First Intel 7 nm wonder comes to life, Raja Koduri reveals full die shot of the Ponte Vecchio Xe HPC 2-tile GPU with 8192 cores - Notebookcheck.net

proc.indah.link

Intel's Raja Koduri recently tweeted a photo of a 2-tile Ponte Vecchio Xe HPC die claiming that it includes seven advanced silicon technologies and called it a "thing of beauty". This Xe HPC GPU offers 16 clusters with 64 execution units (EUs) each for a total of 1,024 EUs i.e. 8,192 cores.

While Raja didn't exactly detail which seven advanced silicon technologies he was alluding to, Wccftech managed to get annotations for this die that offer more insights. Wccftech claims that it has verified this annotation from at least two of its sources. According to the publication, the seven advanced silicon technologies could likely be:

• Intel 7 nm process
• TSMC 7 nm process
• Foveros 3D packaging
• Embedded multi-die interconnect bridge (EMIB)
• Intel 10 nm Enhanced SuperFin process
• Rambo Cache
• HBM2 VRAM

On the top left and bottom right corners we have the Xe Link I/O tile fabbed on the TSMC 7 nm process. We also see two pairs of differently sized HBM2 dies on either side of the compute tiles. The compute units are fabbed on Intel's own 7 nm process, which means we are seeing Intel's new process technology for the first time in the wild.

Running in the middle of each of the 8-compute unit clusters is what appears to be the Rambo Cache fabbed on Intel's 10 nm Enhanced SuperFin process. Surrounding each of the clusters on three sides are passive die stiffeners that do not contain any logic.

Wccftech notes that beneath the tiles is a 10 nm base die while the EMIB interconnect is beneath the passive die stiffeners and the HBM2 memory. The whole package uses Intel's Foveros 3D packaging technology, so there's a lot more complexity to it than meets the eye.

Raja said that this particular package is now ready for power-on. Ponte Vecchio will launch some time in late 2021 or early 2022 and is primarily aimed at datacenter and HPC applications such as the Aurora exascale supercomputer.

At some point, we may also see this technology percolate down to consumer GPUs from Intel. Currently, Intel's consumer dGPU offering is the Xe DG1 with DG2-based Xe HPG expected later this year.

The Link Lonk


January 30, 2021 at 08:38AM
https://ift.tt/2YtTeaB

First Intel 7 nm wonder comes to life, Raja Koduri reveals full die shot of the Ponte Vecchio Xe HPC 2-tile GPU with 8192 cores - Notebookcheck.net

https://ift.tt/2YXg8Ic
Intel

The new Intel NUC 11 Panther Canyon packaging is just the cutest - Notebookcheck.net

proc.indah.link

One particular aspect we don't like about the new design is that it is significantly larger than the boxes from previous years even though the actual mini PC and its accessories remain small. It's not the most environmentally friendly way to package a product as a result. Maybe future NUCs can come in smaller houses as a compromise.

Expect our review on the NUC 11 in the coming weeks. The latest model is notable not just for its Core i7-1165G7 Tiger Lake CPU but also for its support for PCIe 4.0 NVMe drives and its built-in Qi wireless smartphone charger.

The Link Lonk


January 30, 2021 at 05:13AM
https://ift.tt/3pDbJoL

The new Intel NUC 11 Panther Canyon packaging is just the cutest - Notebookcheck.net

https://ift.tt/2YXg8Ic
Intel

Intel DG1 GPU Teardown, Failed Benchmarks, and Why It Won't Work on Most Systems - Tom's Hardware

proc.indah.link

German publication Igor's Lab has nailed a world-exclusive look at Intel's DG1 discrete graphics card. The chipmaker showcased the DG1 last year at CES 2020, running Warframe, but Intel's entry-level Iris Xe development graphics cards are exclusively available to system integrators and OEMs. In fact, Intel has put up some barriers in place to make sure that the DG1 only works on a handful of selected systems. Therefore, you really can't just rip out the DG1 from an OEM system and test the graphics card on another PC. After analyzing the images, the teardown helps explain why. 

Wallossek managed to get his hands on a complete OEM system with the original DG1 SDV (Software Development Vehicle). In order to protect his sources, Igor only shared the basic specifications of the system, which includes a Core i7 non-K processor and a Z390 mini-ITX motherboard.

