Tag Archives: Integrated Circuits

Opinion: This record number in Nvidia earnings is a scary sight

Nvidia Corp.’s financial results had a bit of a surprise for investors, and not on the good side — product inventories doubled to a record high as the chip company gears up for a questionable holiday season.

Nvidia reported fiscal third-quarter revenue that was slightly better than analysts’ reduced expectations Wednesday, but the numbers weren’t that great. Revenue fell 17% to $5.9 billion, while earnings were cut in half thanks to a $702 million inventory charge, largely relating to slower data-center demand in China.

Gaming revenue in the quarter fell 51% to $1.57 billion. Nvidia said it is working with its retail partners to help move the currently high-channel inventories.

While the company was writing off the inventory for China, its own new product inventory was growing. Nvidia
NVDA,
-4.54%
reported that its overall product inventory nearly doubled to $4.45 billion in the fiscal third quarter, compared with $2.23 billion a year ago and $3.89 billion in the prior quarter. Executives cited its coming product launches, designed around its new Ada and Hopper architectures, when asked about the inventory gains.

In the semiconductor industry, high inventories can make investors nervous, especially after the industry had so many supply constraints in recent years that quickly swung to a glut of chips in 2022. With doubts about demand for gaming cards and consumers’ willingness to spend amid sky-high inflation this holiday season, having all that product on hand just amps up the nerves.

Full earnings coverage: Nvidia profit chopped in half, but tweaked servers to China offset earlier $400 million warning

Chief Financial Officer Colette Kress told MarketWatch in a telephone interview Wednesday that the company’s high level of inventories were commensurate with its high levels of revenue.

“I do believe….it is our highest level of inventory,” she said. “They go hand in hand.” Kress said she was confident in the success of Nvidia’s upcoming product launches.

Nvidia’s revenue reached a peak in the April 2022 quarter with $8.3 billion, and in the past two quarters revenue has slowed, with gaming demand sluggish amid a transition to a new cycle, and a decline in China data-center demand due to COVID-19 lockdowns and U.S. government restrictions.

For its data-center customers, the new architectures promise major advances in computing power and artificial-intelligence features, with Nvidia planning to ship the equivalent of a supercomputer in a box with its new products over the next year. Those types of advanced products weigh on inventory totals even more, Kress said, because of the price of the total package.

“It’s about the complexity of the system we are building, that is what drives the inventory, the pieces of that together,” Kress said.

Bernstein Research analyst Stacy Rasgon believes that products based on Hopper will begin shipping over the next several quarters, “at materially higher price points.” He said in a recent note that he believes Nvidia’s numbers were likely hitting a bottom in this quarter.

“We remain positive on the Hopper ramp into next year, and believe numbers have at this point likely reached close to bottom, with new cycles brewing and an attractive secular story even without China potential,” Rasgon said in an earnings preview note Tuesday.

Read also: Warren Buffett’s chip-stock purchase is a classic example of why you want to be ‘greedy only when others are fearful’

Nvidia Chief Executive Jensen Huang reminded investors on a conference call that the company’s inventories are “never zero,” and said everyone is enthusiastic about the upcoming launches. But it doesn’t take too long of a memory to conjure up a time when Nvidia went into a holiday with an inventory backlog that included new architecture and greatly disappointed investors: Four years ago, Huang had to cut his forecast for holiday earnings twice amid a “crypto hangover” with similar dynamics to the current moment

Investors need faith that this holiday season will not be the same, even as demand for some videogame products declines after a pandemic boom just as the market for cryptocurrency — some of which has been mined with Nvidia products — hits a rough patch. Huang said that Nvidia’s RTX 4080 and 4090 graphics cards based on the Ada Lovelace architecture had an “exceptional launch,” and sold out.

Nvidia shares gained more than 2% in after-hours trading Wednesday, suggesting that some are betting that this time will be different. That enthusiasm needs to translate into revenue for Nvidia so that this big gain in inventories does not end up being part of another write-down at some point in the future.

