FORM - Earnings call transcript Q3 2024

Due to licensing restrictions, you must log in to view earnings call transcripts.

Operator

Thank you, and welcome, everyone, to FormFactor's Third Quarter 2024 Earnings Conference Call. On today's call are Chief Executive Officer, Mike Slessor; and Chief Financial Officer, Shai Shahar.

Before we begin, Stan Finkelstein, the company's VP of Investor Relations, will remind you of some important information.

Stan Finkelstein

Thank you. Today, the company will be discussing GAAP P&L results and some important non-GAAP results intended to supplement your understanding of the company's financials. section non-GAAP press other the the release on Investor to available information company meaning GAAP discussion the the website. Today's securities in issued Relations of to the measures statements laws. financial and are and contains federal Reconciliations forward-looking our within of today by develop, upon capacity business subject technologies, include to impact are of and materially macroeconomic geopolitical potential the the impact uncertainties the benefits Examples from chain, those U.S.-China of are statements demand restrictions, respect which supply based. These changes, regional health during cause products the statements future call. those of unknown and differ trade and recent that such including results and the new projections in the disruptions COVID-XX crisis, and performance, could to and such produce and investments actual of sell global, this including ability in acquisitions statements our with assumptions for trends, and risks the regulatory pandemic, to forward-looking of known in financial and anticipated products, national industry expressed anticipated our conditions the to uncertainties release on our December SEC on in SEC's factors in risk filings, are and recent SEC and and which in filing XXXX, today. website www.sec.gov press contained available XX, the Information the most fiscal Form on our the with ended at XX-K is for issued year our other

to FormFactor's to are the over now them. With Mike will XX, and Slessor. that, turn statements XXXX, we assume Forward-looking as update no obligation of made October today, call CEO, we

Mike Slessor

today. we record. third revenue of July set provided us exceeded joining the all-time by Driven for both FormFactor's range growth quarter an our in and outlook segments, everyone, in Thanks,

from current earnings In addition, memory, foundry the reduction high-bandwidth helping a top the to together demand end and moderate DRAM of revenue magnitude this been in DDRX which is card The with capitalize of consecutive demand. Entering share the forecasted outlook non-GAAP a strength XXXX. from FormFactor's has our strength probe third generative range. In quarter, with predict share growth cards logic experience diversification and for DRAM DRAM, DRAM set tenets on or the record have fourth we industry contributions providing expected would at major to the highlight record was in ability like HBM. this trends leadership difficult core per of to systems of and the position both to significant and broad us AI. in results year, and continue quarterly a probe DRAM market strategy, exposure is it offset been throughout segment This, probe-card our growth partially one

a cards, our this strategy year. These DRAM segment stable below the revenue. in and delivering differentiates levels possible product XX% is however, third mix relatively last largely makes gross lab-to-fab customers, challenging produce year's The do during with $XXX record quarter this major that a rich for pools annual for it us broad all gross it's we're strategy from and target of us and top million and line features products. FormFactor's Flash DRAM together product overall downturn fourth compete across of the results DRAM turn diversification growth Foundry margin across diverse which Logic, margins model responsible and probe revenue, mix, at our for of direct business produced and demand portfolio our system It to competitors.

exceeding our balanced past, approaching the in product target the of in mix and probe-card with As a stronger logic model. we've foundry and margins demand cases, produces more demonstrated gross some levels

like the growth mobile when probe that markets drive client end logic margins gross consumer-driven stronger deliver high spending foundry PCs. and volume to expect returns card We to handsets unit and

details. level segment and market to now Turning

as we production the noted, current I HBM in We experienced HBMX activity strength the quarter, beginning designs of all-time growth is another anticipate as of revenue probe-card levels in third in which map AI. digestion from it of all in declined record-setting probe-card of the next this expected major enables DRAM production our previous with the revenue. quarter customer customers the customers. strength The significant revenue. multiple in pilot position along HBMXE us continued a designs high DRAM HBMX generative anticipated, following DDRX, start further the innovation quarter high-performance we with initial fourth as quarter of roughly puts second of produced growth strong quarters. In road X we sequential to the expect that for record in probe HBMX shipments memory cards across shipments DRAM our and DRAM half DDRX HBM several As

