Nvidia 7nm: Nvidia Places 7nm And 5nm Orders For Next-Gen Ampere GPUs

NVIDIA Ampere Uses TSMC’s 7nm, Next-Gen Hopper GPUs Uses Samsung’s 5nm EUV Process

In a new report, ChinaTimes has reported that NVIDIA’s Ampere GPUs are indeed going to utilize TSMC’s 7nm process node technology whereas the next-generation Hopper GPU family would be based on the Samsung 5nm EUV process node.

NVIDIA’s Ampere Fabricated on TSMC’s 7nm Process, Next-Gen Hopper GPUs on Samsung’s 5nm EUV Node, Alleges Rumor

NVIDIA will be publishing a pre-recorded version of its GTC 2020 ‘Get Amped’ keynote featuring Jensen Huang, the company’s CEO, who will be presenting new innovations, announcements, and technologies. The core segment of the online keynote will focus on the Ampere GPU which will be unveiled along with technical information such as specifications and products that it would be featured inside.

According to ChinaTimes, NVIDIA has been confirmed as one of the biggest customers of TSMC’s 7nm process node. This seems to indicate that NVIDIA’s Ampere GPUs, at least for the HPC & the data center segment, will be based on TSMC’s advanced chip manufacturing process. As soon as the news came out, Samsung released a statement in its Q1 earnings report, confirming that it will begin production of its 5nm (5LPE) EUV process in Q2 2020.

Samsung has already reportedly received orders for processors, graphics cards and various other products to utilize its 5nm process node. According to the report, most of these products are going to be utilized in consumer applications. Based on internal roadmaps, AMD is not expected to use Samsung foundries for its next-generation processors or graphics cards but it is very likely that the next-generation Hopper GPUs from NVIDIA will utilize Samsung’s 5nm EUV node which would be replacing Ampere GPUs.

NVIDIA Tesla Graphics Cards Comparison

Tesla Graphics Card Name NVIDIA Tesla M2090 NVIDIA Tesla K40 NVIDIA Telsa K80 NVIDIA Tesla P100 NVIDIA Tesla V100 NVIDIA Tesla Next-Gen #1 NVIDIA Tesla Next-Gen #2 NVIDIA Tesla Next-Gen #3
GPU Architecture Fermi Kepler Maxwell Pascal Volta Ampere? Ampere? Ampere?
GPU Process 40nm 28nm 28nm 16nm 12nm 7nm? 7nm? 7nm?
GPU Name GF110 GK110 GK210 x 2 GP100 GV100 GA100? GA100? GA100?
Die Size 520mm2 561mm2 561mm2 610mm2 815mm2 TBD TBD TBD
Transistor Count 3. 00 Billion 7.08 Billion 7.08 Billion 15 Billion 21.1 Billion TBD TBD TBD
CUDA Cores 512 CCs (16 CUs) 2880 CCs (15 CUs) 2496 CCs (13 CUs) x 2 3840 CCs 5120 CCs 6912 CCs 7552 CCs 7936 CCs
Core Clock Up To 650 MHz Up To 875 MHz Up To 875 MHz Up To 1480 MHz Up To 1455 MHz 1.08 GHz (Preliminary) 1.11 GHz (Preliminary) 1.11 GHz (Preliminary)
FP32 Compute 1.33 TFLOPs 4.29 TFLOPs 8.74 TFLOPs 10.6 TFLOPs 15.0 TFLOPs ~15 TFLOPs (Preliminary) ~17 TFLOPs (Preliminary) ~18 TFLOPs (Preliminary)
FP64 Compute 0.66 TFLOPs 1.43 TFLOPs 2.91 TFLOPs 5.30 TFLOPs 7.50 TFLOPs TBD TBD TBD
VRAM Size 6 GB 12 GB 12 GB x 2 16 GB 16 GB 48 GB 24 GB 32 GB
VRAM Bus 384-bit 384-bit 384-bit x 2 4096-bit 4096-bit 4096-bit? 3072-bit? 4096-bit?
VRAM Speed 3. 7 GHz 6 GHz 5 GHz 737 MHz 878 MHz 1200 MHz 1200 MHz 1200 MHz
Memory Bandwidth 177.6 GB/s 288 GB/s 240 GB/s 720 GB/s 900 GB/s 1.2 TB/s? 1.2 TB/s? 1.2 TB/s?
Maximum TDP 250W 300W 235W 300W 300W TBD TBD TBD

It will be definitely interesting to see an updated roadmap that clears this up during the GTC 2020 keynote. There also exists an opportunity for NVIDIA to go with Samsung’s 5nm for its consumer-based graphics card lineup since mass production is indeed happening this quarter and GeForce cards replacing the Turing lineup could arrive by the end of 2020.