First up, let's look at why the card won't work on most motherboards. 

Intel DG1

Image 1 of 2

(Image credit: Igor's Lab)

Image 2 of 2

(Image credit: Igor's Lab)

Intel has limited support for the card to a handful of OEM systems and motherboard chipsets, sparking speculation about why the company isn't selling the cards on the broader retail market. It turns out there's a plausible technical explanation. 

Hardware-hacker Cybercat 2077 (@0xCats) recently tweeted out (below) that the DG1 cards lack the EEPROM chip that holds the firmware, largely because they were originally designed for laptops and thus don't have the SPI lines required for connection. These EEPROM chips are present on the quad-GPU XG310 cards for data centers that use the same graphics engines, but as we can see in the naked PCB shot from Igor's Lab above, those same chips aren't present on the DG1 board. 

According to Cybercat 2077, that means the card's firmware has to be stored on the motherboard, hence the limited compatibility. Intel hasn't confirmed this hypothesis, but it makes perfect sense. 

While it's technically possible to shoehorn SPI eeproms on via some tricks, Intel has chosen not to do so on the DG1 OEM/Consumer cards. Image here of a 4chip Xe XG310 for hyperscalers where you can see an eeprom (red dot on them denotes pin 1) located next to each GPU chip. pic.twitter.com/Dq8HG4GLsrJanuary 28, 2021

Image 1 of 2

Intel DG1 (Image credit: Igor's Lab)

Image 2 of 2

Intel DG1 (Image credit: Igor's Lab)

The DG1 SDV reportedly features a DirectX 12 chip produced with Intel's 10nm SuperFin process node and checks in with 96 Execution Units (EUs), which amounts to 768 shaders. That's 20% more shaders than the cut-down version that Asus and other partners will offer. The DG1 features 8GB of LPDDR4 memory with a 2,133 MHz clock speed. The memory is reportedly connected to a 128-bit memory interface and supports PCIe 4.0, although it's limited to x8 speeds. 

At idle, the graphics card runs at 600 MHz with a power consumption of 4W. The fans spin up to 850 RPM and keep the graphics card relatively cool at 30 degrees Celsius. With a full load, the clock speed jumps up to 1,550 MHz, and the power consumption scales to 20W. In terms of thermals, the graphics card's operating temperature got to 50 degrees Celsius with the fan spinning at 1,800 RPM. Wallossek thinks that the DG1's total power draw should be between 27W to 30W.

The DG1 is equipped with a light alloy cover with a single 80mm PWM cooling fan and an aluminum heatsink underneath. Design-wise, the DG1 leverages a two-phase power delivery subsystem that consists of a buck controller and one PowerStage for each phase. The Xe GPU is surrounded by four 2GB Micron LPDDR4 memory chips.

Given the low power consumption, the DG1 draws what it needs from the PCIe slot alone and doesn't depend on any PCIe power connectors. Display outputs include one HDMI 2.1 port and three DisplayPort outputs. 

However, Wallossek noted that while you can get an image from the HDMI port, it causes system instability. He thinks that the firmware and driver prevent you from establishing a direct connection with the DG1, which explains why Intel recommends using the motherboard display outputs instead. The DG1 in Wallossek's hands is a test sample. Despite the many driver updates, the graphics card is still finicky, and its display outputs are unusable.

Image 1 of 3

Intel DG1 Benchmarks (Image credit: Igor's Lab)

Image 2 of 3

Intel DG1 Benchmarks (Image credit: Igor's Lab)

Image 3 of 3

Intel DG1 Benchmarks (Image credit: Igor's Lab)

The DG1's performance should be right in the alley of Nvidia's GeForce GT 1030, but there are no benchmarks or tests to support this claim. Wallossek couldn't provide any, either. Apparently, benchmarks simply crash the system, or they end up in an infinite loop. Wallossek could only get AIDA64's GPGPU benchmark to budge, but that doesn't really tell us anything meaningful about graphics performance.

