News Posts matching #GA102

Return to Keyword Browsing

NVIDIA's Next-Gen Big GPU AD102 Features 18,432 Shaders

The rumor mill has begun grinding with details about NVIDIA's next-gen graphics processors based on the "Lovelace" architecture, with Kopite7kimi (a reliable source with NVIDIA leaks) predicting a 71% increase in shader units for the "AD102" GPU that succeeds the "GA102," with 12 GPCs holding 6 TPCs (12 SMs), each. 3DCenter.org extrapolates on this to predict a CUDA core count of 18.432 spread across 144 streaming multiprocessors, which at a theoretical 1.80 GHz core clock could put out an FP32 compute throughput of around 66 TFLOP/s.

The timing of this leak is interesting, as it's only 3 months into the market cycle of "Ampere." NVIDIA appears unsettled with AMD RDNA2 being competitive with "Ampere" at the enthusiast segment, and is probably bringing in its successor, "Lovelace" (after Ada Lovelace), out sooner than expected. Its previous generation "Turing" architecture saw market presence for close to two years. "Lovelace" could leverage the 5 nm silicon fabrication process and its significantly higher transistor density, to step up performance.

NVIDIA Announces RTX A6000 48 GB Professional Graphics Card Accelerators

NVIDIA today announced their RTX A6000 series of graphics cards, meant to perform as graphics accelerators for professional workloads. And the announcement marks a big departure for the company's marketing, as the Quadro moniker has apparently been dropped. The RTX A6000 includes all raytracing resources also present on consumer RTX graphics cards, and marks a product segmentation from the company's datacenter-geared A40. The RTXA6000 features a full-blown GA102 chip - meaning 10752 CUDA cores powering single-precision compute performance of up to 38.7 TFLOPs (3.1 TLFOPs higher than that of the GeForce RTX 3090). Besides offering NVIDIA's professional driver support and features, the RTX A6000 features 48 GB of GDDR6 (note the absence of the X) memory - ensuring everything and the kitchen sink can be stored in the cards' VRAM. GDDR6X doesn't currently offer the per-chip density of GDDR6 solution, hence why NVIDIA opted for the lower-performing, yet denser memory variant.

The RTX A6000 features a classic blower-type cooler, and presents a new low-profile NVLink bridge that enables two of them to work in tandem within the same system. NVIDIA vGPU virtualization technologies are supported as well; display outputs are taken care of by 4x DisplayPort connectors, marking the absence of HDMI solutions. The card is currently listed for preorder at a cool and collected $5,500, but with insufficient silicon to offer even to its highest-margin datacenter customers, it remains to be seen exactly how available these will be in the market.

NVIDIA GeForce RTX 3080 Ti Graphics Card Launch Postponed to February

In the past, we heard rumors about NVIDIA's upcoming GeForce RTX 3080 Ti graphics card. Being scheduled for January release, we were just a few weeks away from it. The new graphics card is designed to fill the gap between the RTX 3080 and higher-end RTX 3090, by offering the same GA102 die with the only difference being that the 3080 Ti is GA102-250 instead of GA102-300 die found RTX 3090. It allegedly has the same CUDA core count of 10496 cores, same 82 RT cores, 328 Tensor Cores, 328 Texture Units, and 112 ROPs. However, the RTX 3080 Ti is supposed to bring the GDDR6X memory capacity down to 20 GBs, instead of the 24 GB found on RTX 3090.

However, all of that is going to wait a little bit longer. Thanks to the information obtained by Igor Wallosek from Igor's Lab, we have data that NVIDIA's upcoming high-end GeForce RTX 3080 Ti graphics card is going to be postponed to February for release. Previous rumors suggested that we are going to get the card in January with the price tag of $999. That, however, has changed and NVIDIA allegedly postponed the launch to February. It is not yet clear what the cause behind it is, however, we speculate that the company can not meet the high demand that the new wave of GPUs is producing.