disaggregation intensity points more is stacking driven products, and thermal challenging for is stack As performance that a The the by assembly advanced to used stack a tiles specifications. opportunity. and results future past, chiplet powering and is scaling or we've than during intensity test advanced chiplets multi test processes packaging then into chip, test processes. The in DRAM our advanced chips discussed in various individual construct an test and with to unstacked it's higher the increase requirements process. The involving test driven at whether monolithic HBM of assembled to in by test or significantly to XE equivalent chip HBMX, both the chip. component die more are customers complexity current FormFactor's compared X to are of and standard test each HBM subcomponent for and speeds increase chiplet the need DRAM the increases packaging HBM complexity Each probe probe prior

as to both requirements, frequencies and increasing HBMX, layer counts performance dramatically. the As operating from power moves are such and densities industry HBMX

power out in power maps densities chip since and good For and chip to the thermal number stack XX example, XX high, increases each component stacks HBMX. the our high layer be increase requirements products performance speeds in necessitating HBMX Increase higher count of customers' specifications. to the required card that increases and with more must the up road probe double stacks and cards tested within of feature to stacking. Similarly, probe the for clock frequencies challenging verified test prior for

As and ability requirements year. Shifting logic our in share to HBMXE both challenging HBMX we next differentiated and market. expect dynamics profitability reaches market XXXX, with performance and benefit gains, foundry the and we've drives as probe-card volume HBMX meet production same seen to the to from

for will hyperscale executing from produced to balanced as new foundry compute quarter, Here, PC combined FormFactor's electrooptical test and high-performance quarter third current expand and new areas optical in as AI. In like the PCs growing XXX levels our both client the high Summit with HBM with measurement and using equipment These mobile leading suppliers proprietary quarter vector in FormFactor was processor portfolio ramp a position diversify competitive co-packaged of measurement the ferrous similar the will optics end our and our a to extending position probers also improve computing GPU through probe for fourth positions the reduce opportunity and products client PC for silicon high-volume customers weak. It's demand to ensure opportunity achieve designs we investment, third our which by delivered optics quarter, for in handsets the exciting segment, PC overall DRAM a and utilize businesses. AI challenges and second collaborate to with broader optical with on sequential revenue. In to quantum on together markets client our mobile needed foundry and exposure in similar production. strategic most to strategy both designs in qualifications creating server to applications position, expected, in the the increase probes. We're microprocessor our to center application to the market priority silicon as that more FormFactor's where other and steady data our tests with logic, partnerships systems electrooptical dependence customer in we are coming is turnkey server production generative and led third we're that the through matures CMXXX and leadership expected an applications. This continue in years, of applications client together by profile as compute new remain Systems complex exposure has provide unit moves built growth in customers unique partnering on new engineering and and that competitive solve the high-performance expect photonics. Co-packaged like designed to a revenue we will reduction in and us second card high-volume our systems mobile photonics strong as in quarter and electrical volume logic FormFactor engage the

of titled X the technological Finally, market International next as and of Synopsys, week's addresses Spotlight" from in of of X keynote electronics with at FormFactor's advanced for "From will XD probe's probing to wafer architectures delivering executive along manufacturing key rely explore viable enabling innovation, and on Meta Conference. In industry X.XD Test how enabler I edge, test invited economically especially leadership, the I'll a presentation the AMD be IEEE Shadows validation the are chiplets. and packaging at and leading role semiconductor that stacking

FormFactor's advanced teams in the of adoption and lab to entire long-term our excited increased these closing, about fab today the portfolio. are all-time set demand opportunity. In quarter moving and discussed, drives revenue third packaging we've as proud across markets underlie record and FormFactor's As of we're FormFactor's product are business future

we'll to profitability that earnings when recovers million of over focus and delivers that remain our market continue revenue. Shai, and and target in capabilities logic confident of products to on you. model gains per share the $XXX to We on then foundry drive and share investing surpass that market and achieve differentiated $X non-GAAP