Read original article here

AMD’s RDNA 3 GPUs are Way Cheaper Than the RTX 4090

Screenshot: AMD

Back in August, Dr. Lisa Su introduced the world to AMD’s newest iteration of CPU technology, the Ryzen 7000 Series–but she didn’t stop there. We got an announcement for an announcement: RDNA 3, AMD’s next generation of GPU technology. Well, today is November 3rd, and we now know more about AMD’s answer to the RTX 40 Series.

Dr. Su began the presentation by reflecting on the Ryzen 7000 Series release and stating AMD’s ambitious goals, much like she did during the previous presentation. For RDNA3, she reiterated the company’s commitment to energy efficiency and performance.

“At the forefront of what we’re doing, it’s all about power and energy efficiency. We want to make sure that we continue to innovate around performance-per-watt leadership to enable all of the gamer upgrades with fantastic performance, but at a reasonable power.” – Dr. Lisa Su

Starting with the new chiplet design, RDNA 3 takes a modular approach with the intention of optimizing the efficiency of the overall GPU design. Much like the Ryzen CPU family, RDNA3 will utilize a mixed chiplet architecture. With a 5nm graphics die (GCD) compute unit containing all of the shaders, display engines, and updated media engine, the GCD is paired with a 6nm Memory Cache Die (MCD) that consists of the GDDR6 controller as well as 96 MB of AMD’s Infinity Cache–2nd Generation Infinity Cache, that is!

With this new design, the RNDA 3 chiplet will have an interconnect speed of up to 5.3 TB/s (a 2.7x increase over RDNA 2), enabling up to 61 TFLOPS of compute. All of this will be supported by up to 24 GB of GDDR6 with a memory bus up to 384-bit (not the GDDR6X we saw in the RTX 4090) and, according to Dr. Su, will enable RDNA 3 GPUs to achieve up to 54% greater performance-per-watt over the previous generation.

So, what is this magical mystery GPU? Well, it’s actually two GPUs: the Radeon RX 7900 XTX and RX 7900 XT.

The Radeon RX 7900 XTX comes with 24 GB of GDDR6 and Radeon RX 7900 XT comes with 20 GB of GDDR6. Engineered as both 4K and 8K gaming GPUs, the Radeon RX 7900 XTX and 7900 XT have a number of updates over the previous generation that will help push it into the future of gaming.

Starting with dedicated AI acceleration, RDNA 3 is said to improve AI-based functions of the GPU by 2.7x and ray tracing instructions by 1.8x over RDNA 2. In rendering applications–including ray tracing, this new architecture is said to obtain up to 50% more performance per compute unit and double the instructions per clock. This is a much-needed leap for Radeon graphics to compete within this particular space of graphics processing!

Screenshot: AMD

But, there’s more:

AMD’s new Radiance Display is the engine pushing all of that data into the display. The engine will support 12 bit-per-channel color with up to 68 billion colors as well as higher refresh rates. Thanks in no small part to DisplayPort 2.1 and HDMI 2.1 adoption, the RX 7900 XTX and RX 7900 XT will support refresh rates of up to 900 Hz at 1440p, 480 Hz at 4K, and 165 Hz at 5K.

Alongside of the Radiance Display engine, AMD also unveiled a new dual media engine for simultaneous encode and decode for both AVC and HEVC formats. This engine will also support AV1 encoding and decoding, with a max resolution of 8K60. Later on in the presentation, AMD teased future support for AV1 encoding support within OBS as well as other popular video streaming and editing software. This teaser also included a future feature called SmartAccess Video, which will leverage Ryzen CPUs and Radeon GPUs together to supposedly provide up to a 30% uplift in 4K multi stream encoding.

When it comes to gaming performance, the RX 7900 XTX is purported to operate up to 1.7x better than AMD’s former flagship GPU, the Radeon RX 6950 XT, in rasterization and up to 1.6x better in ray tracing games. Using FidelityFX Super Resolution (FSR), the charts AMD showed for the 7900 XTX showed several titles running at frame rates far exceeding 200 FPS at 4K. One standout was Valorant, which showed the RX 7900 XTX running at 704 FPS! Lots of bold claims here, but we’ll have to see for ourselves once the cards are in the hands of 3rd-party reviewers!