NVIDIA RTX 30-series GA102 IR Photographs Appear, Expose Silicon Inner Workings

IR photographs are one of the best ways to take a look at the inner silicon etchings inside the world's most powerful accelerators. Of course, one could always just "shove off the top", but that would be a crime unto itself. Fritchens Fritz has pointed an IR gun at the heart of NVIDIA's GA102, the crown jewel of the RTX 30-series, and the results are an intimate look at some of the world's most powerful hardware. The 642 mm² chip showcases all of its precision-engineered resources that allow gamers to achieve (almost) 60 FPS in 4K resolution in Cyberpunk 2077, with rows upon rows of transistor building blocks mirroring NVIDIA's (representative) chip resource breakdown in the Ampere whitepaper.

The seven vertical rows showcase the GA102's Graphics Processing Clusters (GPCs) - and in each of those rows, one can count 12 Streaming Multiprocessors (SMs). Image definition goes slightly downhill from there; but if you compare side-by-side with NVIDIA's own Ampere shots, you'll certainly be able to plot the different transistor arrangements that form the Tensor, RT, and CUDA cores.

New HP OEM Driver References RTX 3080 Ti, RTX 3070 Ti, and RTX 3080 Variants

A new HP OEM GeForce driver points to the two distinct approaches NVIDIA is possibly taking to develop its new high-end GeForce RTX 30-series SKU positioned between the $699 RTX 3080 and the $1,499 RTX 3090; particularly in the wake of the $999 AMD Radeon RX 6900 XT launch. The OEM driver's GPU support list references a number of unreleased graphics cards based on the "GA102" silicon, including engineering samples of 11 GB and 12 GB variants of the RTX 3080; and an RTX 3080 Ti.

The 11 GB and 12 GB variants of the RTX 3080 (which are unreleased engineering samples at this point) could possibly be SKUs carved out with the same core-configuration as the RTX 3080, but with slightly wider memory interfaces, with the 11 GB variant using a 352-bit interface, and the 12 GB variant maxing out the full 384-bit interface of the "GA104," albeit with 8 Gbit memory chips, unlike the RTX 3090, which uses twenty four 8 Gbit chips (2 per 32-bit path), to achieve 24 GB. The RTX 3080 Ti appears to be a whole different beast. Although the HP document doesn't mention its core-configuration or memory size, older reports have pointed at the possibility of this SKU featuring 9,984 CUDA cores, and the full 384-bit wide memory bus (possibly with 12 GB of memory). Even older reports point to the likelihood of the RTX 3080 Ti retaining the 320-bit memory bus of the RTX 3080, but doubling the memory amount to 20 GB.

GALAX Teases New Graphics Card Series: Work The Frames

GALAX via its Virtual Online Expo has revealed a new graphics card series coming to customers. The new Work The Frames (WTF) cards will no longer bring the HOF series' signature white looks, but instead seems to be more about that RGB bling. The tease is just a render at this point, though - your guess is as good as ours as to when these will be available in the market - and even then, and at what supply levels. The renders currently lack any power connectors - so we shouldn't look too hard into the absence of any NVLink fingers either. Following GALAX's usual business decisions, it seems likely that the Work the Frames family will be available in GA102 flavors.

The cooler shroud seems to be all black with RGB accents throughout the heatsink, backplate and even at the card's rear. It's a triple-fan solution seemingly coming in at around 2.2-slot thickness. Check the graphics card for yourself by following the source link and entering the appropriate GALAX Expo room.

NVIDIA GeForce RTX 3080 Ti Landing in January at $999

According to the unknown manufacturer (AIB) based in Taiwan, NVIDIA is preparing to launch the new GeForce RTX 3000 series "Ampere" graphics card. As reported by the HKEPC website, the Santa Clara-based company is preparing to fill the gap between its top-end GeForce RTX 3090 and a bit slower RTX 3080 graphics card. The new product will be called GeForce RTX 3080 Ti. If you are wondering what the specification of the new graphics card will look like, you are in luck because the source has a few pieces of information. The new product will be based on GA102-250-KD-A1 GPU core, with a PG133-SKU15 PCB design scheme. The GPU will contain the same 10496 CUDA core configuration as the RTX 3090.

The only difference to the RTX 3090 will be a reduced GDDR6X amount of 20 GB. Along with the 20 GB of GDDR6X memory, the RTX 3080 Ti graphics cards will feature a 320-bit bus. The TGP of the card is limited to 320 Watts. The sources are reporting that the card will be launched sometime in January of 2021, and it will come at $999. This puts the price category of the RTX 3080 Ti in the same range as AMD's recently launched Radeon RX 6900 XT graphics card, so it will be interesting to see how these two products are competing.

