 1080 Ti KPE features an all new design with a revised 10-phase VRM, PCI Express power connections on the back edge of the PCB (something we talked about in the Hardware Asylum Podcast), voltage test points, EVBot connector and a couple of new things including an SLI Style finger at the top and a pin block at the bottom. They are claiming that every card is guaranteed to overclock to 2025Mhz out of the box meaning that the GPUs have been prequalified to run a certain speed. I would fully expect that to warrant a price premium similar to what EVGA did with the 980 Ti KPE cards.
Of course the photo I found is of a bare PCB. Final versions will feature a triple fan cooler like on the FTW3, copper plated heatsink (like on previous KPE cards), ICX cooling technology so you can monitor temps of everything on the board, and cast style backplate for that added EVGA style.
Funny thing about the backplate. They were always “hop up” parts before everyone started making them standard and now that every card is expected to have a backplate EVGA has gone over the top with a cast style plate with even more thermal contacts.
Current rumors are that this card will be available in July so get your LN2 containers ready.
Published: Tuesday, May 30, 2017 | By: Dennis
A close look at the EVGA X299 Dark
Computex 2017 is in full effect and while I was unable to attend this year there is still plenty to talk about and do some armchair analysis.
The first up is the new X299 Dark from EVGA. As many of you may have noticed the X299 is the chipset supporting the new generation of Skylake-X and Kaby Lake-X processors which promises more PCI Express lands, More Cores and generally more performance. The EVGA Dark has traditionally been a ultra high-end motherboard design that sits above the Classified with a unique spin on how Overclocking and Multi GPU configurations are handled.
I’m sure some more accurate information will be available as reports from Computex get released but, until this let's speculate.
The first thing you will notice is the large MCP/Chipset cooler with a very large blower style fan. It is unlikely that the X299 chipset will require any cooling however M.2 drives typically do. By the looks of it exhaust air from the blower will pass over the M.2 drives to help regulate their temperature.
VRM is located at the top of the motherboard between the banks of DDR4 memory. This motherboard only comes with four memory slots and as we have learned from past motherboard designs, if you want high speed memory you have to drop the extra “crybaby” DIMM slots. As a result 4133Mhz is supported.
Along all of the sides you will notice every connector is set at 90 degrees with several of them “cut” into the motherboard making them slightly recessed. You’ll notice dual debug LEDs, I suspect dual temperature montitors OR, maybe a selectable display where it would show everything from POST codes to CPU temp, CPU voltage, VRM temp and maybe even frequency. Once we see the X299 Dark in action that speculation should become clear.
I also see triple BIOS, PCI Express disable switches, dual U.2 connectors, onboard power and reset and two very strange connectors located under the last PCI Express slot.
- Form Factor – Optimized EATX for clean cable management
- Supports x16 PCIe Graphics with all CPUs
- Audio – Creative Sound Core3D Quad-Core Audio Processor + Optical
- Highly-Efficient 14 Phase Digital VRM
- 12 Layer PCB Design
- Bundled Bench Stand Plate
- Dedicated M.2 Cooling
- Onboard Power, Reset and Clear CMOS
- EVGA E-LEET X Tuning Utility
- Triple BIOS Support
- M.2 NVMe PCI-E SSD Support
- U.2 NVMe SSD Support
- Intel® Optane™ Memory Ready
- Switchable DC / PWM Fan Control
- 10-zone Temperature Monitor
- Onboard Voltage Display
- Intel® X299 Express Chipset
- SLI – Up to 4-Way
- Memory Support – 4 DIMM Quad-Channel DDR4 4133MHz+ (Up to 64GB)
- Ethernet – 2x Intel® Gigabit NIC
Saying I am excited to see this motherboard in action is an understatement!
Published: Wednesday, May 24, 2017 | By: Dennis
One of the coolest things to watch during Computex is the GSKill world stage overclocking demonstrations. It really serves no purpose than to draw in crowds and give the press something to take pictures of BUT, it is still fun to watch.
