Anna Bliokh
In our earlier evaluation of Lattice Semiconductor Company (NASDAQ:LSCC), we delved into the outstanding upsurge of the corporate’s progress in 2022, which surged at a formidable 28.1%, fuelled by strong performances within the Communications & Computing (26.1%) and Industrial & Automotive (41.2%) segments, the place Lattice expanded its product portfolio and elevated its competitiveness. We believed these elements enabled the corporate to edge out different opponents. Moreover, our evaluation indicated its progress was supported by the rising use of low-power FPGAs in EVs.
On this newest evaluation, we centered on Lattice’s newest FPGA developments that would drive progress past 2024. Firstly, we analyzed the enlargement of Lattice’s FPGA portfolio, increasing its low-power portfolio with Nexus and introducing mid-range FPGA options with the launch of Avant-G and Avant-X. We performed a comparative evaluation of their options in opposition to opponents’ choices to gauge their aggressive edge. Moreover, we examined the potential for Lattice to capitalize on AI-driven purposes using its mid-range FPGAs. Lastly, we delved into the outlook for FPGA adoption throughout numerous finish markets, updating our income projections for Lattice.
Growth of FPGA Product Portfolio
Firstly, we examined Lattice’s FPGA portfolio and in contrast it with prime opponents (excluding Renesas) within the desk for Low to Mid-Vary FPGAs, with low-range FPGAs of fewer than 100,000 logic cells and mid-range spanning 100,000 to 500,000 logic cells based on Lattice.
Comparability of Low to Mid-Vary FPGAs
Lattice
AMD
Microchip
Intel
Variety of Low-Vary FPGAs (<100K Logic)
3,454
2,790
1,150
2,981
Variety of Mid-Vary FPGAs (100K – 500K Logic)
199
854
301
3,573
Complete Low to Mid-range FPGAs
3,653
3,644
1,451
6,554
Click on to enlarge
Supply: DigiKey, Khaveen Investments
Primarily based on the desk, Lattice has the best variety of Low-Vary FPGAs (<100K Logic parts), highlighting its benefit on this section. In our earlier evaluation, we highlighted the corporate’s deal with “small measurement and power-efficient FPGA as a distinction to its opponents within the FPGA market together with Xilinx and Intel.”
Lattice Nexus Platform
Throughout the low-range FPGA portfolio, Lattice lately expanded its portfolio with its CrossLinkU-NX FPGA household, claiming to be the trade’s first low-power FPGAs that includes built-in USB, to speed up USB-equipped system designs and embedded imaginative and prescient purposes. As well as, the corporate highlighted in its earnings briefing Nexus as a “main contributor” to Lattice’s progress and the introduction of recent FPGAs throughout the Nexus household.
A few examples of that’s on Nexus, we have launched seven system households based mostly on Nexus. 5 of these system households are already in manufacturing. However there’s two extra that go into manufacturing this 12 months.” – James Anderson, CEO
Avant FPGAs
We recognized that Lattice at present has solely 199 mid-range FPGAs, which is considerably decrease than opponents AMD, Microchip (MCHP), and Intel (INTC). Amongst these firms, Intel leads the way in which with a complete of three,573 mid-range FPGAs. Nonetheless, from its earnings briefing, administration introduced its enlargement within the mid-power FPGAs with the launch of Avant-G and Avant-X lately, following Avant-E. Moreover, Lattice expects the Avant portfolio to develop to between 15% to twenty% of its whole income in 3 to 4 years. From its Investor Presentation, administration claimed Avant gives 2.5x decrease energy consumption and 2x sooner efficiency than opponents.
Lattice
Therefore, we in contrast Lattice’s newest mid-range FPGA merchandise, Avant-G and Avant-X, with comparable mid-range FPGAs from opponents to find out their efficiency competitiveness based mostly on their reminiscence charges, reminiscence kind, transceivers knowledge charges, and course of node.