When it comes to the actual specifications, the RX 7900 XTX will have 96 compute units with a game clock of 2.3 GHz. All of this is said to run at a total board power draw of 355 W. For context, that is 95 W less than NVIDIA’s RTX 4090 Founder’s Edition while hovering close to the RX 6950 XT’s typical board power. The RX 7900 XT will have 84 compute units with a game clock of 2 GHz and a total board power of 300 W.

Oh yeah, and neither one of them will require a special cable to power them.

This is just the hardware, though. AMD took some time to talk about FSR adoption, the uplift in performance seen within FSR2, and how well RDNA 3 operates with it enabled. One example the company showed off was Ubisoft’s Assassin’s Creed: Valhalla running at 96 frames per second at 8K. It was a short clip, but pretty wild to see all the same. But AMD wanted us to know that FSR isn’t stopping at FSR2. There’s a new iteration, FSR3. At the time of the announcement, AMD said that Radeon users can expect up to 2x more frames per second over FSR2 and that the technology would be available in 2023.

Continuing the conversation about software, AMD’s Frank Azor shared updates to Radeon Adrenaline software, including a new feature coming in the first half of 2023 called AMD HYPR-RX. This feature will be a one-button optimizer to give AMD systems the best possible performance without having to make all of the adjustments yourself.

Screenshot: AMD

Team Red also shared its commitment to providing the best CPU and GPU unified experience by working with system integrators to bring the AMD Advantage line to the desktop platform. This means pairing AMD GPUs and CPUs together in system configurations carefully curated by AMD for the supposed best possible AMD experience.

The Radeon RX 7900 XTX and Radeon RX 7900 XT will be available on December 13th, 2022. And the price? $999 and $899 USD, respectively. This is an amazing distinction from NVIDIA’s $1599 flagship.

We look forward to seeing just how these GPUs perform out in the wild! Let us know in the comments what you are most excited for in AMD’s announcement today and if you plan on making an upgrade before the year is through.

Read original article here

EVGA, Big Graphics Card Maker, Has Messy Breakup With Nvidia

Times have been tough.
Image: Kotaku / San Francisco Chronicle / Hearst Newspapers (Getty Images)

And now for something that no one saw coming: EVGA, one of the most prominent third-party PC graphics card manufacturers, and a favorite brand among PC gamers for quality parts and reliable warranties backed by solid customer service, is terminating its longtime relationship with Nvidia. What’s more, the company reportedly said that it won’t be pursuing partnerships with competing silicon giants like AMD or Intel, either. It seems like EVGA is just done with GPUs.

Kotaku has reached out to EVGA for comment.

News of EVGA’s seemingly sudden decision to stop manufacturing GPUs broke via the popular YouTubers GamersNexus and Jayztwocents. Personalities from both channels say that they were invited to a private meeting with EVGA staff, including CEO Andrew Han. In the meeting, EVGA reportedly laid out its desire and intention to break away from Nvidia, citing multiple frustrations with the partnership.

These sore spots mostly concern what Han describes as Nvidia’s reluctance to share essential information about its products with partners until that same information is made available to the public, often onstage at a press conference; that it believes Nvidia is undercutting partners like EVGA by selling its own “Founders’ Edition” cards at a lower price; and a sense among partners that Nvidia just doesn’t value their patronage.

GamersNexus has a very thorough breakdown of the meeting and this news in its video.

GamersNexus

EVGA’s most senior management made its decision to break away from Nvidia back in April, but kept the decision strictly confidential. Though EVGA, a company that is so often known and valued for great GPUs and reliable customer service, is leaving the GPU market, the company reportedly intends to stay in business. However, it won’t be expanding into new product categories, GamersNexus reports. And while the company does make and sell other PC components such as motherboards, cases, and power supplies, the loss of the GPU side of its business is likely to pose challenges for its 280 worldwide staffers.