NVIDIA Reportedly Working on GeForce RTX 3080 Ti Graphics Card with 20 GB GDDR6X VRAM

A leak from renowned (and usually on-point) leaker Kopite7kimi claims that NVIDIA has finally settled on new graphics cards to combat AMD's RX 6800 threat after all. After the company has been reported (and never confirmed) to be working on double-memory configurations for their RTX 3070 and RTX 3080 graphics cards (with 16 GB GDDR6 and 20 GB GDDR6X, respectively), the company is now reported to have settled for a 20 GB RTX 3080 Ti to face a (apparently; pending independent reviews) resurgent AMD.

The RTX 3080 Ti specs paint a card with the same CUDA core count as the RTX 3090, with 10496 FP32 cores over the same 320-bit memory bus as the RTX 3080. Kopite includes board and SKU numbers (PG133 SKU 15) along a new GPU codename: GA102-250. The performance differentiator against the RTX 3090 stands to be the memory amount, bus, and eventually core clockspeed; memory speed and board TGP are reported to mirror those of the RTX 3080, so some reduced clocks compared to that graphics card are expected. That amount of CUDA cores means NVIDIA is essentially divvying-up the same GA-102 die between its RTX 3090 (good luck finding one in stock) and the reported RTX 3080 Ti (so good luck finding one of those in stock as well, should the time come). It is unclear how pricing would work out for this SKU, but pricing comparable to that of the RX 6900 XT is the more sensible speculation. Take this report with the usual amount of NaCl.

NVIDIA Allegedly Already Preparing an RTX 3080 Ti Graphics Card

This generation's GPU release is shaping up to be one of the most interesting in late years; for some good reasons, and bad reasons alike. We've heard - keep in mind, not seen - NVIDIA back down from multiple graphics card releases (the double VRAM versions of RTX 3070 and 3080 come to mind); postponing the RTX 3070 until after they have gleaned exactly what AMD will be offering with their RX 6000 series; preparing to launch an RTX 3060 Ti with no announcement whatsoever and before the RTX 3060 is ever launched; and now, apparently, the company is readying a response to AMD's as-of-yet-unannounced RX 6000 series in the form of the RTX 3080 Ti.

Recent performance leaks have placed an unclear AMD GPU (and apparently, not even the fastest Big Navi chip at that) at the same performance level as NVIDIA's RTX 3080, which is a tremendous increase in performance for the red team, coming from years of only being able to effectively compete in the midrange offerings. Now, Kopite7kimi, a known leaker with a proven track record, has claimed that NVIDIA is already prepping a new GA102-based graphics card, sitting in performance between the RTX 3080 and RTX 3090. He lists identifies the GPU as GA102-250-A1, 9984FP32, 384bits GD6X - let's call it the RTX 3080 Ti. If those details are correct, this is yet another product demanding the same 628 mm² GA102 GPU be available for it - in a scenario with inadequate availability of the RTX 3080, 3090, and likely 3070 Ti graphics cards already, should that later one actually materialize. The memory bus on this prospective RTX 3080 Ti is apparently inheriting the same design as the RTX 3090, with a 384-bit solution (compared to the RTX 3080's 320 bit), and likely 12 GB of GDDR6X memory.

NVIDIA Reportedly Cancels Launch of RTX 3080 20 GB, RTX 3070 16 GB

Fresh reports floating in the rumor mill's circulatory system claim that NVIDIA backtracked on its plans to launch higher VRAM capacity versions of their RTX 3080 and the (in the meantime, delayed) RTX 3070. These cards launched with 10 GB VRAM for the RTX 3080 and 8 GB VRAM for the RTX 3070, with reports circulating as early as their announcement that there would be double-capacity versions hitting the market just a few months later - specifically, in December of this year. Videocardz, however, claims that these long-rumored 20 GB and 16 GB SKUs have now been canceled by NVIDIA, who sent this news to its AIB partners - and the usage of canceled, not postponed, is perfunctory.