Comparability of Mid-Vary FPGA
Avant-G (Lattice)
Avant-X (Lattice)
Spartan UltraScale+ (AMD)
PolarFire (Microchip)
Agilex 5 D-series (Intel)
Agilex 5 E-series (Intel)
Reminiscence Charges (RAM)
2.4 Gbps
2.1-2.4 Gbps
2.4-4.3 Gbps
1.6 Gbps
3.2-4.3Gbps
2.4-3.7Gbps
Reminiscence Sort (RAM)
LPDDR4/DDR4
DDR5/LPDDR 4/DDR4
LPDDR5/DDR4/LPDDR4
DDR4/DDR3/LPDDR3
DDR5/ LPDDR5/DDR4/LPDDR4
DDR5/ LPDDR5/DDR4/LPDDR4
Transceivers Knowledge Charges
10Gbps
25Gbps
16.3 Gbps
12.7 Gbps
28 Gbps
28Gbps
Course of
16nm
16nm
16nm
28nm
10nm
10nm
Click on to enlarge
Supply: Firm Knowledge, Khaveen Investments
Primarily based on the desk, the 2 Avant FPGAs have a slower reminiscence pace (as much as 2.4 Gbps) in comparison with AMD (2.4 – 4.3 Gbps) and Intel (as much as 4.3 Gbps), sooner reminiscence permits for faster knowledge entry and thus decreased latency. As well as, the reminiscence kind might have an effect on the ability consumption and bandwidth of the FPGAs, and probably the most superior varieties available in the market are DDR5 and LPDDR5. Therefore, Intel’s Agilex 5 sequence leads on this side, adopted by AMD and Lattice. In relation to transceivers’ knowledge charges, Intel’s Agilex 5 sequence is the perfect with 28 Gbps. Excessive transceiver charges enable for sooner knowledge change and bandwidth-intensive purposes. Lattice’s Avant-X has a barely decrease charge of 25 Gbps, however the firm claims it additionally affords “energy effectivity, superior connectivity, and optimized compute”. Different FPGAs’ knowledge charges are considerably decrease at a variety from 10 Gbps to 16.3 Gbps. Lastly, when it comes to course of node know-how used, Intel’s FPGAs are extra superior (10nm). Extra superior course of know-how can supply elevated logic density, decreased energy utilization and enhanced efficiency. Then again, Microchip falls behind compared to Lattice, AMD and Intel. We then ranked every of the elements and calculated the common rating of all elements.
Comparability of Mid-Vary FPGA
Avant-G (Lattice)
Avant-X (Lattice)
Spartan UltraScale+ (AMD)
PolarFire (Microchip)
Agilex 5 D-series (Intel)
Agilex 5 E-series (Intel)
Reminiscence Charges (RAM)
5
4
2
6
1
3
Reminiscence Sort (RAM)
5
3
4
6
1
1
Transceivers Knowledge Charges
6
3
4
5
1
1
Course of
3
3
3
6
1
1
Common
4.75
3.25
3.25
5.75
1
1.5
Click on to enlarge
Supply: Khaveen Investments
In response to the desk above, Intel’s Agilex 5 D-series has the best rating in all facets, main it to be ranked in first place, adopted by Intel’s, Agilex 5 E-series. Avant-X and Spartan Ultrascale+ even have the identical rank, however Avant X has the next reminiscence pace whereas Spartan has higher reminiscence kind and better transceivers knowledge charges. Avant-G ranks fourth and Microchip’s PolarFire ranks final. Due to this fact, we imagine that despite the fact that Intel’s FPGAs edge out different opponents, Lattice’s FPGAs nonetheless have some aggressive benefits in comparison with Microchip, contemplating the corporate has simply expanded its portfolio to mid-range FPGAs. General, we decided our efficiency comparability outcomes contradict the administration’s declare from its presentation, and we as a substitute imagine Intel remains to be the superior firm within the mid-range section.
Outlook
General, we examined the low to mid-range FPGAs of Lattice and its opponents. The corporate dominates with the biggest product breadth in low-range FPGA with nearly 3,500 merchandise. Moreover, the corporate launched a number of new merchandise on this vary which additional solidifies its product breadth. Nonetheless, the corporate has not stood out as a mid-range FPGA producer as a consequence of its low variety of mid-range FPGAs as a brand new entrant competing in opposition to bigger gamers like AMD, Intel and Microchip on this section. In our earlier evaluation of Lattice, we additionally highlighted that AMD and Intel stay the highest FPGA producers with 53% and 26.7% market share in 2022. Lattice’s newest mid-range FPGA, Avant-G and Avant-X, lack competitiveness because it lags behind Intel’s and AMD’s mid-range FPGAs. Nonetheless, we imagine the corporate’s launch of Avant establishes its entry into the mid-range FPGA market and the roadmap beneath exhibits the deliberate improvement of future Avant FPGAs which might result in the corporate enhancing a few of its product specs missing akin to reminiscence charges, reminiscence kind, transceiver knowledge charges and course of based mostly on our FPGA spec comparability evaluation above.
