Your questions answered Vol 2 - IM Private Clients
Discussion
As a general question, my income has tripped the High Income Child Benefit Charge and I had to fill in a self assessment tax return which resulted in me having to pay the benefit back again for this year.
There were several other options on the self assessment that I ignored as I was 'invited' to fill one in due to the HICBC so focused on that. Is this area in general something that IM can help with as can losses on investments, etc also be put on the self assessment to reduce the tax due?
There were several other options on the self assessment that I ignored as I was 'invited' to fill one in due to the HICBC so focused on that. Is this area in general something that IM can help with as can losses on investments, etc also be put on the self assessment to reduce the tax due?
Edited by KTF on Monday 20th November 14:34
KTF said:
As a general question, my income has tripped the High Income Child Benefit Charge and I had to fill in a self assessment tax return which resulted in me having to pay the benefit back again for this year.
There were several other options on the self assessment that I ignored as I was 'invited' to fill one in due to the HICBC so focused on that. Is this area in general something that IM can help with as can losses on investments, etc also be put on the self assessment to reduce the tax due?
Hi Peter,There were several other options on the self assessment that I ignored as I was 'invited' to fill one in due to the HICBC so focused on that. Is this area in general something that IM can help with as can losses on investments, etc also be put on the self assessment to reduce the tax due?
Edited by KTF on Monday 20th November 14:34
Are you referring to the tax year ending 5 April 23? It's not my area of expertise but I think the following is correct. Capital losses would be able to offset against capital gains, not income unless it's an unlisted security.
If you are looking at the tax year end 5 April 24 you can still make pension contributions (assuming you have unused allowance) to reduce your income further and potentially get under the start of the taper.
Regards
Adam
AdamIM said:
Hi Peter,
Are you referring to the tax year ending 5 April 23? It's not my area of expertise but I think the following is correct. Capital losses would be able to offset against capital gains, not income unless it's an unlisted security.
If you are looking at the tax year end 5 April 24 you can still make pension contributions (assuming you have unused allowance) to reduce your income further and potentially get under the start of the taper.
Regards
Adam
Hi Adam,Are you referring to the tax year ending 5 April 23? It's not my area of expertise but I think the following is correct. Capital losses would be able to offset against capital gains, not income unless it's an unlisted security.
If you are looking at the tax year end 5 April 24 you can still make pension contributions (assuming you have unused allowance) to reduce your income further and potentially get under the start of the taper.
Regards
Adam
Ok, if its not classed as income then I guess thats a separate tax issue? This is a new area for me so not sure what sort of things I should (or should not) be putting on the self assessment form.
My question related to the tax year 22/23. I went over the threshold in that tax year and HMRC sent me a nice letter 'inviting' me to register for self assessment in order to pay back the child benefit
For this tax year I am looking at increasing my pension contributions via my employers scheme as this is then taken off at source which will get me back under the threshold. Just need to create/look for a spreadsheet to work out what I need to increase the contributions by bearing in mind that I am already part way through the tax year already...
KTF said:
AdamIM said:
Hi Peter,
Are you referring to the tax year ending 5 April 23? It's not my area of expertise but I think the following is correct. Capital losses would be able to offset against capital gains, not income unless it's an unlisted security.
If you are looking at the tax year end 5 April 24 you can still make pension contributions (assuming you have unused allowance) to reduce your income further and potentially get under the start of the taper.
Regards
Adam
Hi Adam,Are you referring to the tax year ending 5 April 23? It's not my area of expertise but I think the following is correct. Capital losses would be able to offset against capital gains, not income unless it's an unlisted security.
If you are looking at the tax year end 5 April 24 you can still make pension contributions (assuming you have unused allowance) to reduce your income further and potentially get under the start of the taper.
Regards
Adam
Ok, if its not classed as income then I guess thats a separate tax issue? This is a new area for me so not sure what sort of things I should (or should not) be putting on the self assessment form.
My question related to the tax year 22/23. I went over the threshold in that tax year and HMRC sent me a nice letter 'inviting' me to register for self assessment in order to pay back the child benefit
For this tax year I am looking at increasing my pension contributions via my employers scheme as this is then taken off at source which will get me back under the threshold. Just need to create/look for a spreadsheet to work out what I need to increase the contributions by bearing in mind that I am already part way through the tax year already...
If the other tax matter is capital gains and losses. These will be realised GIA gains and losses. I believe you can carry losses forward until they are utilised. Nik knows more on this and if I am mistaken he will chime in. I hope this helps
Adam
This is only for the child benefit charge rather than the investment aspect. Will need to crunch some numbers about how much I need to increase the contributions by and then submit another tax return next year for the part payment anyway...
I just wondered about what other things could offset tax as this is all new to me.
I just wondered about what other things could offset tax as this is all new to me.
KTF said:
This is only for the child benefit charge rather than the investment aspect. Will need to crunch some numbers about how much I need to increase the contributions by and then submit another tax return next year for the part payment anyway...
I just wondered about what other things could offset tax as this is all new to me.
The main items are:I just wondered about what other things could offset tax as this is all new to me.
Personal allowance
Pension contributions
charitable donations
There are a few smaller items like travel, work from home, travel...potentially. That's it
Mr Hunt might throw you some tax relief later this week
2Btoo said:
NVIDIA has an earnings call tomorrow. Any predictions of what it might bring? 