GamersNexus’ Steve Burke reports that EVGA is looking to reallocate staff to different projects to keep everyone employed. The company laid off 20 percent of its Taiwan employees earlier this year, and now several people whose jobs solely revolved around GPU manufacturing and development don’t have an obvious job to perform.

While EVGA will continue to sell RTX 30-series cards, it expects to run out of stock by the end of the year, and will be hanging on to an additional stock to service warranties and repairs. EVGA’s pledging to honor warranties for existing customers of those cards.

Today is a bittersweet day for PC gamers, as EVGA’s presence in the GPU arena will be sorely missed. On the flip side, the crypto-mining craze that has plagued the industry by buying up countless cards for mining rigs seems to be coming to an end. The prominent crypto Ethereum has finally, finally moved away from the GPU-hungry “proof of work” algorithms that contributed to the virtual decimation of available GPU stock over the last two years. As you’ve probably noticed, GPUs are once again available to buy and pricing has finally started to fall back to Earth. With the Ethereum switch, hopefully that trend will only accelerate.

Read original article here

Nvidia stock stumbles after Baird voices concern about graphics business

Shares of Nvidia Corp. shares were slipping Monday after an analyst took a more cautious outlook on the company’s gaming business.

Baird’s Tristan Gerra downgraded Nvidia’s stock
NVDA,
-5.20%
to neutral from outperform, writing of his concerns about excess inventory for consumer graphics processing units (GPUs). The shares were off 5.1% in afternoon trading.

“We believe order cancellations recently started in consumer GPUs, driven by excess inventories, a slowdown in consumer demand (reflected by an ongoing reduction in graphic cards pricing), slowdown in PC demand, and the Russia embargo,” he wrote in a note to clients.

Gerra thinks that the consensus view might be underestimating the percentage of Nvidia’s consumer GPU business that’s linked to Russia. Additionally, he believes that demand has dampened in China, which he said represents an estimated 25% to 30% of the market for consumer GPUs.

Further, he worries about the upcoming Ethereum fork, which is expected to drive reselling of GPUs no longer needed for Ethereum mining. That dynamic could put additional pressure on prices, Gerra wrote.

Read: Ethereum’s major upgrade is coming. Should you be more bullish on it than bitcoin?

He now projects that Nvidia’s gaming revenue could be flat to slightly up on a sequential basis in the fiscal second quarter, as well as down by a mid-single-digit rate in the fiscal third quarter.

Gerra said that the forecast for Nvidia’s data-center revenue is still “very strong,” though he has questions about “whether new expected delays in the ramp of Sapphire Rapids could impact mainstream server refreshes in C2H and affect shipments later this year.”

He lowered his price target on Nvidia’s stock to $225 from $360. The shares have lost 21.2% over the past three months, as the S&P 500
SPX,
-1.69%
has dropped 6.0%.

Read original article here

SoftBank Pitches IPO for Arm After Deal With Nvidia Falls Through

TOKYO—After a deal that could have been worth $80 billion to his company fell apart,

SoftBank Group Corp.

9984 5.85%

Chief Executive

Masayoshi Son

is playing salesman for Plan B—an initial public offering of chip designer Arm.

Mr. Son sounded as if he were on a roadshow for investors at a news conference in Tokyo on Tuesday. He said Arm is entering a “golden period” of high demand for the chips it helps create in smartphones, electric vehicles and computer-server farms operated by the likes of

Amazon.com Inc.

The pitch came hours after the Japanese investment and technology conglomerate said it was abandoning plans to sell Arm to Nvidia Corp.—in what would have been the largest semiconductor deal on record—because antitrust concerns stood in the way.

Mr. Son said he was surprised to see the backlash not only from U.S. regulators who sued to block the deal in December but also big tech companies that rely on Arm’s chip designs.