Taipei, Taiwan (19 May 2017) – G.SKILL International Enterprise Co., Ltd., the world’s leading manufacturer of extreme performance memory and gaming peripherals, is thrilled to announce its largest annual LN2 extreme overclocking events at Computex – the 6th Annual OC World Record Stage 2017 and 4th Annual OC World Cup 2017. In addition, an Extreme OC Workshop event co-hosted with HWBOT.org at the G.SKILL booth this year. These 3 events will be held throughout Computex week at the G.SKILL Computex booth in the Nangang Exhibition Hall, 1F, Booth J0806.
6th Annual OC World Record Stage 2017
The G.SKILL OC World Record Stage event has reached its 6th year in the running and has become one of the biggest annual overclocking events in the community. This year, 18 world-class professional overclockers will be attending and aim to break as many benchmark records as possible, showing off the astounding performance of the latest hardware from Intel, leading motherboard vendors, and G.SKILL.
4th Annual OC World Cup 2017 – $10K USD for the Champion
G.SKILL OC World Cup awards the highest single cash prize in overclocking competition history. The contest rules and format are specifically designed for extreme memory overclockers. The top 6 extreme overclockers qualified from the online qualifier competition stage on HWBOT will compete live at the G.SKILL booth and demonstrate advanced LN2 overclocking skills at its finest. The Grand Final OC Champion will be walking away with the ultimate $10,000 USD cash prize.
Extreme Overclocking Workshop
This year, in addition to the two G.SKILL annual overclocking events, G.SKILL joins forces with HWBOT and co-hosts a LN2 extreme OC workshop at the G.SKILL booth. This workshop gives Computex visitors the rare opportunity to experience LN2 overclocking and try to achieve record-breaking high-speed CPU frequency or DDR4 frequency, guided step-by-step by professional overclockers.
Published: Wednesday, May 24, 2017 | By: Dennis
One of the biggest cases I ever reviewed was the Cooler Master Cosmos II. The thing was huge and did plenty of things right from the outward styling, retention of what made the Cosmos such a popular case and the ability to hold just about any system you could think of.
To help mark their 25th birthday Cooler Master has re-released the Cosmos II 25th Annivesary edition and while it looks almost identical there are some rather dramatic changes to help make the design a little more modern.
The Cooler Master Cosmos II was launched in 2012 and became one of the brand's most recognizable enclosures. Now, half a decade later, Cooler Master presents the Cosmos II 25th Anniversary edition to celebrate a quarter century in business and to bring the iconic chassis back to a new generation of enthusiasts and gamers.
I think I know what my Throwback Thursday will be That is asuming I can get back to where the Cosmos II is currently chillin.
Published: Wednesday, May 24, 2017 | By: Dennis
Everyone loves an inexpensive chassis proof of that can be seen in the steady declining PC component sales and the endless number of reddit threads asking for help building a "budget pc".
For a hardware enthusiast budget is often secondary over component choice which doesn't necessarily remove cost from the equation but brings style, features and compatibility forward to ensure the selected product will match the need.
ModSynergy is taking a look at the CMT210 case from FSP. It isn't a bad looking case but does seem to be "sampled" from a few older Cooler Master designs. Maybe that is what CMT stands for.
I’ll be reviewing FSP’s first ever PC chassis with their CMT210 gaming oriented ATX mid-tower PC case. The CMT210 aims to provide those on a budget a modern PC chassis with lots of cooling capabilities, a sleek interior layout providing lots of expansion abilities including showcasing your SSDs up front, support for both tall air coolers and water cooling, support for lengthy high-end graphics cards, and closing it up with a clear side panel. All of this can be had for a competitive $59.99 USD and I’ll be testing the FSP CMT210 to see if it makes your money go the extra mile and is a worthy platform to invest in.
Even if this case is not for you, you cannot fault them for the lack of photos.
Published: Thursday, May 18, 2017 | By: Dennis
Gorgeous Tempered Glass Case, space for two vertical graphics cards, featuring USB 3.1 Type-C
May 18, 2017, Keelung, Taiwan - Lian-Li Industrial Co. Ltd launches the PC-O12; a compact mid-tower chassis that combines sleek tempered glass panels with strong, but lightweight, steel and aluminum. This new addition to Lian Li’s latest generation O-series chassis range offers unsurpassed style, plus slim design with ample space for a powerful but compact PC build. Thanks to its unique design, it offers space for two vertically placed graphics cards in a separate compartment for gorgeous PC builds.