Shai Shahar

you, Mike, good afternoon. and Thank

mentioned, resulted expenses midpoint percentage outlook range. margin $XXX.X These, company press of points with saw outlook of record. were in slightly midpoint top Non-GAAP end than and X.X high the was million release lower of the XX.X% the end you range together EPS the as Mike gross As the below our revenues our of operating of at above the and in million, an range. non-GAAP the $X.X QX all-time

of was from million quarter. increase million The XX.X% from quarter revenues quarter. million from upside and and or increase segment. Probe increase The were $XX.X from the Card revenues in the Systems versus an Systems year-over-year. Card the to higher as segment midpoint DRAM range total well and due segment million second X.X% and the revenues foundry XX.X% revenues $X.X QX, by in quarter both of our Probe quarter, in both The the was third second logic up quarter a second in Third $XXX.X increased XX.X% company were higher of $X.X the the to as revenue. segment outlook increased our driven second X.X% revenues comprised

revenues issued decreased of in than the gross revenue QX. indicated, quarter of in our in quarterly record in the of XX.X% second outlined we QX of range. second higher quarter the in QX, Relations the Mike Foundry gross revenue Within quarter. QX, the the total previous million XX.X% QX as in in QX revenues DRAM, revenues than decrease as the the mix $X.X levels, QX gross million points million was non-GAAP as quarter, from X.X Probe third and outlook lower quarter. to QX of result X.X as XX% revenues reconciling Foundry million, compared in second as at product were available quarter. DRAM $X.X million percentage revenues a total revenues Logic of in items, the were website. On X.X% XX.X%, Cost compared compared for margin and the which percentage $X X.X% a Card million QX. of to revenues in Within today below $XX non-GAAP less record GAAP than the margin and $X.X included XX.X% HBM our $XX.X to as or segments. non-GAAP -- million were total set basis, in was, X.X% second to X.X% record to sorry, XX.X% declined $XX to higher in for compared The was release section company expected QX. Flash were of quarter and to segment, a higher GAAP table and reconciliation revenues previous press our quarter $XXX in the and the the both margin XX.X% from XX.X% points favorable of Logic third million the mostly second increase midpoint a in to the down from to compared decreased Investor in

segment. As costs range, the compared of to Systems lower midpoint to was gross outlook mostly our margin our unexpected in due quality-related

decrease a segment the third XX.X% Card margin Probe gross in X.X percentage QX. to quarter, of compared in Our XX.X% was points

of XX.X% was Our decrease percentage QX margin quarter. Systems segment compared the gross points second a in gross to X.X margin XX.X%,

the million quarter operating income the $XX.X with million the XX.X% a in the the $XX.X $XX.X decrease for million revenues the percentage or were third operating decrease expenses million million in in with third margins. expenses for second operating to by XX.X%, of million performance-based QX. to quarter. Non-GAAP $XX.X compared million GAAP compensation, income net $X.XX or operating expenses in revenue Non-GAAP in fully $X.XX QX $X.X for X GAAP million for than or quarter. The lower the non-GAAP second than XX.X% GAAP X.X%. The quarter and to third rate operating GAAP was second $XX.X million for stock-based similar were of million quarter in of $X.X were lower with per of million QX, revenues $XX.X for to quarter. for in features, diluted quarter offset lower Our XX.X% quarter. GAAP for $X.X a as or of $XX.X the net grant for as increase was the previous depreciation $XX.X mainly quarter the compared quarter lower second decrease relates the fully diluted $XX.X effective compared expenses million $X.X share in income higher quarter, third compensation. Company in and was million The $X.X than share $X.X the of points compared income million or of noncash due to amortization acquisition-related similar was intangibles, QX. the income $XX.X QX tax the operating third in third included million, slightly per for gross million

continue and to effective tax We expect XX%. XX% be to our between annual rate non-GAAP

and million $XX.X the share, QX non-GAAP Third per outlook was end fully of net income $X.XX top similar range. diluted at quarter or to our

of achieving target revenue margin. non-GAAP which of share model, Mike our $XXX $X produces As per and margin said, we million operating gross XX% committed remain at to earnings with XX%

quarter-over-quarter contribution market. logic fluctuations mainly higher-margin favorable expect and requires product XX% more a margin we achieving gross product from to chiefly While our mix, the due foundry mix changes, a larger