Hi O, Expectations are 16.11B and 3.39 eps (non gaap). No one knows what the reaction will be. They will beat by a lot, how much, who knows. I’ll be happy if they report 19B and 3.90. I’d be thrilled if they manage 20B but it’s a big ask. Guidance is key and we know(from cfo) every visible quarter will be higher than the previous, so reported Nr +4B would be within scope. Gross margins of 72% and net profit margins of 50% at least, maybe 52%.
There will be plenty of opinions after the fact. In our opinion it’s not slowing down (earnings growth). Wallstreet has them delivering 70B in calendar 24 give or take. We think it will be well over 100B, possibly 120.
What the stock does tomorrow after hours is a coin flip but I’ll go out on a limb and say I’d be surprised if it fell. It won’t change our opinions if the report card is as expected.
I expect the ceo to literally say, growth has found another gear, a second wave. It’s what I see in the field. NVidia will transition to a software business over time and expand their margins.
Regardless, it’s very early in their journey, we will take in the conference call and move from there.
AdamIM said:
2Btoo said:
NVIDIA has an earnings call tomorrow. Any predictions of what it might bring?
Hi O, Expectations are 16.11B and 3.39 eps (non gaap). No one knows what the reaction will be. They will beat by a lot, how much, who knows. I’ll be happy if they report 19B and 3.90. I’d be thrilled if they manage 20B but it’s a big ask. Guidance is key and we know(from cfo) every visible quarter will be higher than the previous, so reported Nr +4B would be within scope. Gross margins of 72% and net profit margins of 50% at least, maybe 52%.
There will be plenty of opinions after the fact. In our opinion it’s not slowing down (earnings growth). Wallstreet has them delivering 70B in calendar 24 give or take. We think it will be well over 100B, possibly 120.
What the stock does tomorrow after hours is a coin flip but I’ll go out on a limb and say I’d be surprised if it fell. It won’t change our opinions if the report card is as expected.
I expect the ceo to literally say, growth has found another gear, a second wave. It’s what I see in the field. NVidia will transition to a software business over time and expand their margins.
Regardless, it’s very early in their journey, we will take in the conference call and move from there.
On a Y/Y basis Revenue is up +206% and EPS +593%
The stock was all over the place last night from -$30 to +$10 and currently sits at -$8 which isn't really surprising as the market digests the information. The take away is calendar 2024 revenue will be $100B min and probably more. We are modelling a mid point of $110B and a net income figure of at least $60B which equates to a PE of 20. When you look at growth rates it looks low. For comparison MSFT/Apple are in the 28 range, ADBE closer to 40 and TSLA > 70.
More to come.
I’m sure Adam is typing up a NVIDIA report but reading on quote on line bewildered me,
“This caused some wild swings in after-hours trading, its stock sinking under $475.00 at one stage - quite a move given that each dollar is worth $2.5 billion in market cap for the $1.2 trillion company”
Can that be true? (The $1 = $2.5 billion bit)
“This caused some wild swings in after-hours trading, its stock sinking under $475.00 at one stage - quite a move given that each dollar is worth $2.5 billion in market cap for the $1.2 trillion company”
Can that be true? (The $1 = $2.5 billion bit)
tight fart said:
I’m sure Adam is typing up a NVIDIA report but reading on quote on line bewildered me,
“This caused some wild swings in after-hours trading, its stock sinking under $475.00 at one stage - quite a move given that each dollar is worth $2.5 billion in market cap for the $1.2 trillion company”
Can that be true? (The $1 = $2.5 billion bit)
Yes $1 is $2.5B in value for nvidia. $1 move for Apple is $16.5B and $7.5B for MSFT“This caused some wild swings in after-hours trading, its stock sinking under $475.00 at one stage - quite a move given that each dollar is worth $2.5 billion in market cap for the $1.2 trillion company”
Can that be true? (The $1 = $2.5 billion bit)
After hours is not a liquid market. It's a platform for speculators, mainly. It's only an indication of what direction a stock will take when traded on the exchange. The big players do not participate in AH trading. The wild swings as noted just show you it's often a casino table. I mean who in their right mind sold for $475 literally 5 sec after seeing the news wire. On an absolute basis NVDA earnings make it 30% cheaper than MSFT and Apple and more on a relative basis because NVDA growth is much much higher.
Results and notes taken from the conference call. For those interested it is worth reading how the entire ecosystem is developing and you will start to appreciate this is just the beginning. The biggest mistake naysayers make is that they think all these GPU orders are front loaded and it will die out. They are wrong because they don't understand the demand drivers. LLMs complexity is growing so fast that demand for compute power is growing 100% every 6 months. It may take years for supply to catch up to demand. And before it does, the company's investments in many other verticals will throw of huge cashflows, things like its big bet on drug discovery and computer vision to hunt cancers, or Omniverse which is helping train surgeons on intricate brain surgery. The ability to virtually perform a surgery before you actually do it-100% perfect simulation. The company isn't just selling hardware to big pharma, they are partnering and developing the IP, so they get paid in the future when it becomes ubiquitous.
7 out of 10 vehicle OEMs use Nvidia for driver assist systems and regardless, their tech is as good as Tesla. So if rob taxis become a thing, Nvidia is in a good position to generate significant revenues. In the interim they have agreements to supply $15B in assist systems between now and 2028
The world is changing and every industry is moving quickly to accelerated computing. Nvidia is by far the Nr 1 platform provider. Trading now at a Fwd PE of around 20, we consider the stock to be reasonably priced in absolute terms and under valued when looking at their growth.