The Link Lonk


January 30, 2021 at 10:24PM
https://ift.tt/39BBvEp

Intel DG1 GPU Teardown, Failed Benchmarks, and Why It Won't Work on Most Systems - Tom's Hardware

https://ift.tt/2YXg8Ic
Intel

The AMD Radeon RX 6700 XT leaks again with its logo, VRAM capacity, marketed gaming resolution and expected release dates outed - Notebookcheck.net

proc.indah.link

The Radeon RX 6700 XT has reared its head again, courtesy of Andreas Schilling. Taking to Twitter, Schilling published a photo of the RX 6700 XT series' logo that, presumably, has been cribbed from materials that AMD has sent to its board partners. Additionally, Schilling has hinted that AMD will market the RX 6700 XT series for 1440p gaming. By contrast, AMD currently advertises the RX 6000 series for 4K gaming.

Moreover, Schilling states that the RX 6700 XT series will have 12 GB of GDDR6 VRAM, reiterating previous leaks. As we have reported previously, AMD is believed to have based the RX 6700 series on Navi 22 GPUs with up to 40 Compute Units (CUs) and a 192-bit memory bus. Like the RX 6800 series, the RX 6700 XT is thought to have an XT GPU, and the RX 6700 a lesser XL variant.

The former is expected to offer a TGP of 186 W to 211 W, while AMD may fix the latter at between 146 W and 156 W. Furthermore, Moore's Law Is Dead claimed earlier this month that the RX 6700 XT series would have 64 MB or 96 MB of Infinity Cache and clock speeds in excess of 2.5 GHz.

Previously, Moore's Law is Dead asserted that AMD had scheduled to release the RX 6700 XT in May 2021, but Schilling has hinted that Navi 22 cards could arrive a few months earlier. According to Schilling, AMD will announce RX 6700 XT sometime in the first half of 2021 and possibly even as soon as February or March. Either way, the RX 6700 and RX 6700 XT should allow AMD to compete with the RTX 3060 and the RTX 3060 Ti, assuming that it and NVIDIA continue to match each other's product stacks.

JayzTwoCents may have provided a look at the RX 6700 series, too. In September, the YouTuber published renders of what were, at the time, broadly spoken of as RX 6000 series cards. The triple-fan card became the current RX 6000 series, so it should follow that the dual-fan card is the RX 6700, the RX 6700 XT, or both. If AMD releases the RX 6700 series within the next few months, we imagine more information will leak next month.

Purchase the Sapphire Nitro+ AMD Radeon RX 6900 XT on Amazon

The Link Lonk


February 01, 2021 at 01:47AM
https://ift.tt/36seHFh

The AMD Radeon RX 6700 XT leaks again with its logo, VRAM capacity, marketed gaming resolution and expected release dates outed - Notebookcheck.net

https://ift.tt/2ZDueh5
AMD

Intel’s New Processor Beats AMD’s Ryzen 5000 In Latest Benchmarks - Forbes

proc.indah.link

AMD (NASDAQ: AMD) has gone from underdog to having to the fastest processors available for gaming and content creators, but there are signs Intel has something fast in the works. It's 11th Gen desktop processors, which are expected to land on shelves in the next month or two, are beating AMD's fastest Zen 3 Ryzen 5000 processors in the PassMark benchmark.

The latest benchmarks to be posted in the PassMark database point at Intel having a lead in single ... [+] thread performance

www.cpubenchmark.net

The above tells us two things. Firstly, that Intel's new processors are likely to be faster in single-threaded and lightly-threaded applications than AMD's latest Ryzen CPUs. This is significant as it means that the work Intel has done to include an improved architecture as well as slightly higher frequencies has had a noticeable impact. The Rocket Lake-S CPUs you can see at the top of the chart - the Core i9-11900K and Core i7-11700K - aren't just a frequency boost. They sport a new architecture and that's significant. Intel is still using a 14nm manufacturing process here, but improvements in performance can come from areas other than lithography and frequency.

Secondly, we get more of an idea about these improvements looking at the results from Intel's 10th Gen CPUs lower in the graph such as the Core i9-10900K. This scores 3,174 points while the new Core i9-11900K scores 3,764. That's a sizable 19 percent performance boost and at frequencies that are only slightly higher than the previous generation. This indicates a big shift in performance that could be mirrored in multi-threaded workloads too as there seems to be much more going on here that just frequency-induced performance increases.