Tempered glass adds a touch of class
The PC-O12’s flawless tempered glass front and side panels make it a sleek and sophisticated showcase for the latest cutting edge computing technologies. Tempered glass is tough, safe and very durable, providing a ‘fresh from the showroom’ appearance indefinitely. The PC-O12’s alluring black aluminum outer body and panels complete the picture. Internally, a rigid steel frame provides a firm foundation for state of the art features.
Ideal balance of chassis size and features
Despite it’s space-saving format, this mid-tower enclosure offers plenty of room for the most powerful hardware. The 440mm full bandwidth PCI Express 16x riser cable allows flexible vertical graphics card mounting to enhance cooling and to show off the latest graphics technology through the tempered glass side panel. The roomy case interior fits graphics cards up to 340mm long and CPU coolers up to 75mm high.
There’s internal space for up to eight hard disk and SSD drives for terabytes of fast storage capacity. In addition, the newest ultra speedy, powerful external USB 3.1 type C devices are supported, and there are a total of four external USB connectors as standard.
A case with great low-noise cooling performance
With up to five large-format fans, this chassis ensures valuable PC components keep running cool, prolonging life, enhancing performance and reducing noise. There’s space for three 120mm fans at the top of the case, plus two 140mm or 120mm fans at the front. With so many airflow options, users are able to reduce fan speed and reduce noise.
In addition, removable mesh dust filters cover the primary fan mounts. The drive cages and PSU mount include rubber vibration dampeners to minimize noise.
Price and Availability
The PC-O12 is now available at Newegg for $399.99
Additional PCI Express riser cables are available at Performance PC starting in June 2017
About Lian Li
Lian Li Industrial Co. Ltd is an Aluminum Chassis and Chassis peripheral manufacturer and provider of industrial wholesale as well as OEM & ODM services for the PC industry as required. With over 20 years of service excellence, ISO 9001 certification and a team of professional staff willing to do the utmost to deliver durable, superior quality products, all backed by a *2 year guarantee; your satisfaction is guaranteed (*conditions apply). To learn more about Lian Li, please visit the official website: http://www.lian-li.com/. You can also visit Lian Li on Facebook or follow Lian Li on Instagram: @LianLiHQ
Published: Monday, May 15, 2017 | By: Dennis
This information has been making the rounds today after having been posted in the Anandtech forums last Friday. It would seem that Skylake-X and Kaby Lake-X processor specs have been "leaked" and instead of keeping with the familiar Core i7 name they changed it to.
Now, if memory serves Intel swore up and down and even sent some "thumb breakers" to editors claiming that Core i9 was going to happen.
The Intel response was "Stop trying to make Core i9 happen, its not going to happen"
Well, the joke is on 'cause we got screenshots claiming otherwise. Here is the breakdown.
Oh, and a photo of a screen showing a powerpoint slide with all of the info.
I would say that this is rumor mill at best. In the photo we can see a rather messy "desktop", I mean who moves their icons away from the start shortcut (it’s not a button, at least anymore) and still retains credibility in their intelligence?. On top of that, who actually uses an ACER monitor(s) anymore?? Maybe some engineer in Taiwan but you'd think that one of the largest chipmakers in the world could spring for a decent panel.
Of course, the last and final clue is the Core i9 name. Let us assume that Intel is keeping with their tagline that Core i9 isn't going to happen so, why release it now? I can see where it might work with the top end chips given that it is 12 cores and comes with 44 PCI Express lanes but, everything else on that list currently has a Core i7 counterpart.
Then again, maybe this is how Intel will regain some of their consumer cash by saying Core i7 is "Consumer" while Core i9 is "Professional". If that was their play then Intel could jack up the price and effectively stop consumers from using anything based on the X299 chipset in the retail space.
I suspect we will find out more come Computex when the first X299 motherboards are released along with processor details.
Published: Monday, May 15, 2017 | By: Dennis
I will be the first to admit that I suck at memory overclocking. I have overclocked memory, tweaked timings and playing with frequency but for the most part that is the extent of what I do.