The best is yet to come.
Q3 was another record quarter. Revenue of $18.1 billion was up 34% sequentially and up more than 200% year-on-year and well above our outlook for $16 billion.
The continued ramp of the NVIDIA HGX platform based on our Hopper Tensor Core GPU architecture, along with InfiniBand end-to-end networking, drove record revenue of $14.5 billion, up 41% sequentially and up 279% year-on-year.
NVIDIA HGX with InfiniBand together are essentially the reference architecture for AI supercomputers and data center infrastructures. Some of the most exciting generative AI applications are built and run on NVIDIA, including Adobe Firefly, ChatGPT, Microsoft 365 Copilot, CoAssist, now assist with ServiceNow and Zoom AI Companion. Our Data Center compute revenue quadrupled from last year and networking revenue nearly tripled.
Investments in infrastructure for training and inferencing large language models, deep learning, recommender systems and generative AI applications is fueling strong broad-based demand for NVIDIA accelerated computing. Inferencing is now a major workload for NVIDIA AI computing. Consumer Internet companies and enterprises drove exceptional sequential growth in Q3, comprising approximately half of our Data Center revenue and outpacing total growth.
Companies like Meta are in full production with deep learning, recommender systems and also investing in generative AI to help advertisers optimise images and text. Most major consumer Internet companies are racing to ramp up generative AI deployment. The enterprise wave of AI adoption is now beginning. Enterprise software companies such as Adobe, Databricks, Snowflake and ServiceNow are adding AI copilots and the systems to their platforms. And broader enterprises are developing custom AI for vertical industry applications such as Tesla in autonomous driving.
Cloud service providers drove roughly half of Data Centre revenue in the quarter. Demand was strong from all hyperscale CSPs(cloud service providers), as well as from a broadening set of GPU-specialised CSPs globally that are rapidly growing to address the new market opportunities in AI. NVIDIA H100 Tensor Core GPU instances are now generally available in virtually every cloud with instances in high demand.
We have significantly increased supply every quarter this year to meet strong demand and expect to continue to do so next year! This is where every quarter will be higher than the preceding. The company is still significantly constrained. Current orders are booking 2025 deliver.
We will also have a broader and faster product launch cadence to meet the growing and diverse set of AI opportunities. This is next Gen chips, Blackwell and then 'X'.Towards the end of the quarter, the U.S. government announced a new set of export control regulations for China and other markets, including Vietnam and certain countries in the Middle East. These regulations require licenses for the export of a number of our products, including our Hopper and Ampere 100 and 800 series and several others.
Our sales to China and other affected destinations derived from products that are now subject to licensing requirements have consistently contributed approximately 20% to 25% of Data Centre revenue over the past few quarters. We expect that our sales to these destinations will decline significantly in the fourth quarter. So we believe will be more than offset by strong growth in other regions. The U.S. government designed the regulation to allow the U.S. industry to provide data centre compute products to markets worldwide, including China. Takeaway, they still guided 20B, think what they could have delivered and the chip issues re china will be resolved with accepted redesigns. Nvidia is working on it now with Govt assistance.
Continuing to compete worldwide as the regulations encourage, promotes U.S. technology leadership, spurs economic growth and supports U.S. jobs. For the highest performance levels, the government requires licenses. For lower performance levels, the government requires a streamlined prior notification process. And for products even lower performance levels, the government does not require any notice at all.
Following the government's clear guidelines, we are working to expand our Data Centre product portfolio to offer compliance solutions for each regulatory category, including products for which the U.S. government does not wish to have advance notice before each shipment. We are working with some customers in China and the Middle East to pursue licenses from the U.S. government. It is too early to know whether these will be granted for any significant amount of revenue. The takeaway here is non of this speculative revenue is counted in their guide. So, scope for material upward revisions. It is conceivable that they could deliver close to $25B in Q4!
Many countries are awakening to the need to invest in sovereign AI infrastructure to support economic growth and industrial innovation. With investments in domestic compute capacity, nations can use their own data to train LLMs and support their local generative AI ecosystems. For example, we are working with India's government and largest tech companies including Infosys, Reliance and Tata to boost their sovereign AI infrastructure. And French private cloud provider, Scaleway, is building a regional AI cloud based on NVIDIA H100 InfiniBand and NVIDIA's AI Enterprise software to fuel advancement across France and Europe.
National investment in compute capacity is a new economic imperative and serving the sovereign AI infrastructure market(this is govt spending) represents a multi-billion dollar opportunity over the next few years. From a product perspective, the vast majority of revenue in Q3 was driven by the NVIDIA HGX platform based on our Hopper GPU architecture with lower contribution from the prior generation Ampere GPU architecture.
The new L40S GPU built for industry standard servers began to ship, supporting training and inference workloads across a variety of consumers. This was also the first revenue quarter of our GH200 Grace Hopper Superchip, which combines our ARM-based Grace CPU with a Hopper GPU. Grace and Grace Hopper are ramping into a new multi-billion dollar product line. Grace Hopper instances are now available at GPU specialised cloud providers, and coming soon to Oracle Cloud.
Grace Hopper is also getting significant traction with supercomputing customers. Initial shipments to Los Alamos National Lab and the Swiss National Supercomputing Center took place in the third quarter. The UK government announced it will build one of the world's fastest AI supercomputers called Isambard-AI with almost 5,500 Grace Hopper Superchips.