Lattice
AI Affect to FPGA Demand
Subsequent, we examined how AI advantages Lattice’s FPGAs. Moreover, as administration highlighted from its earnings briefing that AI-related income was round $100 mln in 2023, accounting for about 14% of Lattice’s whole income, we additionally seemed into what segments the corporate derives its AI-related income. In response to the corporate’s CEO from its earnings briefing, Lattice estimated AI-related income from a variety of AI-related purposes together with:
“AI-optimized servers within the knowledge middle…Lattice gadgets are used within the management, administration, and safety of the AI computing system” “AI-enabled PCs, the place Lattice options are used to run the AI inference algorithm that gives options akin to consumer presence and gaze detection in PC methods just like the Lenovo ThinkPad” “AI-enabled automotive ADAS methods, the place Lattice options are used to mixture and pre-process important knowledge that’s used for AI processing”.
In response to Informa Tech, FPGAs can function {hardware} accelerators in knowledge facilities, and they’re particularly useful for AI inferencing duties as they’ve larger throughput and decrease latency in comparison with common processors or GPUs. As well as, Lattice highlighted that its FPGAs have a 2.5x decrease energy consumption than different FPGAs. In our earlier evaluation of Nvidia (NVDA), we derived a forecasted common progress charge for knowledge middle chip progress of 39%, pushed by demand from cloud service suppliers ramping up infrastructure. Moreover, the corporate lately introduced a partnership with Nvidia to make the most of its power-efficient FPGAs to facilitate sensor fusion and bridging in Nvidia’s Jetson Orin and IGX Orin platforms, which fall below Nvidia’s Knowledge Middle section. In response to Nvidia, Jetson Orin is designed to speed up energy-efficient autonomous methods, akin to robotics, imaginative and prescient AI, and edge purposes, whereas the IGX Orin is right for AI-powered industrial and medical purposes.
Apart from that, Lattice has additionally developed SensAI, a software program stack that makes use of its FPGAs for inferencing, permitting prospects to “develop and deploy FPGA-based Machine Studying / Synthetic Intelligence options”. In PCs, Lattice’s SensAI can assist lengthen battery life by 28% or detect consumer presence and a focus utilizing its Face Framing Expertise. Furthermore, Lattice claimed that its software program also can improve safety and optimize methods in PCs. In response to Counterpoint Analysis, the worldwide AI PCs market is anticipated to develop at a CAGR of fifty% reflecting the sturdy demand for AI PCs’ superior safety and elevated effectivity.
Lattice additionally has mVision that helps deploy options akin to “machine imaginative and prescient, robotics, ADAS, video surveillance, and drones”. Moreover, Lattice has highlighted that its FPGAs are additionally utilized in automotive ADAS and Infotainment methods by enabling parallel sensor processing, low-power interfaces and “hardware-based safety” capabilities. For instance, Lattice’s FPGAs are additionally used to assist ADAS “facet radar purposes” for Mazda’s (OTCPK:MZDAY) CX-60 and CX-90 SUV. We highlighted a robust demand for ADAS purposes with a market CAGR of 18.7%.
Outlook
Finish Market
Forecast CAGR
Knowledge Middle
39%
AI-enabled PC
50%
ADAS
18.7%
Common
35.9%
Click on to enlarge
Supply: Counterpoint Analysis, Acumen Analysis, Khaveen Investments
The desk above consists of the market CAGR of AI purposes together with knowledge facilities, AI-enabled PCs, and the ADAS market. General, we calculated the common CAGR to be 35.9%. As well as, the corporate administration additionally expects larger AI-related income within the subsequent few years as highlighted by the quote beneath.
We anticipate our AI-related income to greater than double over the subsequent few years based mostly on the rising pipeline of AI-related design wins. – James Anderson, CEO.
Therefore, from the administration steering of Lattice’s AI-related income to be $200 mln, we assumed the corporate to attain this within the subsequent three years and accordingly calculated the CAGR to be 26%. In contrast with our derived common market CAGR of 35.9%, it surpasses administration’s forecast and due to this fact we imagine administration’s steering of $200 mln in AI-related income could possibly be cheap.