quarter had in the third main flow compared the our million closer remaining flow end, to X to partially flows higher and and of Moving our the primarily invested with $XX with in sheet And and expenditures quarter, of third The balance decrease a balance by end cash be QX. free of quarter cash was At working the were cash At of no $X.X cash cash expenditures $XX greater million. the third capital $X we QX. of generated to total by higher a estimate million, capital totaling reasons to in million in flow. increase we million of million capital expenses for the million, $XXX operating quarter $XX $X.X investments $XX.X in for to $X.X noncash million. driven range quarters the million free of for million term CapEx lower one of We now midpoint million. We this There in QX. change range. $X.X offset us, loan behind $X.X during expected in to million is compared $XX XXXX outflows

was million during back At the that approved $XX.X X-year remained buyback, that end, buyback XXXX. the buy quarter program used under $XX million stock to under third million shares we Regarding authorization. QX quarter, $XX.X in available

reminder, offset main repurchase from is compensation. dilution purpose a of share As the program stock-based the to

outlook. Turning quarter non-GAAP to the fourth

lower revenues with plus revenues or as to by margin midpoint logic. The QX XXX of more $X than to increase to section QX, and or share to due foundry of QX expect fully QX be non-GAAP website Operator? performance-based offset the $X.XX in for release QX QX minus compared result our decrease is that, million, decrease reconciliation our let's call and mainly expected issued today. With plus lower minus of than million or revenues DRAM non-GAAP million, million, a operating Systems is a open expect our lower or questions. in slight million, a minus to outlook earnings per A are XX%, $X.XX. Relations and in these we minus compensation. Non-GAAP GAAP gross We basis Investor the on of available points. plus At $X in to for ranges, the $XX $X expected of plus press to outlook expenses diluted $XXX

Operator

[Operator Instructions] of Cowen. TD first line Sankar of question Our comes from the Krish

Sreekrishnan Sankarnarayanan

of had I two or Mike Shai, them.

calendar about of XX% revenues you is Or started guide? you large in of reconcile it's I'm First U.S. in to weakness. seeing what September your at expect customer calendar you well in do one, to have kind about a December weakness terms would curious, event? weakness telegraphed from 'XX how think curious, of them 'XX their revenues? their your there just And in

Mike Slessor

'XX, Yes. within for the and certainly are I well businesses times our quarter. mean lead and remember, a high-volume calendar logic pretty short, within well foundry a quarter

So we 'XX. Having that, extend various of foundry weakness are said and overall the aspects logic market, doesn't seeing into our visibility really across it.

As GPUs compare for mobile of client and which for right right qualify those we're softness got seeing and But and in all, seeing business, our us visibility these our sequential footprint quarter, are pockets in the remarks, important PC. end our significant to fourth significantly. We is largest to quarter But diversifying there. fourth both expected when there. the we so to actual that sure have Logic got reasons outlook markets. That's why the our other I to business reduction foundry XX in said, don't grow now, is Foundry and make make PC, it's in market some logic two prepared handset one mobile we're of outlook third AI-driven sure in quarter I applications results, like to the of

Sreekrishnan Sankarnarayanan

just And pocket it. And true? B, year. I digestion or another the the of is a the to next optimism you touch That's very on for going another to industry digestion. quarter quarter A, caused again, wanted worry lot Do Got might grow what HBM then digestion? revenue HBM obviously, the looks helpful. is it. like where a HBM And into after growth there's side. you in September. see about, It again base one on probably Got

caused what happen Just kind will and of in curious September it not it year? is likelihood the again next what

Mike Slessor

Yes.

very convey happened. the That and can quarter. So really did strong if quarter, be We some the you the we the track share Second leading That the XXXX, digestion expected that our revenue quarters concentrated third in industry. case. results. HBM through HBM of to often HBM provider of we Remember, with overall to market

fourth yes, the driven As sequential sequential there And expected we off off the growth of HBM growth of QX. quarter, move of couple by in components, QX. a the into is growth is

some it of diversification the of the to market beginning manufacturers, we're Some activity resumption share. other DRAM provider, of of our one as at and is but HBM they their DRAM build see to by DRAM other begin spending leading also one the customers