German supercomputing center, Julich, also announced that it will build its next-generation AI supercomputer with close to 24,000 Grace Hopper Superchips and Quantum-2 InfiniBand, making it the world's most powerful AI supercomputer with over 90 exaflops of AI performance. All-in, we estimate that the combined AI compute capacity of all the supercomputers built on Grace Hopper across the U.S., Europe and Japan next year will exceed 200 exaflops with more wins to come.
Inference is contributing significantly to data centre demand, as AI is now in full production for deep learning, recommenders, chatbots, copilots and text to image generation and this is just the beginning. NVIDIA AI offers the best inference performance and versatility, and thus the lower power and cost of ownership. We are also driving a fast cost reduction curve. With the release of TensorRT-LLM, we now achieved more than 2x the inference performance for half the cost of inferencing LLMs on NVIDIA GPUs.
Announced the latest member of the Hopper family, the H200, which will be the first GPU to offer HBM3e, faster, larger memory to further accelerate generative AI and LLMs. It moves inference speed up to another 2x compared to H100 GPUs for running LLMs like Norma2 (ph). Combined, TensorRT-LLM and H200, increased performance or reduced cost by 4x in just one year. With our customers changing their stack, this is a benefit of CUDA and our architecture compatibility. Built in redundancy.
Compared to the A100, H200 delivers an 18x performance increase for inferencing models like GPT-3, allowing customers to move to larger models and with no increase in latency. Amazon Web Services, Google Cloud, Microsoft Azure and Oracle Cloud will be among the first CSPs to offer H200-based instances starting next year. This will be a huge capex in 2024-Growth accelerating
At last week's Microsoft Ignite, we deepened and expanded our collaboration with Microsoft across the entire stack (hardware, software, API, networking). We introduced an AI foundry service for the development and tuning of custom generative AI enterprise applications running on Azure. Customers can bring their domain knowledge and proprietary data and we help them build their AI models using our AI expertise and software stock in our DGX cloud, all with enterprise grade security and support.
SAP and Amdocs are the first customers of the NVIDIA AI foundry service on Microsoft Azure. In addition, Microsoft will launch new confidential computing instances based on the H100. The H100 remains the top performing and most versatile platform for AI training and by a wide margin, as shown in the latest MLPerf industry benchmark results.
There training cluster included more than 10,000 H100 GPUs or 3x more than in June, reflecting very efficient scaling. Efficient scaling is a key requirement in generative AI, because LLMs are growing by an order of magnitude every year. Microsoft Azure achieved similar results on a nearly identical cluster, demonstrating the efficiency of NVIDIA AI in public cloud deployments. Remember scaling! Intel chips don't scale. Not sure about AMDs because no one is using it yet. They are years behind.
Networking now exceeds a $10 billion annualised revenue run rate. Strong growth was driven by exceptional demand for InfiniBand, which grew fivefold year-on-year. InfiniBand is critical to gaining the scale and performance needed for training LLMs. Microsoft made this very point last week, highlighting that Azure uses over 29,000 miles of InfiniBand cabling, enough to circle the globe. Nvidia purchased Mellanox a few years ago for $7B-a great investment
Expanding NVIDIA networking into the Ethernet space. Our new Spectrum-X end-to-end Ethernet offering with technologies, purpose built for AI, will be available in Q1 next year. With support from leading OEMs, including Dell, HPE and Lenovo. Spectrum-X can achieve 1.6x higher networking performance for AI communication compared to traditional Ethernet offerings.
Our latest DGX cloud customer announcement was this morning as part of an AI research collaboration with Gentech, the biotechnology pioneer also plans to use our BioNeMo LLM framework to help accelerate and optimise their AI drug discovery platform. We now have enterprise AI partnership with Adobe, Dropbox, Getty, SAP, ServiceNow, Snowflake and others to come.
Gaming revenue of $2.86 billion was up 15% sequentially and up more than 80% year-on-year with strong demand in the important back-to-school shopping season with NVIDIA RTX ray tracing and AI technology now available at price points as low as $299. We entered the holidays with the best-ever line-up for gamers and creators.
Gaming has doubled relative to pre-COVID levels even against the backdrop of lackluster PC market performance.
Generative AI is quickly emerging as the new pillar app for high performance PCs. NVIDIA RTX GPUs to find the most performance AI PCs and workstations. We just released TensorRT-LLM for Windows, which speeds on-device LLM inference up by 4x. With an installed base of over 100 million, NVIDIA RTX is the natural platform for AI application developers.
Moving to the Pro Visualisation. Revenue of $416 million was up 10% sequentially and up 108% year-on year. NVIDIA RTX is the workstation platform of choice for professional design, engineering and simulation use cases and AI is emerging as a powerful demand driver. Early applications include inference for AI imaging in healthcare and edge AI in smart spaces and the public sector.
On Omniverse, the software platform for designing, building and operating 3D virtual worlds. Mercedes-Benz is using Omniverse powered digital twins to plan, design, build and operate its manufacturing and assembly facilities, helping it increase efficiency and reduce defects. Oxxon (ph) is also incorporating Omniverse into its manufacturing process, including end-to-end simulation for the entire robotics and automation pipeline, saving time and cost.