Finish Markets Development Outlook for FPGAs
Lastly, we examined the tip market outlook for FPGAs and subsequently forecasted the corporate’s income. Lattice’s Avant and Nexus are low and mid-range FPGAs, that concentrate on purposes in Communications, Computing, Industrial & Automotive.
Lattice
Communications and Computing
In response to Lattice, the corporate gives options for “computing methods akin to servers and shopper gadgets, 5G wi-fi infrastructure, switches, routers, and different associated purposes” on this section. For Computing, FPGAs can speed up knowledge facilities and their computing capabilities by releasing up processor assets and lowering energy consumption. In response to Intel, FPGAs can carry out higher than GPUs when “the applying calls for low latency and low batch sizes”. In our earlier evaluation of Microsoft (MSFT), we estimated that the cloud market will likely be rising by a ahead common of 23% pushed by knowledge quantity progress. We imagine knowledge middle demand to be a superb gauge of FPGA demand in Computing, as FPGAs are utilized in knowledge middle servers to dump “compute-intensive duties from the CPU and GPU” as a consequence of their low latency and power effectivity. For Communications, given their low latency and the flexibility for parallel processing, FPGAs are appropriate for processing large knowledge stream required within the 5G RAN. We highlighted in our earlier evaluation of Qualcomm that the penetration charge of 5G gadgets is projected to “attain 83% by 2027 from 62% in 2023”, boding nicely for the demand for 5G providers.
Industrial and Automotive
For this section, the corporate affords options to “industrial Web of Issues (“IoT”) and “Trade 4.0”, machine imaginative and prescient, robotics, manufacturing unit automation, superior driver help methods (“ADAS”), and automotive infotainment.”
For Industrial, from our earlier evaluation of ADI (ADI), we highlighted sturdy progress within the semicon industrial market with a CAGR of seven.21%, pushed by good grid deployments and digitalization of factories. Intel’s case research explains how FPGAs might allow the transformation of good factories with their parallel computing capabilities. Furthermore, Lattice’s Automate 3.0 answer stack can be used to speed up good automation system improvement, together with “robotics, embedded real-time networking, predictive upkeep, useful security (FuSa) and safety”.
Lattice
For Automotive, the ADAS market is anticipated to develop at a CAGR of 18.7%. Furthermore, McKinsey expects the automotive software program market to develop at a CAGR of 9.4%, pushed by Software program-defined Automobiles as shoppers demand safer and extra superior automobiles. FPGAs might assist automobile producers enhance numerous options and allow SDVs, akin to zonal architectures, energy effectivity, real-time networking, and infotainment interfaces. Lattice claimed that its FPGA affords 2x sooner connectivity pace and 5x extra unbiased sensor interfaces. The chart beneath summarizes how Lattice’s software program is utilized in numerous automotive purposes.
Lattice
Client
Within the Client section, the corporate’s merchandise are utilized in purposes together with “good dwelling gadgets, prosumer gadgets, sound bars, high-end projectors, Augmented Actuality (“AR”) / Digital Actuality (“VR”), and wearables”, FPGAs can speed up processing energy, allow real-time sensor fusion and object recognition for AR purposes, and optimize energy consumption. We highlighted that the AR/VR market CAGR is 42.24% in our earlier evaluation of Apple (AAPL), pushed by demand in shopper and enterprise purposes.
Outlook
Phase Income ($’000s)
Our Earlier Forecast (2023)
Precise Income (2023)
Distinction
Communication and Computing
326,092
257,536
-68,556
Development % (YoY)
18.7%
-6.3%
-25.0%
Industrial and Automotive
397,793
433,482
35,689
Development % (YoY)
24.5%
35.7%
11.2%
Client
49,064
46,136
-2,928
Development % (YoY)
0.0%
-6.0%
-6.0%
Licensing and Companies
15,730
-15,730
Development % (YoY)
-8.2%
8.2%
Complete Income
788,678
737,154
-51,524
Development % (YoY)
19.4%
11.6%
-7.8%
Click on to enlarge
Supply: Firm Knowledge, Khaveen Investments
In our earlier protection, we projected the corporate’s 2023 income based mostly on its historic 5-year common progress charge. Our forecast was larger in comparison with its precise 2023 income progress of 11.6%, attributed primarily to the unanticipated market headwinds highlighted by administration.