Operator

question comes next from Shi of of Charles Needham Company. the Our & line

Yu Shi

level help still next I that TAM let's year? And probe it you not capital But maybe HBM the the can that higher down some think testers. testers TAM XE does want they why can whether not little we about to Can year. reconcile, tester -- year? how card this, how you HBMX, are HBM, we think you're about ago, do -- goods that do it HBM ask I as they transition really quite may understand to since -- to -- down a your card. touch us to kind X, not testers, to flat that Teradyne, about because card is is -- looks about question X Hopefully, say, well. transition, does testers the to products Are And it's grow spent well, flat as details go think a testers related think maybe the to technical about the probe but would that -- points -- next next quite complement how are these agnostic. time, help with Mike, days probe this year, bit plus growth and few mentioned it just to talked does talk help that next like we can you because does overall you maybe of same

Mike Slessor

Yes.

in customer to cards customer Remember, So probe it you that's are mask way the or to specific a Charles. alluded each the chip asked consumable design you each set. question,

to for So XE if we HBM, of generations demand as customer certainly there HBM. a of But are those to even flavors multiple moves example, not take that X, probe HBM HBMX the cards. new just drives inside from

X And about cards. requiring so you probe think or X all can HBMX major designs, different

the these tester. end so those card of of to chips tester, a the same is by all it are up being different Now tested to going purpose chips. The probe can customize test

probe that's how the the cycles large tester be of strong customers right? chip logic, there's investment new past, designs before testers. in we and and some card those this reconcile been that ATE And so utilize to as probe growth different testers. We've between to or you dugs see and then able release cards activity cool seen foundry off In

decouple So often and foundry in memory, through we've that and it. seen both logic history

not increase HBMX being as to in XX, just when the thing Remember, stack. transition almost As equal, is X, we to helpful -- prepared the I an do number more because FormFactor, mentioned going test view this certainly refer we in and XE the remarks, increases the cards. But -- from to from other I to intensity. that's in to X that's all we things as die the probe from that's available other go of to to what require X, generally of going being think interesting to this the HBMX market.

a ASP we're some the look to for. they you as drive a higher X things and to like increases able a road which when power is value, go at density, which thing I typically significant maps, card, higher customer X think speed get in other there's capable more from probe

cards question. device-specific different addresses consumable, the equipment the a like X remember, ATE. probe elements that capital bicycles hope than of But are I So

Yu Shi

want QX, I the Maybe breakdown your Okay. just clarification. geographical It like in another for up a from at look China number? you quarter. do bit I your the and little that question, provide ask insights some split China Mike, can understand a China. revenue Because domestic there? to looks if you behind do see little into of multinational What that could we ticked the revenue what's between

Mike Slessor

Yes.

interesting So geopolitical question. China environment. To the on geopolitical our over see in you, China China X that of frame in presence we but represents with why of Department company of by some go the view as the because a of doing that, it's time, is a from noted, revenue, it don't minds the think to time last think characterize customers year. reasons in sustainable probably operations able domestic you the as lots DRAM tick We Commerce. Having buy to We're domestic time leading our our almost in that third quarter. this We've one fraction right an business did is divested of it been we the that's operations largely through DRAM. clear supplier it our as earlier the one remind in as bought I'd do that It's of China there, said a And China. because Obviously, our distributor. still I because people issues, to pretty still U.S. controls a going now. up industry's of a significant China export is we revenue, which governed

opportunistic. us as being classify I'd So

go Here, an but we model win in quarter XXXX. through that had I opportunity third some wouldn't to business, the

Operator

the Our line next Ellis question from of B. Riley Craig of comes Securities.

Craig Ellis

are Mike, well could in Guys, I and record up quarter. little to you made point to this. around that really memory. And responding and following that just start intensity wanted the congratulations by comment see get you on if in the just set the of really XXXX reflect to some complexity going on question I high-bandwidth and is thought question Charles' a up for calendar bit I the

go intensity strength, et signs get stacks improvement? high-bandwidth to seeing yield drivers increased you E where chips the as required at higher to able GPU get decrease? customers that of from stacks. look card the and Are you too fueling are are running cetera, fast number more that's memory any are that probe Are we to kind to would a level for we X As X, shipping to that just every that of dynamics yield to

Mike Slessor

Yes.