Moving to Automotive. Revenue was $261 million, up 3% sequentially and up 4% year-on year, primarily driven by continued growth in self-driving platforms based on NVIDIA DRIVE Orin SOC (system on chip) and the ramp of AI cockpit solutions with global OEM customers. Extended our automotive partnership of Foxconn to include NVIDIA DRIVE for our next-generation automotive SOC. Foxconn has become the ODM for EVs. Our partnership provides Foxconn with a standard AV sensor and computing platform for their customers to easily build a state-of-an-art safe and secure software defined car. The company has $15B in signed contracts to deliver this technology. The small revenues today is because they are waiting for regulatory approval and model releases to roll it out. This area has significant potential to be many billions per year. The majority of auto OEMs use Nvidia Drive.
GAAP gross margin expanded to 74% and non-GAAP gross margin to 75%, driven by higher Data Centre sales and lower net inventory reserve, including a 1 percentage point benefit from the release of previously reserved inventory related to the Ampere GPU architecture products. Sequentially, GAAP operating expenses were up 12% and non-GAAP operating expenses were up 10%, primarily reflecting increased compensation and benefits. FYI, reserves against inventory are a way of making an allowance for lower selling prices so they would charge the P&L with the estimated cost of the write down. They are saying here, given high sales prices they didn't need the provision and have written it back as a gain to the PNL via COGS.
Total revenue is expected to be $20 billion, plus or minus 2%. We expect strong sequential growth to be driven by Data Centre, with continued strong demand for both compute and networking. Gaming will likely decline sequentially as it is now more aligned with notebook seasonality.
GAAP and non-GAAP gross margins are expected to be 74.5% and 75.5%, respectively, plus or minus 50 basis points. non-GAAP operating expenses are expected to be approximately $2.2 billion. This is where the operating leverage 'fly-wheel' comes from. Guided revenue +$2B (@75% margin), net 1.5B and only 150M additional operating expenses.
7 out of 10 vehicle OEMs use Nvidia for driver assist systems and regardless, their tech is as good as Tesla. So if rob taxis become a thing, Nvidia is in a good position to generate significant revenues. In the interim they have agreements to supply $15B in assist systems between now and 2028
The world is changing and every industry is moving quickly to accelerated computing. Nvidia is by far the Nr 1 platform provider. Trading now at a Fwd PE of around 20, we consider the stock to be reasonably priced in absolute terms and under valued when looking at their growth.
The best is yet to come.
Q3 was another record quarter. Revenue of $18.1 billion was up 34% sequentially and up more than 200% year-on-year and well above our outlook for $16 billion.
The continued ramp of the NVIDIA HGX platform based on our Hopper Tensor Core GPU architecture, along with InfiniBand end-to-end networking, drove record revenue of $14.5 billion, up 41% sequentially and up 279% year-on-year.
NVIDIA HGX with InfiniBand together are essentially the reference architecture for AI supercomputers and data center infrastructures. Some of the most exciting generative AI applications are built and run on NVIDIA, including Adobe Firefly, ChatGPT, Microsoft 365 Copilot, CoAssist, now assist with ServiceNow and Zoom AI Companion. Our Data Center compute revenue quadrupled from last year and networking revenue nearly tripled.
Investments in infrastructure for training and inferencing large language models, deep learning, recommender systems and generative AI applications is fueling strong broad-based demand for NVIDIA accelerated computing. Inferencing is now a major workload for NVIDIA AI computing. Consumer Internet companies and enterprises drove exceptional sequential growth in Q3, comprising approximately half of our Data Center revenue and outpacing total growth.
Companies like Meta are in full production with deep learning, recommender systems and also investing in generative AI to help advertisers optimise images and text. Most major consumer Internet companies are racing to ramp up generative AI deployment. The enterprise wave of AI adoption is now beginning. Enterprise software companies such as Adobe, Databricks, Snowflake and ServiceNow are adding AI copilots and the systems to their platforms. And broader enterprises are developing custom AI for vertical industry applications such as Tesla in autonomous driving.
Cloud service providers drove roughly half of Data Centre revenue in the quarter. Demand was strong from all hyperscale CSPs(cloud service providers), as well as from a broadening set of GPU-specialised CSPs globally that are rapidly growing to address the new market opportunities in AI. NVIDIA H100 Tensor Core GPU instances are now generally available in virtually every cloud with instances in high demand.
We have significantly increased supply every quarter this year to meet strong demand and expect to continue to do so next year! This is where every quarter will be higher than the preceding. The company is still significantly constrained. Current orders are booking 2025 deliver.
We will also have a broader and faster product launch cadence to meet the growing and diverse set of AI opportunities. This is next Gen chips, Blackwell and then 'X'.Towards the end of the quarter, the U.S. government announced a new set of export control regulations for China and other markets, including Vietnam and certain countries in the Middle East. These regulations require licenses for the export of a number of our products, including our Hopper and Ampere 100 and 800 series and several others.
Our sales to China and other affected destinations derived from products that are now subject to licensing requirements have consistently contributed approximately 20% to 25% of Data Centre revenue over the past few quarters. We expect that our sales to these destinations will decline significantly in the fourth quarter. So we believe will be more than offset by strong growth in other regions. The U.S. government designed the regulation to allow the U.S. industry to provide data centre compute products to markets worldwide, including China. Takeaway, they still guided 20B, think what they could have delivered and the chip issues re china will be resolved with accepted redesigns. Nvidia is working on it now with Govt assistance.