Motherboard manufacturers such as Asus, ASRock, Gigabyte and MSI are all launching new models that ... [+] use the Z590 chipset

Gigabyte

Along with 11th Gen CPUs, Rocket Lake-S will bring PCIe 4.0 support to current generation motherboards such as those that use the Z490 chipset, and also a new chipset - Z590. Motherboards using the Z590 chipset will see wide use of Thunderbolt 4 as well as native USB 3.2 Gen 2x2 support. However, The Core i9-11900K will only offer eight cores, while the Core i9-10900K had 10 cores. This means that AMD will retain a significant lead in multi-threaded performance on the mainstream desktop PC and it remains to be seen if Intel will counter this with future architectures or a renewed high-end desktop platform later this year.

MORE FOR YOU

I'’ll be reviewing Intel's new CPUs when they arrivce as well as looking at some of the interesting new Z590 motherboards so follow me here at Forbes, YouTube, Twitter, Facebook,  Instagram or Reddit for more PC hardware news and reviews.

The Link Lonk


February 01, 2021 at 01:28AM
https://ift.tt/2McNnnD

Intel’s New Processor Beats AMD’s Ryzen 5000 In Latest Benchmarks - Forbes

https://ift.tt/2ZDueh5
AMD

AMD: Time to Buy This Hot Growth Stock - Motley Fool

proc.indah.link

Advanced Micro Devices (NASDAQ:AMD) wowed Wall Street with its fourth-quarter results, posting huge year-over-year growth in revenue and earnings thanks to solid demand for its products across various verticals.

The chipmaker's revenue jumped 53% year over year in the fourth quarter to $3.24 billion. This was the first time AMD's quarterly revenue crossed the $3 billion level. Non-GAAP net income jumped 66% to $636 million, and earnings per share soared to $0.52 from $0.32 in the prior-year period. Analysts were expecting AMD to earn $0.47 per share on revenue of $3.03 billion.

AMD believes this terrific momentum will continue in the new year, primarily driven by growth in the data center, video gaming, and the personal computer (PC) businesses. These segments could help AMD deliver more upside in 2021. Let's see how.

Image source: Getty Images.

AMD's guidance sets the stage for a terrific year

AMD expects 37% revenue growth in 2021. That's impressive considering that the chipmaker delivered a 45% jump in the top line last year to $9.76 billion. Additionally, AMD expects a non-GAAP gross margin of 47% this year, up 2 percentage points over 2020 levels.

AMD's guidance is way better than Wall Street's expectation of 25.7% revenue growth for 2021. That's not surprising, as the catalysts that powered its growth in 2020 could get even better in the new year.

The computing and graphics business, for instance, is sitting on a couple of strong tailwinds. Accounting for nearly two-thirds of the top line, this segment's revenue shot up nearly 37% year over year on the back of strong sales of its Ryzen desktop processors.

AMD said on its latest earnings conference call that the sales of its new Ryzen 5000 desktop processors that were launched in October 2020 were more than double "the launch quarter sales of any prior-generation Ryzen desktop processor." On the other hand, sales of AMD's laptop-focused Ryzen 4000 chips increased in the double digits on a year over year basis, thanks to the 100 design wins the company had scored in early 2020 for these processors. In fact, AMD closed 2020 with record shipments of its mobile processors.

AMD is looking to turn things up a notch in 2021. The company recently released its Ryzen 5000 mobile processors based on the updated Zen 3 architecture. The company claims that the new laptop chips can deliver up to 23% performance gains and a longer battery life of up to 17.5 hours.

What's more, AMD management says that it aims to "increase the number of notebook designs powered by our new Ryzen 5000 processors by 50% compared to our prior generation." In simpler words, laptop and notebook original equipment manufacturers (OEMs) are going to launch more models with AMD processors.

Meanwhile, AMD looks set to take away more market share from Intel in the desktop processor space as well. Wall Street analysts expect AMD to boost its CPU market share to 50% in 2021 as compared to 22.4% in the third quarter of 2020, driven by the chipmaker's technological advantage over Intel.

The graphics business is also in fine form for AMD. The RX 6000 series graphics cards that the company launched late last year are witnessing overwhelming demand, with "launch quarter shipments three times larger than any prior AMD gaming GPU priced above $549." The chipmaker has been unable to produce enough cards to meet the strong demand so far, and it continues to ramp up production to address the supply constraint.

All of this indicates that AMD's computing and graphics business is about to have a solid year ahead.