For cards theoretically shipping come December, this is indeed a small advance notice, but it might be enough for AIB partners to feed all their GA102-200 (RTX 3080) and GA104-400 (RTX 3070) silicon towards the already - if not readily - available models. This report, Videocardz claims, has been confirmed by two of their sources, and comes at the exact same day specifications for AMD's RX 6000 series leaked. It's likely NVIDIA already had knowledge of its competition's designs and performance targets, however, so this could be seen as nothing more than a coincidence. One of the publications' sources claims GDDR6X yields might be the cause for the cancellation, but this doesn't help explain why the alleged RTX 3070 16 GB card (with its GDDR6 chips) was also canceled. Remember: these are rumors on cards that were never announced by NVIDIA themselves, so take these with the appropriate salt-mine level of skepticism.

NVIDIA Unveils RTX A6000 "Ampere" Professional Graphics Card and A40 vGPU

NVIDIA today unveiled its RTX A6000 professional graphics card, the first professional visualization-segment product based on its "Ampere" graphics architecture. With this, the company appears to be deviating from the Quadro brand for the graphics card, while several software-side features retain the brand. The card is based on the same 8 nm "GA102" silicon as the GeForce RTX 3080, but configured differently. For starters, it gets a mammoth 48 GB of GDDR6 memory across the chip's 384-bit wide memory interface, along with ECC support.

The company did not reveal the GPU's CUDA core count, but mentioned that the card's typical board power is 300 W. The card also gets NVLink support, letting you pair up to two A6000 cards for explicit multi-GPU. It also supports GPU virtualization, including NVIDIA GRID, NVIDIA Quadro Virtual Data Center Workstation, and NVIDIA Virtual Compute Server. The card features a conventional lateral blower-type cooling solution, and its most fascinating aspect is its power input configuration, with just the one 8-pin EPS power input. We will update this story with more information as it trickles out.
Update 13:37 UTC: The company also unveiled the A40, a headless professional-visualization graphics card dedicated for virtual-GPU/cloud-GPU applications (deployments at scale in data-centers). The card has similar specs to the RTX A6000.

Update 13:42 UTC: NVIDIA website says that both the A40 and RTX A6000 a 4+4 pin EPS connector (and not 8-pin PCIe) for power input. An 8-pin EPS connector is capable of delivering up to 336 W (4x 7 A @ 12 V).

NVIDIA's Ampere-based Quadro RTX Graphics Card Pictured

Here is the first picture of an alleged next-generation Quadro RTX graphics card based on the "Ampere" architecture, courtesy YouTube channel "Moore's Law is Dead." The new Quadro RTX 6000-series shares many of its underpinnings with the recently launched GeForce RTX 3080 and RTX 3090, in being based on the 8 nm "GA102" silicon. The reference board design retains a lateral blower-type cooling solution, with the blower drawing in air from both sides of the card, through holes punched in the PCB, "Fermi" style. The card features the latest NVLink bridge connector, and unless we're mistaken, it features a single power input near its tail end, which is very likely a 12-pin Molex MicroFit 3.0 input.

As for specifications, "Moore's Law is Dead," shared a handful of alleged specifications that include maxing out of the "GA102" silicon, with all its 42 TPCs (84 SMs) enabled, working out to 10,752 CUDA cores. As detailed in an older story about the next-gen Quadro, NVIDIA is prioritizing memory size over bandwidth, which means this card will receive 48 GB of conventional 16 Gbps GDDR6 memory across the GPU's 384-bit wide memory interface. The 48 GB is achieved using twenty four 16 Gbit GDDR6 memory chips (two chips per 32-bit wide data-path). This configuration provides 768 GB/s of memory bandwidth, which is only 8 GB/s higher than that of the GeForce RTX 3080. The release date of the next-gen Quadro RTX will depend largely on the supply of 16 Gbit GDDR6 memory chips, with leading memory manufacturers expecting 2021 shipping, unless NVIDIA has secured an early production batch.