And, the reason? Well, it takes a LONG time to fully overclock a memory module and unlike with motherboards where very little on the board will physically block you from overclocking a memory module depends on chip quality, CPU quality and CPU Memory Controller quality. All of which can kill your overclock and you may not know which is the culprit.
When you consider the "real world" benefits to memory overclocking some tweaks are difficult to attain and net you no benefit which is why you go after the big gains like frequency and tune from there. The smaller tweaks are usually best served when doing competitive overclocking like what is being discussed over at MadShrimps.
Since the launch of the Intel Sky Lake platform, we were being flooded with a multitude of different DDR4 memory ICs. After you finally decided to buy that particular tweakable kit, there were again screenshots popping up of something new and at first glance even better performing memory. At launch date Hynix was king of the hill for the X99 quad channel setups. however the latest memory architecture release for Sky and Kaby Lake are the Samsung B-DIE ICs. When tweaked, they are able to run at high speeds in combination with ultra tight timings. These are a real treat to gain them extra precious points to climb in the HWBOT ranking. Besides the fact that the latter are 8GB modules and thus a bit more pricey than the other 4GB variants, one drawback is that these B-DIE based memories are still pretty inconsistent in quality. Wading through different OC forums we noted DOA's, degraded sticks and even many sticks just failing to do tight timings at what we consider to be standard OC speeds, hence why memory binners jumped on the wagon and are selling, logically for a little extra margin, sticks that can truly deliver the goods. Today we explore what memory is great and which one is the best for your high end Intel LGA 2011-3 platform.
Some good stuff over there be sure to check it out.
Published: Thursday, May 11, 2017 | By: Dennis
This is really exciting.
SAN JOSE, Calif.—Nvidia refreshed its lineup of GPUs for deep learning and artificial intelligence applications on Wednesday with the new 5,120-core, 7.5 teraflop Tesla V100 Volta.
The new processor is part of Nvidia's quest to come up with a new way to consistently improve computing performance in the aftermath of Moore's Law, which many industry leaders agree is pretty much dead. Instead of boosting processor speeds or cramming more transistors onto already-crowded silicon, Nvidia is championing GPU-accelerated computing, which the company's CEO Jensen Huang (pictured above) said can offer a 150 percent performance boost every year.
The Tesla V100, with a brand-new architecture called "Volta," represents that latest boost. It's the "next giant leap into the new world" of AI and high-performance computing, Huang said at Nvidia's developers conference here. The V100 will start shipping by the end of the year to data centers owned by Amazon, Microsoft, and other cloud computing providers in several different configurations.
Be sure to check out the entire story over at PCMag. According to NVIDIA they are building a platform for deep learning which is the precursor to AI and other exciting and scary things. From a hardware standpoint this new GPU is pretty amazing as it features HBM2 memory, 5120 CUDA cores and is built on a 12nm process. All of that fits on a small board that can then be scaled in datacenter systems with no upper level.
Published: Wednesday, May 10, 2017 | By: Dennis
NVIDIA is on a roll these days. First with the launch of Pascal, the new GPU architecture powering the GTX 1000 series GPUs and Titan Xp and now with a quarterly earnings revenue of $1.94 billion.
Basically a 48% gain.
I wouldn't expect every NVIDIA fanboi to follow the stock market but a revenue of almost 2 billion dollars tends to get investors excited and they immediately start dumping money into the stock. That makes good sense for everyone including existing stock holders who are just watching the numbers go up.
The result is something like this.
Nvidia stock is soaring, up more than 14% in after-hours trading, beating Wall Street’s expectations for its first quarter.
The graphics processing unit (GPU) company announced its first quarter earnings today with revenue of $1.94 billion, a 48% gain year over year. While gaming drives most of the company’s sales, its largest area of growth over the past year was selling GPUs that power the artificial intelligence and graphics processing in datacenters. That sector of business saw 186% growth year over year, just a dip below last quarter’s 205% growth.
Lets hope that NVIDIA can keep this going. I suspect things will slow down for awhile given that all of the major verticals have been released but, you never know.
Of course that doesn't include automotive and datacenter systems unless you count the $1200 Titan Xp and the Tesla V100 and variety of configurations you can buy it in.