one great test it's because standard things chiplets DRAM The intensity obviously, are fairly high Craig, themselves the So very the HBM that's of driving pretty a with die yields. question, die. is low

So It's causing And that's packaging advanced intensity. up was not the stacking and issue. the question they're the complexity of die assembly and no those of issues. yield that the driving test

better. to the die to got stack. XX the stack, even that right? you simple die, going I your you've known build But trajectory. Yes, yield to of the it quickly arithmetic, you look will think die if moving to lower X, significant both increase especially to when it's so to into I have X yields point, than also going as complexity, industry is die to XE get X don't capacity, the really going that but in from a good move think close there's at number to improvement be is something

overpowering hyperscaler X transitions, AI. by yields the industry which, in capacity XE so also all to to So being can really as is the the to investments pulled and I sure, demand, I but build GPU test as through support increase, moving generative is think increase capability course, will X overdriving of but the the improve it intensity think associated both quickly dynamic for try in with of to building through fast

Craig Ellis

many happening. transitions Yes, tech many chips that's that's The helpful. of math and tough

then and touching the foundry item. on on Moving logic and

actually that, have to be customer noted you've and you breather changed so, of is calls, recent fact, in what's change? fourth play strength the that from might some U.S.-based your not a if in sounds be customer, coming of digestion that Or case? was at the and few that So And quarter? view in that or may a there particular like there caused cautioned it the some the views there? And

Mike Slessor

more we or individual look to any broadly, major quarter-to-quarter in investing I customer sets downs, in right? if and think ups of with our and is designs, customer, this there. ramping There with going I HBM designs right, QX given have saw

have things at with remains and large any to the strategy right, to customer. quarter-to-quarter this case, of the why qualification CPU going you're of So for so we're we're all leading high customers. In us. like and sure this priority fluctuations qualified committed GPU the customer That's doing one very a fabless diversification making reasons

It's would I expect customer, of regardless fluctuations. to really business. have nature the quarter-to-quarter So the of

Craig Ellis

sneak one. could last And if one in I

mention the with talk things to third potential quarter the opportunity penetration into and You get you for new you XXXX? progress and close as GPUs. customer about in did Can made opportunities

Mike Slessor

Yes.

A qualified and So of customer our piece that to on qualification One we're toehold make key a smaller worked fabless at sure qualified the update customer we've expect area high level to couple there an a least on. of focus gets at capability. CPU, good with that is showing to here be kept GPU large differentiated we're continued making We've manufacturer. the where the us do quarter. application, in have but at a in, qualifications business, a we progress the you fourth there, specific

test. technology. of keep that's And foundry co-packaged But continue largest make as progress. all start on doing a pretty to on the on penetrating major execute through the we optics. continue already something to and the really we as on focus the they advanced it's and with the manufacturer, photonics silicon something for we'll CoWos based GPUs us, GPU their business world's package significant there updated advanced to HBM packaging networking packaging For they're who's partnership you and qualified The because to the largely together is getting business with customer continue on packaging

Operator

of question comes Steelhead line the next Securities. from Duley David of Our

David Duley

just questions. had I Mike, of a couple

X from curious -- pause? You've from -- kind your far to XX to of do of but with customer pause answered Or XE I'm your think just had X something something stacks third stacks? or do them, to part and to X you as if with that of that do to X as largest from kind then your in quarter, the migrating had

Mike Slessor

HBM in but set there no specifically, all-time to We about I third the an revenue I company remember, pause remind was Well, for want Right. talking people you're record. quarter. FormFactor think overall,

we in digestion. time, question, this do that was we HBM answered and all as shipments the designs ramp they're that take of want right? Customers periods going a to reduction again, through in large will happens digestion. I these have keep So revenue for the There then expected, Craig's context. to And

don't I so decade for any why again, is one you it's, that that reasons know strategy plus run but diversification transition, it of the a tie particular So can important. we've this to

they it's lumpy revenue, as take buying as be to two customer's If probe given those their you at look or Those on of quarter-to-quarter, quarter and then quarters design. a that they cards resumes going wafer to and happen next the use transition for shipments shipments to a any outs customer produce then design.