NVIDIA GeForce RTX 3090 "CEO Edition" Rears its Head, Most Likely a Fake

Social media is abuzz with a screengrab of a regional webpage of the NVIDIA website purporting a "GeForce RTX 3090 CEO Edition" graphics card. Positioned a notch above the "Founders Edition," this $3,499 card, if even remotely real, could be a limited edition product. The screengrab references "48 GB of G6X" memory. We're not sure how this is even possible. The RTX 3090 already uses 8 Gbit GDDR6X chips, piggybacking two chips per 32-bit memory channel, unless Micron has done the unthinkable by launching a 16 Gbit G6X chip within 2020. Frankly, we're very interested to see how the next-gen Quadro RTX even achieves its alleged 48 GB of GDDR6.

That aside, the alloy frame now comes with a gold finish. We wonder if memory and a fancy trim is all that NVIDIA is asking the extra 2 Grand for, or if it even maxed out the "GA102" ASIC (there are two more TPCs left to unlock). As for the name "CEO Edition," there have been instances of tech CEOs flexing their vanity on limited edition products. Limited edition Ryzen and Radeon products, for example, bear the AMD CEO's signature. So the name "CEO Edition" by itself isn't implausible. Just not with these specs, and not this price.

NVIDIA's Top "Ampere" Based Quadro RTX Features 10,752 CUDA Cores, 48GB Memory

Possible specifications of NVIDIA's next-generation flagship Quadro RTX professional graphics card leaked to the web. The SKU is possibly based on the same 8 nm "GA102" silicon as the GeForce RTX 3090, but features more of the silicon unlocked. It apparently features 10,752 CUDA cores, or exactly one TPC (two SMs) more than the RTX 3090. With 84 SM (42 TPC), the unnamed Quadro RTX should feature 84 RT cores, 336 Tensor cores, and 336 TMUs.

NVIDIA's choice for memory for the upcoming Quadro RTX flagship is interesting, as it's prioritizing memory size over bandwidth (which is more relevant in the professional visualization use-case dealing with large data sets). The card features 48 GB of conventional GDDR6 memory clocked at 16 Gbps over the chip's 384-bit wide memory interface, which should work out to 768 GB/s of memory bandwidth. The max GPU Boost frequency is pegged at 1860 MHz. There's no word on availability. Pictured below is the previous-gen Quadro RTX 5000.

NVIDIA Announces GeForce Ampere RTX 3000 Series Graphics Cards: Over 10000 CUDA Cores

NVIDIA just announced its new generation GeForce "Ampere" graphics card series. The company is taking a top-to-down approach with this generation, much like "Turing," by launching its two top-end products, the GeForce RTX 3090 24 GB, and the GeForce RTX 3080 10 GB graphics cards. Both cards are based on the 8 nm "GA102" silicon. Join us as we live blog the pre-recorded stream by NVIDIA, hosted by CEO Jen-Hsun Huang.

EVGA Teases the GeForce RTX 3090 KINGPIN: 360mm AIO Hybrid Cooler, RGB OLED Screen

NVIDIA's announcement of the new RTX 3000-series cards has led to simultaneous announcement of add-in card partners with their solutions and custom takes on the same. These typically come in the form of a press release, with common specifications shared and emphasis on the cooling systems as well as other brand-specific features. EVGA was no different, and the PR shows off their new iCX3 cooling technology in it. More interesting to the enthusiasts and overclockers among us is newer information on their halo card- the RTX 3090 K|NGP|N (Kingpin)- with product management director Jacob revealing it will continue to use a hybrid AIO cooler as with the RTX 2080 Ti variant, but with a massive 360 mm radiator and three 120 mm fans for the behemoth GA102 die and accompanying power delivery solution. No word on pricing or retail availability yet, but look forward to your friendly neighborhood TPU reviews in due course of time to help make a purchase decision on your next GPU.

ASUS ROG Strix GeForce RTX 3090 Ampere Pictured

Here's a much clearer picture of the ROG Strix GeForce RTX 3090 "Ampere" graphics card by ASUS. A render of this card was revealed back in July, when the RTX 3090 was still speculated to be called the "RTX 3080 Ti." The card features the latest iteration of the ROG Strix design scheme by ASUS, with more shiny metal bits that conceal lighting elements. The cooler features a chunky aluminium fin-stack heatsink that's ventilated by three Axial-Tech fans. The card is 3 slots thick, and about an inch taller than what would constitute "full height." VideoCardz, which is also the source of the image, predicts that ASUS could use same board design for both the RTX 3090 and the RTX 3080. We've seen custom-design RTX 3090 cards featuring as many as three 8-pin PCIe power inputs, while those based on the RTX 3080 have been shown with two. With the RTX 3090 and RTX 3080 being based on the common silicon, the GA102, differing in memory and core configurations, the R&D costs for custom-design board partners is greatly reduced.