Finish Market Development
Market CAGR
Lattice’s 5Y Common
Computing and Communication
14.6%
16.6%
Industrial and Automotive
12.96%
23.7%
Client
6.60%
-9.3% (4Y common)
Click on to enlarge
Supply: Grand View Analysis, Khaveen Investments
Evaluating the market CAGR and the corporate’s 5-year historic common income progress, we noticed that these two are barely consistent with the Computing and Communication section (14.6% and 16.6%). Nonetheless, for the Industrial & Automotive and Client section, the market CAGR differs considerably from the corporate’s historic common progress. Due to this fact, we imagine that basing our forecast on its historic progress is extra in line than utilizing market CAGR. Nonetheless, we additionally took under consideration the administration steering for the corporate in 2024 the place administration expects Q1 income to be between $130 mln to $150 mln with Q2 surprising to rise considerably however H2 revenues to be larger than H1, reflecting enhancing market circumstances and its new Nexus and Avant product launches in that interval.
Lattice Income Forecasts ($’000s)
2023
2024F
2025F
2026F
Communications and Computing
257,536
236,241
264,274
305,582
Development % (YoY)
-6.3%
-8.3%
16.6%
15.6%
Industrial and Automotive
433,482
392,581
472,879
580,030
Development % (YoY)
35.7%
-9.4%
23.7%
22.7%
Client
46,136
47,771
41,740
37,854
Development % (YoY)
-6.0%
3.5%
-9.3%
-9.3%
Complete Income
737,154
676,592
778,893
923,466
Development % (YoY)
11.6%
-8.2%
18.9%
18.6%
Click on to enlarge
Supply: Firm Knowledge, Khaveen Investments
General, we forecasted the corporate’s whole income to say no by 8.2% in 2024 based mostly on administration steering reflecting the anticipated market headwinds highlighted within the first half and the potential restoration within the second half. In 2025 and past, we forecasted the income progress for every section based mostly on the up to date 5-year historic progress charge. Significantly for the Client section, we used 4-year previous progress as the corporate restructured this section in 2019 by winding down its smartphone merchandise.
Danger: Competitors in FPGAs
AMD (Xilinx) lately introduced the event of low-cost Spartan UltraScale+ FPGAs highlighting their excessive I/O-to-logic-cell ratios and superior security measures, together with “NIST-approved post-quantum safety” and tamper-resistant circuits. This improvement is prone to pose challenges to Lattice’s low to mid-range FPGAs. As AMD is the main participant within the FPGA market by market share, we imagine Lattice could face elevated competitors, prompting them to innovate additional to take care of their aggressive edge within the low-cost FPGA section. Moreover, we imagine that Renesas, a brand new entrant into the FPGA market, could additional result in intensifying competitors.
Verdict
Khaveen Investments
All in all, regardless of anticipated market progress headwinds in 2024, our evaluation signifies a robust potential for Lattice to maintain progress by increasing its product portfolio in low- and mid-power FPGAs. We anticipate important strides from the corporate’s enlargement into the mid-range FPGA market. Nonetheless, we additionally take word that whereas Lattice’s mid-range FPGAs outperform Microchip’s choices, they nonetheless lag behind trade leaders like AMD and Intel, although the corporate’s roadmaps indicated future product developments which might see enhancements in product efficiency to higher compete in opposition to main gamers. Moreover, we imagine that Lattice may benefit from the sturdy progress in AI-related purposes akin to knowledge facilities, AI-enabled PCs, and ADAS methods as our market evaluation confirmed a median CAGR of 35.9%, surpassing the administration’s forecasts. Delving deeper into end-market prospects for FPGAs, we recognized key progress drivers throughout numerous finish markets and the way Lattice is capitalizing on these drivers. General, we imagine that Lattice is poised to capitalize on the rising demand for AI purposes, additional strengthening the corporate’s place within the FPGA market.
Primarily based on a reduction charge of 13.4% (firm’s WACC), and terminal worth based mostly on its 5-year common EV/EBITDA of 49.77x, our mannequin exhibits an upside of 25.95%. As well as, our up to date mannequin exhibits the next upside as we beforehand used the trade’s EV/EBITDA ratio, which was solely 20.13x. We up to date the ratio as the corporate’s historic EV/EBITDA was considerably larger than the trade common (27.78x). Primarily based on our up to date DCF, we obtained a worth goal of $99.57 and improve the corporate to a Purchase ranking.