More reasons to buy

AMD's enterprise, embedded and semi-custom (EESC) business accounts for the remainder of the company's top line, and it came into its own in 2020. The segment's revenue shot up 65% last year to $3.33 billion. Fourth-quarter growth was even more astronomic, with a year-over-year revenue spike of 176%.

The EESC business has stepped on the gas lately thanks to the arrival of Sony's and Microsoft's latest gaming consoles, which use AMD's chips. The company points out that sales of its semicustom chips used in the latest consoles are growing at a faster pace than the previous console cycle. More importantly, AMD expects sales of its semicustom chips to remain stronger than usual in the first half of 2021.

AMD is also upbeat about the server market, where customers such as Microsoft, Google, and Tencent are buying more of its EPYC server processors. The number of cloud instances powered by AMD's chips doubled in 2020 to more than 200. The company looks all set to carry forward this momentum in 2021 with the impending launch of a new generation of server chips that could help it take more share of the server market and substantially boost revenue.

So investors who missed AMD's impressive rally in 2020 have reason to cheer, as the good times can continue this year. What's more, they now have a good opportunity to buy this hot growth stock after its post-earnings dip and set themselves up for long-term gains.

The Link Lonk


February 01, 2021
https://ift.tt/39x1XPA

AMD: Time to Buy This Hot Growth Stock - Motley Fool

https://ift.tt/2ZDueh5
AMD

AMD Ryzen 5000-series CPU FAQ - Laptop Mag

proc.indah.link

Ryzen 5000 series is AMD’s newest CPU line. The arrival of Ryzen and its underlying Zen platform marked a comeback starting in 2017 for AMD in terms of competitiveness with Intel and a stronger approach to the laptop market. The company made serious inroads last year with its Ryzen 4000-series chips. Since then, the Zen microarchitecture has undergone two generational updates, landing us today at Zen 3’s 7-nanometer process.

Zen 3 and Ryzen 5000 laptops are due to market later this year, so if you’re looking for your next gaming or video rendering mobile workhorse, you’ll want to know why everyone’s talking about the latest Ryzen line.

We've put together a guide for AMD's new Ryzen 5000 CPUs to help you decide if these exciting new chips earn a place in your next laptop.

19% performance uplift, 24% more power-efficient

At AMD’s Ryzen 5000 launch event in October 2020, CEO Lisa Su and CTO Mark Papermaster outlined key improvements for Zen 3 over the prior Zen 2 generation. The standout being an increase in instructions per clock (a general marker of performance) with a 24% improvement to power efficiency, providing up to two hours more battery life. For mobile processors, especially, this is a dream combo that you don’t often get: more performance and longer battery life.

What makes Zen 3 next-gen

Zen 3 chips use a completely different layout of CPU cores surrounding a single L3 cache so all cores have faster and equal access to a single memory pool that’s twice the size of Zen 2’s total split cache. This both provides more resources to each CPU core and helps decrease latency across a wide array of tasks. Zen 3 also uses improved LPDDR4 memory to boost performance and reduce power consumption compared to Zen 2.

Asus ROG Flow X13 (Image credit: Future)

Zen 3 will continue to use the Radeon Vega series of integrated GPUs with improved clock speeds of between 300 and 400 MHz over Ryzen 4000 equivalents.

Ryzen 5000 U, HS and HX series

When looking for a Ryzen 5000 machine, you’ll want to know the basics of the naming scheme AMD uses to identify its processors. In general, higher numbers are better. For example, a Ryzen 9 5980HX represents the top-line CPU out of the whole bunch. The first number after "Ryzen" indicates one of three models: Ryzen 5, 7 and 9. The “5980” after the above example tells us the performance tier, which is ranged from 5600 up to 5980. 

Acer Nitro 5 (AMD, 2020) (Image credit: Future)

“HX” represents the CPU series, which tells us the intended category each CPU is designed for. U-series CPUs are for slim-and-lightweight notebooks. These systems offer the lowest performance, but at a great savings to battery life. HS and HX are the gaming and top-performance series, respectively, so expect to see them in larger notebooks. 

Some Ryzen 5000U-series CPUs still use Zen 2

It’s important to note that not all of the Ryzen 5000 laptop processor lines use the latest Zen 3 microarchitecture. The Ryzen 3 5300U, Ryzen 5 5500U and Ryzen 7 5700U still run off Zen 2. Essentially, the lower performance tier of each model number will be left behind on Zen 2. The higher-tier Ryzen 3 5400U, Ryzen 5 5600U and Ryzen 7 5800U will receive the Zen 3 upgrade. 