Continuing to compete worldwide as the regulations encourage, promotes U.S. technology leadership, spurs economic growth and supports U.S. jobs. For the highest performance levels, the government requires licenses. For lower performance levels, the government requires a streamlined prior notification process. And for products even lower performance levels, the government does not require any notice at all.
Following the government's clear guidelines, we are working to expand our Data Centre product portfolio to offer compliance solutions for each regulatory category, including products for which the U.S. government does not wish to have advance notice before each shipment. We are working with some customers in China and the Middle East to pursue licenses from the U.S. government. It is too early to know whether these will be granted for any significant amount of revenue. The takeaway here is non of this speculative revenue is counted in their guide. So, scope for material upward revisions. It is conceivable that they could deliver close to $25B in Q4!
Many countries are awakening to the need to invest in sovereign AI infrastructure to support economic growth and industrial innovation. With investments in domestic compute capacity, nations can use their own data to train LLMs and support their local generative AI ecosystems. For example, we are working with India's government and largest tech companies including Infosys, Reliance and Tata to boost their sovereign AI infrastructure. And French private cloud provider, Scaleway, is building a regional AI cloud based on NVIDIA H100 InfiniBand and NVIDIA's AI Enterprise software to fuel advancement across France and Europe.
National investment in compute capacity is a new economic imperative and serving the sovereign AI infrastructure market(this is govt spending) represents a multi-billion dollar opportunity over the next few years. From a product perspective, the vast majority of revenue in Q3 was driven by the NVIDIA HGX platform based on our Hopper GPU architecture with lower contribution from the prior generation Ampere GPU architecture.
The new L40S GPU built for industry standard servers began to ship, supporting training and inference workloads across a variety of consumers. This was also the first revenue quarter of our GH200 Grace Hopper Superchip, which combines our ARM-based Grace CPU with a Hopper GPU. Grace and Grace Hopper are ramping into a new multi-billion dollar product line. Grace Hopper instances are now available at GPU specialised cloud providers, and coming soon to Oracle Cloud.
Grace Hopper is also getting significant traction with supercomputing customers. Initial shipments to Los Alamos National Lab and the Swiss National Supercomputing Center took place in the third quarter. The UK government announced it will build one of the world's fastest AI supercomputers called Isambard-AI with almost 5,500 Grace Hopper Superchips.
German supercomputing center, Julich, also announced that it will build its next-generation AI supercomputer with close to 24,000 Grace Hopper Superchips and Quantum-2 InfiniBand, making it the world's most powerful AI supercomputer with over 90 exaflops of AI performance. All-in, we estimate that the combined AI compute capacity of all the supercomputers built on Grace Hopper across the U.S., Europe and Japan next year will exceed 200 exaflops with more wins to come.
Inference is contributing significantly to data centre demand, as AI is now in full production for deep learning, recommenders, chatbots, copilots and text to image generation and this is just the beginning. NVIDIA AI offers the best inference performance and versatility, and thus the lower power and cost of ownership. We are also driving a fast cost reduction curve. With the release of TensorRT-LLM, we now achieved more than 2x the inference performance for half the cost of inferencing LLMs on NVIDIA GPUs.
Announced the latest member of the Hopper family, the H200, which will be the first GPU to offer HBM3e, faster, larger memory to further accelerate generative AI and LLMs. It moves inference speed up to another 2x compared to H100 GPUs for running LLMs like Norma2 (ph). Combined, TensorRT-LLM and H200, increased performance or reduced cost by 4x in just one year. With our customers changing their stack, this is a benefit of CUDA and our architecture compatibility. Built in redundancy.
Compared to the A100, H200 delivers an 18x performance increase for inferencing models like GPT-3, allowing customers to move to larger models and with no increase in latency. Amazon Web Services, Google Cloud, Microsoft Azure and Oracle Cloud will be among the first CSPs to offer H200-based instances starting next year. This will be a huge capex in 2024-Growth accelerating
At last week's Microsoft Ignite, we deepened and expanded our collaboration with Microsoft across the entire stack (hardware, software, API, networking). We introduced an AI foundry service for the development and tuning of custom generative AI enterprise applications running on Azure. Customers can bring their domain knowledge and proprietary data and we help them build their AI models using our AI expertise and software stock in our DGX cloud, all with enterprise grade security and support.
SAP and Amdocs are the first customers of the NVIDIA AI foundry service on Microsoft Azure. In addition, Microsoft will launch new confidential computing instances based on the H100. The H100 remains the top performing and most versatile platform for AI training and by a wide margin, as shown in the latest MLPerf industry benchmark results.
There training cluster included more than 10,000 H100 GPUs or 3x more than in June, reflecting very efficient scaling. Efficient scaling is a key requirement in generative AI, because LLMs are growing by an order of magnitude every year. Microsoft Azure achieved similar results on a nearly identical cluster, demonstrating the efficiency of NVIDIA AI in public cloud deployments. Remember scaling! Intel chips don't scale. Not sure about AMDs because no one is using it yet. They are years behind.