The NVIDIA GeForce RTX lineup was also unveiled in late 2018. We have internally confirmed with several sources that NVIDIA’s AIB partners are aggressively clearing up their GeForce RTX (Turing GPU) inventories ahead of a big launch of a consumer-based GeForce lineup in Q3 2020. Q3 2020 also makes sense because that is where Computex 2020 has been shifted to but NVIDIA might select another venue to host its GPU (GeForce) announcement.

With that said, NVIDIA’s Ampere GPUs are definitely going to shake things up in the HPC market with several variants already leaked and performance being rated at around 30 TFLOPs (FP32). We will keep you updated as more info comes prior to the 14th of May when NVIDIA will be presenting its next-gen GPU lineup.

What process nodes do you think NVIDIA’s Ampere & Hopper GPUs are likely to utilize?

  • Ampere — TSMC 7nm Process Node

  • Ampere — Samsung 5nm Process Node

  • Hopper — Samsung 5nm Process Node

  • Hopper — TSMC 5nm Process Node

Poll Options are limited because JavaScript is disabled in your browser.

Share this story



New Chinese 7nm GPU rivals Nvidia and AMD for performance

Skip to main content

When you purchase through links on our site, we may earn an affiliate commission. Here’s how it works.

(Image credit: Shanghai Tianshu Zhixin Semiconductor)

Audio player loading…

Shanghai Tianshu Zhixin Semiconductor Co’s new 7nm GPU is up and running. The new chip is a beast, packing no fewer than 24 billion transistors. For context, Nvidia’s GA102 GPU, as found in the RTX 3080 and 3090 graphics boards, contains 28 billion transistors.

Your next upgrade

(Image credit: Future)

Best CPU for gaming : the top chips from Intel and AMD
Best graphics card : your perfect pixel-pusher awaits
Best SSD for gaming : get into the game ahead of the rest

Known as Big Island (BI), the new GPU can crank out up to 147 TFLOPS of FP16 performance. As Tom’s Hardware points out, that’s certainly competitive with similar products from Nvidia and AMD. Nvidia’s A100 and the AMD Instinct MI100 are capable of 77.97 TFLOPS and 184.6 TFLOPS of FP16 performance, respectively, though the Nvidia A100 also has Tensor cores with an additional 312 TFLOPS of FP16.

As for production technology, Shanghai Tianshu Zhixin Semiconductor says BI is fabbed on 7nm. The details provided seem to align with TSMC’s 7nm node, but the foundry partner is not identified and it’s unclear if China has domestic chip production technology in that class. Images of the new GPU imply a multi-chip package with the GPU itself housed with onboard memory or cache.

Of course, the BI chip along with the AMD and Nvidia competition mentioned are all GPUs specifically made for enterprise applications, not desktop PCs or gaming. Any GPU for gaming will need a wide array of hardware beyond the computational cores that crank out those TFLOPS, including a full pipeline of texture units, pixel processing and render outputs.  

It’s not clear if the BI GPU contains any of those features and, indeed, there’s little reason for their inclusion given the remit. So, the safe assumption is that Shanghai Tianshu Zhixin Semiconductor’s new GPU poses zero threat to the established competition in gaming graphics. 

That said, the simple fact that a large, ultra-complex GPU with comparable performance in terms of floating point performance to the established players has been produced is significant. It’s a big step from there to a gaming GPU. But it’s certainly a start.

Sign up to get the best content of the week, and great gaming deals, as picked by the editors.

Contact me with news and offers from other Future brandsReceive email from us on behalf of our trusted partners or sponsors

Jeremy has been writing about technology and PCs since the 90nm Netburst era (Google it!) and enjoys nothing more than a serious dissertation on the finer points of monitor input lag and overshoot followed by a forensic examination of advanced lithography. Or maybe he just likes machines that go “ping!” He also has a thing for tennis and cars.

PC Gamer is part of Future plc, an international media group and leading digital publisher. Visit our corporate site .

Future Publishing Limited Quay House, The Ambury,
BA1 1UA. All rights reserved. England and Wales company registration number 2008885.

NVIDIA will have a hard time transitioning Ampere gaming GPUs to the 7nm TSMC process

3DNews Technologies and IT market. News NVIDIA graphics card will be hard to translate gaming GP…

The most interesting in the reviews

10/13/2020 [20:27],

Nikolai Khizhnyak

Last week it was reported that next year NVIDIA may transfer some gaming models of video cards based on the Ampere architecture to TSMC’s 7nm process technology. According to the specialized media, this step could create serious problems for NVIDIA.

Image Source: DigiTimes

The apparently unsuccessful launch of the entire GeForce RTX 30xx series showed that Samsung is not keeping up with the production volumes of quality 8nm GPUs for NVIDIA. According to sources, this factor can serve as one of the reasons for turning to TSMC. In addition, for NVIDIA, the transition from 8nm to 7nm manufacturing process will not be very expensive. The two companies are already collaborating on the use of the 7nm process technology. A Taiwanese semiconductor manufacturer is building GA100 graphics chips for NVIDIA’s non-consumer graphics solutions. It is very likely that both parties will be able to agree on a price for the production of gaming chips.