Acer Aspire 5 (Image credit: Laptop Mag)

As we noted above, the U-series is intended for low-power laptops with portability and battery life in mind. If you’re shopping for high-end performance, you’re not likely to be offered a U-series laptop, so just keep it in mind that not everything that says “Ryzen 5000” necessarily means “Zen 3” as well.

What laptops to look for with Ryzen 5000 CPUs

In February, the first Ryzen 5000 laptops will go on sale. At CES 2021, alongside the official debut of the mobile Ryzen 5000 line, several laptop models were unveiled that are sporting the new Ryzen family of processors:

Today's best gaming laptops deals

Show More Deals

The Link Lonk


January 31, 2021 at 02:00PM
https://ift.tt/3cmrMDM

AMD Ryzen 5000-series CPU FAQ - Laptop Mag

https://ift.tt/2ZDueh5
AMD

LGA RIP: Modder Transforms Old Intel PC Into Coffee Coaster - Tom's Hardware

proc.indah.link

As you can see in the video below, creator Cass 3D Designs turned his old PC motherboard into a coaster, with its LGA 775 socket and Intel chip front and center. 

When I retired my old PC from my student times, I cleaned it out and rebuilt it into the oh-so-beautiful Phanteks Evolv Shift 2, installed some old games on it that I used to play, and kept it on display in the living room. But not all old hardware still works, nor would everyone want to keep an entire PC around for nostalgia's sake.

After running across the parts of his first PC build in storage, rather than throwing them out, he wanted to turn it into a memento, inspired by a friend's project.

He used a Dremel to cut out a square of the motherboard, but the edges didn't end up looking so great. To cover these, he designed and 3D-printed a trim piece for around the motherboard to cover up the edges, and then hot-glued the two together. 

After test-fitting, he then made a silicon mould to put the board and trim piece together and then poured Epoxy resin into it to create the final product.

Of course, by this point, only half the work was done. Anyone that's worked with Epoxy will know that to create a nice clear view into it, the top layer needs to be sanded down, sanded down, and sanded down... and then polished to a crystal clear finish.

All things considered, this is a nice project to occupy some time without costing a ton of money -- this coaster will certainly be a conversation starter when friends are allowed over again. The only thing missing is being able to power up the chip to heat the coffee. And perhaps some RGB bling, too. 

The Link Lonk


January 30, 2021 at 08:15PM
https://ift.tt/3cqaZQ9

LGA RIP: Modder Transforms Old Intel PC Into Coffee Coaster - Tom's Hardware

https://ift.tt/2YXg8Ic
Intel

Intel DG1 GPU Teardown, Failed Benchmarks, and Why It Won't Work on Most Systems - Tom's Hardware

proc.indah.link

German publication Igor's Lab has nailed a world-exclusive look at Intel's DG1 discrete graphics card. The chipmaker showcased the DG1 last year at CES 2020, running Warframe, but Intel's entry-level Iris Xe development graphics cards are exclusively available to system integrators and OEMs. In fact, Intel has put up some barriers in place to make sure that the DG1 only works on a handful of selected systems. Therefore, you really can't just rip out the DG1 from an OEM system and test the graphics card on another PC. After analyzing the images, the teardown helps explain why. 

Wallossek managed to get his hands on a complete OEM system with the original DG1 SDV (Software Development Vehicle). In order to protect his sources, Igor only shared the basic specifications of the system, which includes a Core i7 non-K processor and a Z390 mini-ITX motherboard.

First up, let's look at why the card won't work on most motherboards. 

Intel DG1

Image 1 of 2

(Image credit: Igor's Lab)

Image 2 of 2

(Image credit: Igor's Lab)

Intel has limited support for the card to a handful of OEM systems and motherboard chipsets, sparking speculation about why the company isn't selling the cards on the broader retail market. It turns out there's a plausible technical explanation. 

Hardware-hacker Cybercat 2077 (@0xCats) recently tweeted out (below) that the DG1 cards lack the EEPROM chip that holds the firmware, largely because they were originally designed for laptops and thus don't have the SPI lines required for connection. These EEPROM chips are present on the quad-GPU XG310 cards for data centers that use the same graphics engines, but as we can see in the naked PCB shot from Igor's Lab above, those same chips aren't present on the DG1 board. 