NVIDIA GeForce RTX 3090 and 3080 Specifications Leaked

Just ahead of the September launch, specifications of NVIDIA's upcoming RTX Ampere lineup have been leaked by industry sources over at VideoCardz. According to the website, three alleged GeForce SKUs are being launched in September - RTX 3090, RTX 3080, and RTX 3070. The new lineup features major improvements: 2nd generation ray-tracing cores and 3rd generation tensor cores made for AI and ML. When it comes to connectivity and I/O, the new cards use the PCIe 4.0 interface and have support for the latest display outputs like HDMI 2.1 and DisplayPort 1.4a.

The GeForce RTX 3090 comes with 24 GB of GDDR6X memory running on a 384-bit bus at 19.5 Gbps. This gives a memory bandwidth capacity of 936 GB/s. The card features the GA102-300 GPU with 5,248 CUDA cores running at 1695 MHz, and is rated for 350 W TGP (board power). While the Founders Edition cards will use NVIDIA's new 12-pin power connector, non-Founders Edition cards, from board partners like ASUS, MSI and Gigabyte, will be powered by two 8-pin connectors. Next up is specs for the GeForce RTX 3080, a GA102-200 based card that has 4,352 CUDA cores running at 1710 MHz, paired with 10 GB of GDDR6X memory running at 19 Gbps. The memory is connected with a 320-bit bus that achieves 760 GB/s bandwidth. The board is rated at 320 W and the card is designed to be powered by dual 8-pin connectors. And finally, there is the GeForce RTX 3070, which is built around the GA104-300 GPU with a yet unknown number of CUDA cores. We only know that it has the older non-X GDDR6 memory that runs at 16 Gbps speed on a 256-bit bus. The GPUs are supposedly manufactured on TSMC's 7 nm process, possibly the EUV variant.

NVIDIA Ampere GA102-300-A1 GPU Die Pictured

Here's the first picture of an NVIDIA "Ampere" GA102 GPU die. This is the largest client-segment implementation of the "Ampere" architecture by NVIDIA, targeting the gaming (GeForce) and professional-visualization (Quadro) market segments. The "Ampere" architecture itself debuted earlier this year with the A100 Tensor Core scalar processor that's winning hearts and minds in the HPC community faster than ice cream on a dog day afternoon. There's no indication of die-size, but considering how tiny the 10.3 billion-transistor AMD "Navi 10" die is, the GA102 could come with a massive transistor count if its die is as big as that of the TU102. The GPU in the picture is also a qualification sample, and was probably pictured off a prototype graphics card. Powering the GeForce RTX 3090, the GA102-300 is expected to feature a CUDA core count of 5,248. According to VideoCardz, there's a higher trim of this silicon, the GA102-400, which could make it to NVIDIA's next halo product under the TITAN brand.

Video Memory Sizes Set to Swell as NVIDIA Readies 20GB and 24GB GeForce Amperes

NVIDIA's GeForce RTX 20-series "Turing" graphics card series did not increase video memory sizes in comparison to GeForce GTX 10-series "Pascal," although the memory itself is faster on account of GDDR6. This could change with the GeForce RTX 30-series "Ampere," as the company looks to increase memory sizes across the board in a bid to shore up ray-tracing performance. WCCFTech has learned that in addition to a variety of strange new memory bus widths, such as 320-bit, NVIDIA could introduce certain higher variants of its RTX 30-series cards with video memory sizes as high as 20 GB and 24 GB.

Memory sizes of 20 GB or 24 GB aren't new for NVIDIA's professional-segment Quadro products, but it's certainly new for GeForce, with only the company's TITAN-series products breaking the 20 GB-mark at prices due north of $2,000. Much of NVIDIA's high-end appears to be resting on segmentation of the PG132 common board design, coupled with the GA102 silicon, from which the company could carve out several SKUs spaced far apart in the company's product stack. NVIDIA's next-generation GeForce "Ampere" family is expected to debut in September 2020, with product launches in the higher-end running through late-Q3 and Q4 of 2020.