Networking now exceeds a $10 billion annualised revenue run rate. Strong growth was driven by exceptional demand for InfiniBand, which grew fivefold year-on-year. InfiniBand is critical to gaining the scale and performance needed for training LLMs. Microsoft made this very point last week, highlighting that Azure uses over 29,000 miles of InfiniBand cabling, enough to circle the globe. Nvidia purchased Mellanox a few years ago for $7B-a great investment
Expanding NVIDIA networking into the Ethernet space. Our new Spectrum-X end-to-end Ethernet offering with technologies, purpose built for AI, will be available in Q1 next year. With support from leading OEMs, including Dell, HPE and Lenovo. Spectrum-X can achieve 1.6x higher networking performance for AI communication compared to traditional Ethernet offerings.
Our latest DGX cloud customer announcement was this morning as part of an AI research collaboration with Gentech, the biotechnology pioneer also plans to use our BioNeMo LLM framework to help accelerate and optimise their AI drug discovery platform. We now have enterprise AI partnership with Adobe, Dropbox, Getty, SAP, ServiceNow, Snowflake and others to come.
Gaming revenue of $2.86 billion was up 15% sequentially and up more than 80% year-on-year with strong demand in the important back-to-school shopping season with NVIDIA RTX ray tracing and AI technology now available at price points as low as $299. We entered the holidays with the best-ever line-up for gamers and creators.
Gaming has doubled relative to pre-COVID levels even against the backdrop of lackluster PC market performance.
Generative AI is quickly emerging as the new pillar app for high performance PCs. NVIDIA RTX GPUs to find the most performance AI PCs and workstations. We just released TensorRT-LLM for Windows, which speeds on-device LLM inference up by 4x. With an installed base of over 100 million, NVIDIA RTX is the natural platform for AI application developers.
Moving to the Pro Visualisation. Revenue of $416 million was up 10% sequentially and up 108% year-on year. NVIDIA RTX is the workstation platform of choice for professional design, engineering and simulation use cases and AI is emerging as a powerful demand driver. Early applications include inference for AI imaging in healthcare and edge AI in smart spaces and the public sector.
On Omniverse, the software platform for designing, building and operating 3D virtual worlds. Mercedes-Benz is using Omniverse powered digital twins to plan, design, build and operate its manufacturing and assembly facilities, helping it increase efficiency and reduce defects. Oxxon (ph) is also incorporating Omniverse into its manufacturing process, including end-to-end simulation for the entire robotics and automation pipeline, saving time and cost.
Moving to Automotive. Revenue was $261 million, up 3% sequentially and up 4% year-on year, primarily driven by continued growth in self-driving platforms based on NVIDIA DRIVE Orin SOC (system on chip) and the ramp of AI cockpit solutions with global OEM customers. Extended our automotive partnership of Foxconn to include NVIDIA DRIVE for our next-generation automotive SOC. Foxconn has become the ODM for EVs. Our partnership provides Foxconn with a standard AV sensor and computing platform for their customers to easily build a state-of-an-art safe and secure software defined car. The company has $15B in signed contracts to deliver this technology. The small revenues today is because they are waiting for regulatory approval and model releases to roll it out. This area has significant potential to be many billions per year. The majority of auto OEMs use Nvidia Drive.
GAAP gross margin expanded to 74% and non-GAAP gross margin to 75%, driven by higher Data Centre sales and lower net inventory reserve, including a 1 percentage point benefit from the release of previously reserved inventory related to the Ampere GPU architecture products. Sequentially, GAAP operating expenses were up 12% and non-GAAP operating expenses were up 10%, primarily reflecting increased compensation and benefits. FYI, reserves against inventory are a way of making an allowance for lower selling prices so they would charge the P&L with the estimated cost of the write down. They are saying here, given high sales prices they didn't need the provision and have written it back as a gain to the PNL via COGS.
Total revenue is expected to be $20 billion, plus or minus 2%. We expect strong sequential growth to be driven by Data Centre, with continued strong demand for both compute and networking. Gaming will likely decline sequentially as it is now more aligned with notebook seasonality.
GAAP and non-GAAP gross margins are expected to be 74.5% and 75.5%, respectively, plus or minus 50 basis points. non-GAAP operating expenses are expected to be approximately $2.2 billion. This is where the operating leverage 'fly-wheel' comes from. Guided revenue +$2B (@75% margin), net 1.5B and only 150M additional operating expenses.
Edited by AdamIM on Wednesday 22 November 14:01
Analyst (Big banks) opinions are starting to drop
Bank of America say ' Jensen Huang led Nvidia is still in the first 25% of converting $250B global compute(annually ). He now believes the company could earn as much as $40/share in 3 years. That would be 100B net profit. The analyst boosts his target to $700
Rosenblatts Hans Mosesmann reiterated his Buy and $1,100 target. calling the results 'epic'
Wedbush Securities said 'nvidia results show a positive read through for Super Micro (PHO)'
Bank of America say ' Jensen Huang led Nvidia is still in the first 25% of converting $250B global compute(annually ). He now believes the company could earn as much as $40/share in 3 years. That would be 100B net profit. The analyst boosts his target to $700
Rosenblatts Hans Mosesmann reiterated his Buy and $1,100 target. calling the results 'epic'
Wedbush Securities said 'nvidia results show a positive read through for Super Micro (PHO)'
Morning All,
Due to a sale of my company and a mix of shares (CGT/good) but also ‘bonus’ incentives which fall under PAYE, it looks like I am going to hit the earnings threshold for tapered annual allowance to kick in. As I understand it, due to the amount I am going to receive though PAYE in 23/24 I am going to lose the maximum amount, so my annual relief will fall to the minimum of £10,000.