“We saw strong opposition because Arm is one of the most important and essential companies that most companies in the IT industry or in Silicon Valley rely on, either directly or indirectly,” he said.

SoftBank paid $32 billion when it acquired the U.K.-based chip business in 2016. Mr. Son said the sale to Nvidia, under which SoftBank would have received both cash and Nvidia shares, could have been worth $80 billion because of a rise in Nvidia’s share price.

SoftBank now plans to pursue a public listing of Arm by March 2023. Arm shares will most likely be listed on the tech-heavy

Nasdaq Stock Market

in the U.S. because many of Arm’s clients are based in Silicon Valley, Mr. Son said.

He said SoftBank didn’t intend to keep Arm for itself because he wanted outside investors in the SoftBank-led Vision Fund, which owns a quarter of Arm, to be able to cash in through an IPO and because he wanted to give stock options as incentives to Arm employees.

Uncertainties linger around an Arm IPO, including whether the volatile semiconductor business will stay hot through this year.

Chinese tech stocks popular among U.S. investors have tumbled amid the country’s regulatory crackdown on technology firms. WSJ explains some of the new risks investors face when buying shares of companies like Didi or Tencent. Photo Composite: Michelle Inez Simon

Tech shares have fallen recently because of tightening by the Federal Reserve. Fumio Matsumoto, chief strategist at

Okasan Securities,

said that made the timing for a big IPO less than ideal, and he also observed that a strategic buyer in the chip industry might pay more for Arm because of the potential synergy effects.

Still, Mr. Matsumoto said the downturn in Silicon Valley also offered opportunities for Mr. Son, and it made sense to raise cash for his war chest from an Arm IPO. “Because technology share prices have gone through a sharp correction over the past year, we are seeing a good cycle to consider preparing” for new investments, Mr. Matsumoto said.

After a rough patch a few years ago, Arm is on track for $2.5 billion in revenue this fiscal year, which ends in March, up from $1.98 billion the previous year, SoftBank said. Arm’s operating profit, according to one type of calculation used by SoftBank, more than doubled over the past two years to a projected $900 million this fiscal year.

An array of consumer electronics companies as well as semiconductor companies, including

Apple Inc.,

Samsung Electronics Co.

and

Qualcomm Inc.,

use Arm’s designs in at least some of their chips. The designs are known for their low power consumption, making them nearly ubiquitous in mobile devices.

The collapse of the Arm deal is just one of the challenges Mr. Son is tackling in his globe-spanning investment portfolio. He said “we are in pain” over China’s crackdown on its big tech companies, which hit SoftBank investments including its most valuable one, e-commerce giant Alibaba Group Holding Ltd.

The past two years have seen some of the wildest swings in the four decades since Mr. Son started SoftBank. The pandemic, initially seen as a blow, soon emerged as a boon for many technology businesses including those in which SoftBank has invested. SoftBank shares surged, only to fall by half from their recent peak when the China troubles hit and the Arm deal ran aground.

SoftBank’s net asset value, Mr. Son’s preferred measure of the company’s finances, fell by ¥1.6 trillion, equivalent to about $14 billion, in the October-December quarter to ¥19.3 trillion. That is a fall of 30% from the peak in September 2020 and the lowest level since 2017.

Mr. Son blamed the sharp fall in Alibaba shares. The Chinese company, which once made up the majority of SoftBank’s net assets, now accounts for less than a quarter of the total.

SoftBank said it unloaded a small number of Alibaba shares to settle contracts with its lenders, but Mr. Son said SoftBank’s stake in the Chinese company remained close to a quarter.

Mr. Son, who turns 65 this year, has lost a number of top lieutenants in recent years, including Chief Operating Officer

Marcelo Claure,

who stepped down in January after a pay dispute. Mr. Son said that while he was grooming successors, he didn’t intend to step down soon.

“If I stop, I’d become an old grandpa very quickly,” he said. He boasted that when he went bowling recently, he topped 200 points in two different rounds—a fine score for an amateur. “I thought, ‘Hey, I’m still pretty young,’ ” he said.