NVIDIA GeForce "Ampere" GPUs Built on Samsung 8nm Instead of TSMC 7nm?

NVIDIA's upcoming GeForce "Ampere" family of GPUs will be built almost entirely on Samsung's 8 nanometer silicon fabrication process that's derived from its 10 nm node; rather than TSMC's 7 nm process, according to kopite7kimi, a source with a high hit-rate with NVIDIA rumors in the past. The 8LPP silicon fabrication node by Samsung is an extension of the company's 10LPP (10 nm) node. Both have the same fin pitch, but reductions are made in the areas of gate pitch (down by 6%) resulting in a transistor density of over 61 million/mm². Apparently NVIDIA's entire high-end product stack, including the GA102 silicon that powers at least three high-end consumer SKUs, are expected to be based on Samsung 8LPP.

NVIDIA's Next-Gen Reference Cooler Costs $150 By Itself, to Feature in Three SKUs

Pictures of alleged next-generation GeForce "Ampere" graphics cards emerged over the weekend, which many of our readers found hard to believe. It's features a dual-fan cooling solution, in which one of the two fans is on the reverse side of the card, blowing air outward from the cooling solution, while the PCB extends two-thirds the length of the card. Since then, there have been several fan-made 3D renders of the card. NVIDIA is not happy with the leak, and started an investigation into two of its contractors responsible for manufacturing Founders Edition (reference design) GeForce graphics cards, Foxconn and BYD (Build Your Dreams), according to a report by Igor's Lab.

According to the report, the cooling solution, which looks a lot more overengineered than the company's RTX 20-series Founders Edition cooler, costs a hefty USD $150, or roughly the price of a 280 mm AIO CLC. It wouldn't surprise us if Asetek's RadCard costs less. The cooler consists of several interconnected heatsink elements with the PCB in the middle. Igor's Lab reports that the card is estimated to be 21.9 cm in length. Given its cost, NVIDIA is reserving this cooler for only the top three SKUs in the lineup, the TITAN RTX successor, the RTX 2080 Ti successor, and the RTX 2080/SUPER successor.

NVIDIA RTX 3080 Ti and GA102 "Ampere" Specs, Other Juicy Bits Revealed

PC hardware focused YouTube channel Moore's Law is Dead published a juicy tech-spec reveal of NVIDIA's next-generation "Ampere" based flagship consumer graphics card, the GeForce RTX 3080 Ti, citing correspondence with sources within NVIDIA. The report talks of big changes to NVIDIA's Founders Edition (reference) board design, as well as what's on the silicon. To begin with, the RTX 3080 Ti reference-design card features a triple-fan cooling solution unlike the RTX 20-series. This cooler is reportedly quieter than the RTX 2080 Ti FE cooling solution. The card pulls power from a pair of 8-pin PCIe power connectors. Display outputs include three DP, and one each of HDMI and VirtualLink USB-C. The source confirms that "Ampere" will implement PCI-Express gen 4.0 x16 host interface.

With "Ampere," NVIDIA is developing three tiers of high-end GPUs, with the "GA102" leading the pack and succeeding the "TU102," the "GA104" holding the upper-performance segment and succeeding today's "TU104," but a new silicon between the two, codenamed "GA103," with no predecessor from the current-generation. The "GA102" reportedly features 5,376 "Ampere" CUDA cores (up to 10% higher IPC than "Turing"). The silicon also taps into the rumored 7 nm-class silicon fabrication node to dial up GPU clock speeds well above 2.20 GHz even for the "GA102." Smaller chips in the series can boost beyond 2.50 GHz, according to the report. Even with the "GA102" being slightly cut-down for the RTX 3080 Ti, the silicon could end up with FP32 compute performance in excess of 21 TFLOPs. The card uses faster 18 Gbps GDDR6 memory, ending up with 863 GB/s of memory bandwidth that's 40% higher than that of the RTX 2080 Ti (if the memory bus width ends up 384-bit). Below are screengrabs from the Moore's Law is Dead video presentation, and not NVIDIA slides.
Return to Keyword Browsing