I have been making pension payments this year through salary sacrifice on a monthly basis to the old maximum of 40k. If nothing changed I would have 20k of headroom (£60k 23/24 allowance less £40k contribution).
I have some unused allowance from previous years which I could normally carry forward.
Q: I have done some research online but can’t work out if you can carry forward previous year(s) unused allowances to contribute to the total allowance this year (£60k + any previous year amounts) which is then reduced when I hit the earning threshold. If possible, it might mean I can have more than the minimum of £10k left
Q: I think I should stop contributing this year as it will only make matters worse. Ie I have made 8 pension payments (Apr to Nov) £3.33k through salary sacrifice (£26k total).
I appreciate that this is complex and I will need formal advice I suspect but I need to move quickly if I am to stop matters getting worse from a pension tax clawback perspective, so any thoughts or views much appreciated
Due to a sale of my company and a mix of shares (CGT/good) but also ‘bonus’ incentives which fall under PAYE, it looks like I am going to hit the earnings threshold for tapered annual allowance to kick in. As I understand it, due to the amount I am going to receive though PAYE in 23/24 I am going to lose the maximum amount, so my annual relief will fall to the minimum of £10,000.
I have been making pension payments this year through salary sacrifice on a monthly basis to the old maximum of 40k. If nothing changed I would have 20k of headroom (£60k 23/24 allowance less £40k contribution).
I have some unused allowance from previous years which I could normally carry forward.
Q: I have done some research online but can’t work out if you can carry forward previous year(s) unused allowances to contribute to the total allowance this year (£60k + any previous year amounts) which is then reduced when I hit the earning threshold. If possible, it might mean I can have more than the minimum of £10k left
Q: I think I should stop contributing this year as it will only make matters worse. Ie I have made 8 pension payments (Apr to Nov) £3.33k through salary sacrifice (£26k total).
I appreciate that this is complex and I will need formal advice I suspect but I need to move quickly if I am to stop matters getting worse from a pension tax clawback perspective, so any thoughts or views much appreciated
seapod said:
Morning All,
Due to a sale of my company and a mix of shares (CGT/good) but also ‘bonus’ incentives which fall under PAYE, it looks like I am going to hit the earnings threshold for tapered annual allowance to kick in. As I understand it, due to the amount I am going to receive though PAYE in 23/24 I am going to lose the maximum amount, so my annual relief will fall to the minimum of £10,000.
I have been making pension payments this year through salary sacrifice on a monthly basis to the old maximum of 40k. If nothing changed I would have 20k of headroom (£60k 23/24 allowance less £40k contribution).
I have some unused allowance from previous years which I could normally carry forward.
Q: I have done some research online but can’t work out if you can carry forward previous year(s) unused allowances to contribute to the total allowance this year (£60k + any previous year amounts) which is then reduced when I hit the earning threshold. If possible, it might mean I can have more than the minimum of £10k left
Q: I think I should stop contributing this year as it will only make matters worse. Ie I have made 8 pension payments (Apr to Nov) £3.33k through salary sacrifice (£26k total).
I appreciate that this is complex and I will need formal advice I suspect but I need to move quickly if I am to stop matters getting worse from a pension tax clawback perspective, so any thoughts or views much appreciated
Hi Seapod,Due to a sale of my company and a mix of shares (CGT/good) but also ‘bonus’ incentives which fall under PAYE, it looks like I am going to hit the earnings threshold for tapered annual allowance to kick in. As I understand it, due to the amount I am going to receive though PAYE in 23/24 I am going to lose the maximum amount, so my annual relief will fall to the minimum of £10,000.
I have been making pension payments this year through salary sacrifice on a monthly basis to the old maximum of 40k. If nothing changed I would have 20k of headroom (£60k 23/24 allowance less £40k contribution).
I have some unused allowance from previous years which I could normally carry forward.
Q: I have done some research online but can’t work out if you can carry forward previous year(s) unused allowances to contribute to the total allowance this year (£60k + any previous year amounts) which is then reduced when I hit the earning threshold. If possible, it might mean I can have more than the minimum of £10k left
Q: I think I should stop contributing this year as it will only make matters worse. Ie I have made 8 pension payments (Apr to Nov) £3.33k through salary sacrifice (£26k total).
I appreciate that this is complex and I will need formal advice I suspect but I need to move quickly if I am to stop matters getting worse from a pension tax clawback perspective, so any thoughts or views much appreciated
On Q1, you can still carry forward unused allowances, 3 years subject to tapering in those years.
I will inform Nik of your question, it may be that you have a chat about your specific situation.
Regards
Adam
Adam, you seem to be very active on here so perhaps you could get this addressed.
My daughter has been emailing Nik since earlier this year. After no response to her last 2 emails in September and November, she copied in Steve last weekend.
Still nothing.
Can you give someone a nudge please? She was wanting to move her pension and ISA to you, but it's almost December and she's been asking since June.
If I hadn't recommended you so highly, she'd have given up and gone elsewhere months ago.
My daughter has been emailing Nik since earlier this year. After no response to her last 2 emails in September and November, she copied in Steve last weekend.
Still nothing.
Can you give someone a nudge please? She was wanting to move her pension and ISA to you, but it's almost December and she's been asking since June.
If I hadn't recommended you so highly, she'd have given up and gone elsewhere months ago.
Gassing Station | Finance | Top of Page | What's New | My Stuff