According to Cybercat 2077, that means the card's firmware has to be stored on the motherboard, hence the limited compatibility. Intel hasn't confirmed this hypothesis, but it makes perfect sense. 

While it's technically possible to shoehorn SPI eeproms on via some tricks, Intel has chosen not to do so on the DG1 OEM/Consumer cards. Image here of a 4chip Xe XG310 for hyperscalers where you can see an eeprom (red dot on them denotes pin 1) located next to each GPU chip. pic.twitter.com/Dq8HG4GLsrJanuary 28, 2021

Image 1 of 2

Intel DG1 (Image credit: Igor's Lab)

Image 2 of 2

Intel DG1 (Image credit: Igor's Lab)

The DG1 SDV reportedly features a DirectX 12 chip produced with Intel's 10nm SuperFin process node and checks in with 96 Execution Units (EUs), which amounts to 768 shaders. That's 20% more shaders than the cut-down version that Asus and other partners will offer. The DG1 features 8GB of LPDDR4 memory with a 2,133 MHz clock speed. The memory is reportedly connected to a 128-bit memory interface and supports PCIe 4.0, although it's limited to x8 speeds. 

At idle, the graphics card runs at 600 MHz with a power consumption of 4W. The fans spin up to 850 RPM and keep the graphics card relatively cool at 30 degrees Celsius. With a full load, the clock speed jumps up to 1,550 MHz, and the power consumption scales to 20W. In terms of thermals, the graphics card's operating temperature got to 50 degrees Celsius with the fan spinning at 1,800 RPM. Wallossek thinks that the DG1's total power draw should be between 27W to 30W.

The DG1 is equipped with a light alloy cover with a single 80mm PWM cooling fan and an aluminum heatsink underneath. Design-wise, the DG1 leverages a two-phase power delivery subsystem that consists of a buck controller and one PowerStage for each phase. The Xe GPU is surrounded by four 2GB Micron LPDDR4 memory chips.

Given the low power consumption, the DG1 draws what it needs from the PCIe slot alone and doesn't depend on any PCIe power connectors. Display outputs include one HDMI 2.1 port and three DisplayPort outputs. 

However, Wallossek noted that while you can get an image from the HDMI port, it causes system instability. He thinks that the firmware and driver prevent you from establishing a direct connection with the DG1, which explains why Intel recommends using the motherboard display outputs instead. The DG1 in Wallossek's hands is a test sample. Despite the many driver updates, the graphics card is still finicky, and its display outputs are unusable.

Image 1 of 3

Intel DG1 Benchmarks (Image credit: Igor's Lab)

Image 2 of 3

Intel DG1 Benchmarks (Image credit: Igor's Lab)

Image 3 of 3

Intel DG1 Benchmarks (Image credit: Igor's Lab)

The DG1's performance should be right in the alley of Nvidia's GeForce GT 1030, but there are no benchmarks or tests to support this claim. Wallossek couldn't provide any, either. Apparently, benchmarks simply crash the system, or they end up in an infinite loop. Wallossek could only get AIDA64's GPGPU benchmark to budge, but that doesn't really tell us anything meaningful about graphics performance.

The Link Lonk


January 30, 2021 at 10:24PM
https://ift.tt/2MChqET

Intel DG1 GPU Teardown, Failed Benchmarks, and Why It Won't Work on Most Systems - Tom's Hardware

https://ift.tt/2YXg8Ic
Intel

AMD could outsource some of its APU and GPU production to Samsung in the near future - Notebookcheck.net

proc.indah.link

Please share our article, every link counts!

Anil Satyanarayana, 2021-01-30 (Update: 2021-01-30)

I've been an avid PC gamer since the age of 8. My passion for gaming eventually pushed me towards general tech, and I've been a stereotypical 'nerd' ever since. I have a degree in mechanical engineering and have worked in the manufacturing industry. When I'm not writing news on Notebookcheck, you can find me playing or watching an intense match of DOTA 2

The Link Lonk


January 30, 2021 at 09:08PM
https://ift.tt/3iZvVPl

AMD could outsource some of its APU and GPU production to Samsung in the near future - Notebookcheck.net

https://ift.tt/2ZDueh5
AMD