Image source: NVIDIA

Meanwhile, TechPowerUp resource believes that using two manufacturing processes from two different vendors within the same product line can create many problems for both NVIDIA itself and its customers. Firstly, this will require the company to make changes to the power schemes of video cards, as well as conduct additional tests for the stability of the new chips, including by its partner companies. Secondly, if NVIDIA makes changes to its video cards without any notification to customers, then for it it may turn into litigation from the first buyers of video cards with 8-nm Samsung chips.

According to TechPowerUp , many potential buyers can simply refuse to buy models based on the older process technology and start looking for solutions based on the fresh 7nm process technology. Ultimately, this will exacerbate the problem with the supply of GeForce RTX 30-series graphics cards even more.

Initially, it was not indicated which models of video cards NVIDIA could transfer to the 7nm process technology, however, TechPowerUp suggests that we are talking about older GeForce RTX 3090, RTX 3080 and RTX 3070. Perhaps these will be “Super versions” of the cards, in which not only the manufacturing process itself will be changed, but also the Ampere architecture will be optimized. At the same time, the production of chips for low-end GeForce RTX 30-series NVIDIA cards may leave Samsung. Choosing this approach will save the GPU designer from potential legal issues.


If you notice an error, select it with the mouse and press CTRL+ENTER.

Related materials

Permanent URL:
News Hardware, video cards,

tsmc, samsung, 7nm, 8nm, manufacturing process, GPU, nvidia, ampere

← В
To the future →

NVIDIA GeForce Ampere: 7nm made by TSMC or Samsung?

More and more technical details of the new GeForce Ampere video cards appear. We already know the memory capacity, its connectivity, the number of stream processors, and more. Information about clock frequencies has also been leaked, although in this case we are talking about partner video cards without overclocking.

But questions remain. Boost frequencies of Founders Editions cards are unknown. In addition, the structure of special Ampere GPU function blocks is not clear. For the GA100 GPU, NVIDIA has reduced the number of more efficient Tensor cores from four to two per SM (Streaming Multiprocessor). It is not known in what form they will be implemented on the upcoming GeForce video cards.

The number of RT cores that provide hardware-accelerated ray tracing is also unknown. The Turing architecture uses one RT core per SM. The Ampere architecture will install second generation RT cores. If NVIDIA plans to increase ray tracing performance, then more RT cores per SM will be required.

But until the official presentation of the new GeForce graphics cards, which will take place tomorrow evening, we are unlikely to find out the details of the Ampere architecture.

7 nm from TSMC or Samsung?

In recent years, NVIDIA has used only TSMC’s facilities to manufacture GPUs. Which is true for current Turing GPUs as well. The arrival of the GA100 GPU attracted attention with NVIDIA talking about a dedicated production TSMC 7N (not N7). This fact could be attributed to a typo if the 7N process technology appeared once, but NVIDIA made this «mistake» again during the presentation of the PCI Express variant.

At the plenary talk on the accelerator A100 NVIDIA CEO Jensen Huang spoke about «TSMC 7 nm process that has been optimized for NVIDIA» . It is quite clear that the GA100 GPU is produced at the facilities of TSMC, but there has been no doubt about this for a long time. And the name N7 or N7P in this case is not important.

But more interesting is the production of other GA10x GPUs by TSMC or Samsung. Rumors have been circulating for several months that NVIDIA may well return to the power of Samsung in order to reduce dependence on TSMC. Of course, each manufacturer is always looking for optimal production in terms of quality, technology and costs. In the case of the GA100 GPU, there are no alternatives to TSMC for NVIDIA, CoWoS packaging and HBM2 memory play their role here.

But does Samsung have the capacity to produce Ampere GPUs? For a long time, Samsung only offered 8LPP (low power performance), but now the Korean manufacturer has 7LPP. This process is currently only used for the production of Exynos 9825, but Samsung advertises 7LPP for larger chips as well: «HVM (High Volume Manufacturing) goal of 250W has been demonstrated.» . So from a technical point of view, production at Samsung factories is quite possible.

Samsung has been offering the 7LPP process since mid-2019, so it’s fully developed.

It doesn’t matter in the end

It doesn’t matter in the end whether the 5,248 GeForce RTX 3090 stream processors are made using TSMC’s or Samsung’s 7nm process. NVIDIA severely limits the design specifications within which the GPU frequencies will be. The share of the yield of suitable crystals is also taken into account.

Manufacturing in Samsung factories is quite a popular topic, it pops up from time to time, but with the 8nm process technology, the experience is not very successful.