Write to Megumi Fujikawa at megumi.fujikawa@wsj.com and Peter Landers at peter.landers@wsj.com

Copyright ©2022 Dow Jones & Company, Inc. All Rights Reserved. 87990cbe856818d5eddac44c7b1cdeb8

Read original article here

Nvidia stock jumps as analysts say data-center growth ‘has some room to run’

Nvidia Corp. shares rallied Thursday as more than half the analysts who cover the chip maker hiked their price targets following the company’s record quarter and forecast for more new highs based on data-center gains.

Nvidia
NVDA,
+6.13%
shares rallied more than 7% in morning trading to hit an intraday high of $204.95, while the PHLX Semiconductor Index
SOX,
+1.20%
gained more than 1%, and the S&P 500 index
SPX,
+0.32%
gained 0.4%. Nvidia shares last closed at a split-adjusted record high of $206.99 back on July 6, and are up 68% over the past 12 months.

For more on today’s market action, see Market Snapshot

Nvidia forecast revenue of $6.66 billion to $6.94 billion Wednesday, above Wall Street estimates at the time, and said that the “lion’s share” of the $500 million increase will come from data-center sales. That follows new records for total, gaming and data-center revenues that Nvidia reported for the quarter.

What many analysts picked up on is that demand for graphics processing units (GPUs) for cryptocurrency mining didn’t factor that much into the outlook. That came as a relief to analysts, who noted a lower crypto risk compared with 2018 when a fall in cryptocurrency values prompted many miners to sell their gaming card-powered rigs, flooding the market with second-hand cards.

Full earnings coverage: Nvidia reports record data-center and gaming revenue, but supply constraints still a concern

Nvidia broke out sales of its Cryptocurrency Mining Processors, or CMPs, which are intended to divert mining demand away from GPUs made for gamers and not expected to be material in revenue gains.

Data-center sales took up much more of the attention from analysts, however. Bernstein analyst Stacy Rasgon, who has an outperform rating on the stock and raised his price target to a $230 from $180, said that while “the company is having absolutely no trouble continuing to crush gaming,” Nvidia’s data-center story “still feels like it has some room to run.”

“The data-center story is really coming into its own now, with a sizable inflection in the near term and with prospect for the segment to equal, and potentially exceed, gaming in the not-too-distant future,” Rasgon said.

For more: Nvidia’s ARM acquisition is stalled, and there’s a deadline with more than a billion dollars at stake

Evercore ISI analyst C.J. Muse, who has an outperform rating and a $250 price target, called data-center sales a “key for the stock.”

“Data Center revenues were guided to accelerate in 3Q off a very strong comp based on strength across hyperscale and vertical customers, training and inference applications, and compute and networking technologies – the democratization of AI workloads continues to be a front and center theme here, and one we see Nvidia driving and benefiting from over for the foreseeable future,” Muse said.

Cowen analyst Matthew Ramsay, who has an outperform rating and raised his price target to $220.00 from $176.25, said that the data-center acceleration was “the most important takeaway” from the earnings call.

“We expect sustainable data-center and gaming product cycles that should drive >50%+ organic growth for the company in F’2022,” Ramsay said.

Jefferies analyst Mark Lipacis, who has a buy rating and raised his price target to $223 from $214, addressed the lower risk of another crypto-mining debacle.

Read: Crypto is reshaping the world economy, 50 years after Nixon ended the dollar’s peg to gold. Here’s how some are playing it

“We think crypto-miners are 1/10th the gaming GPU sales vs 2018,” Lipacis said. “We continue to believe the risk of a crypto-driven gaming bust is low, and expect Nvidia’s ecosystem moat and increasing software revenues to lead to additional upside surprises.”

Of the 41 analysts who cover Nvidia, 34 have buy ratings, five have hold ratings, and two have sell ratings. Of those, at least 24 analysts hiked their price targets in response to the earnings and one lowered their target, according to FactSet. That resulted in an average price target of $219.23, up from a previous $204.24.

Read original article here