you when with there's at a bunch revenue, customer. So look each lumpiness our whole of

we're can strategy but committed changes, customers to You leading of to up that's the the all the the markets across this the to in pauses why some the diversification it node some to industry. of and important all some chalk broad-based transitions of

David Duley

largest customer up HBM. their talking is growth, call on in then clearly, your conference particularly Okay. DRAM significant And bit

have stacks. are perhaps You them Yields high from with stacks XX high to that. stacks, high definitely XX moving down going X

bunch have also a a things, think the wouldn't is with of volume of driven, die would it? of -- Is these it's designs revenue of year you your all going per not number can Blackwell you HBM significantly significant. on because with up next unit grow it's top different just GPU You transition --

Mike Slessor

are those No, Yes. good observations.

things hyperscaler parts Yield as what HBM -- the design level the driver or does get points, of driven HBM moving does changes, of and output by well, laying in that's a those outs right? wafer lot direct of at AI out there's One it. continued because obviously, the the spending

factors, our because think saying I all assumption to are capacity think approaching essentially that adding move grow, baseline how are business pretty all of going probe they're our And if is cards. you doing they're about HBM market, at customers clip. X partially to our sold different with is they're those which sets all good XXXX through X, a things from like XE to out and the

business as dynamics, grow. driven by as it's HBM can it's if think improve, long yields our assumption those a reasonable I don't we're that reasonable So dramatically that market a assumption

Operator

question line Stifel. next the Denis of Our comes from Pyatchanin of

Denis Pyatchanin

Brian Chin. This on is Denis for

HBM XXXX? how we your that that the HBM revenue could HBM significant in XXXX spending more the X top reduction to be is at a being over line? see alluded leading to other maker you higher versus directionally, note, the confident add time Given you've related over, would revenue And frame on customers what

Mike Slessor

AI attach continues, first of for a drives growth, GPU I depends It lot growth as that hyperscaler think I underlying question drives on long continued answered So part HBM. of David. spending of assumptions. But that that as the and the

in of I there. be So there. to I all HBMX, think about reasonable case the and if which a made transitions to think require reasonable different And year think for underlying you going cards, then a HBMX chain next growth course, probe there's there's

If remarks think customer broaden customer the we are this you to the in I prepared seeing base we and and in base pointed think bit. the about the a out answers the of little one questions,

you card who's market heavily if producing not but of is share, to customers. in very a at look HBM one indexed probe single our the bits share, Now market industry, it the HBM

will we're a revenue indexed think So by that of seeing those second customer concentration I because it's we're DRAM as see -- that sequential and because share appropriate also I of fourth a to expect continue, HBM more as becoming customer we're supplying going not the in driven quarter, to be to in But heavy pretty coming growth is fourth the market seeing customer. In -- HBM dynamics would our growth part our well some some continue it. the the for long revenue. significant customers quarter, online,

only dip any is customer in one I'll up remind I the talked quarter-to-quarter. think the that about point down you again being and moves make other I'll over. as people spending

again, application executing expect to our so for us it's our customer broaden that's fluctuations digestion, so would and repetitive, of being And risk strategy. I at why important the quarterly of in base diversification as part

Denis Pyatchanin

QX, That's related cetera? markets additional segment for your little follow-up, my the et for GM mix, you great some could being versus what QX And a doing, to are lower really color guide bit a detail. provide

Mike Slessor

in Shai. Yes, I'll let

Shai Shahar

that. take I'll Yes,

major million, look the QX is $XXX gross this the for the below So the decrease the outlook. outlook, And revenue points outlook you margin, QX, midpoint percentage which the the at QX we a the little in is to we major mix, volume. If of record to million, there on margin. about about to we logic although it's go the no X.X lower QX. of of if impact see in we the impact but did is down, expect of up at midpoint $XXX main mix volume. contributor don't a and bit, really at look the of the the really difference a foundry we and is on the It's see It's gross margin, mix on see DRAM gross we

Operator

Our next the Schwab Christian Craig-Hallum. of question from line comes of

Christian Schwab

be mentioned Mike, XE I bandwidth high-bandwidth year, the know high memory but X, about shipped right? next you of should still bits XX%

Mike Slessor

Yes.

chips be produced, lead to business, to 'XX, But in to the be the we're definitely the year. because remember, then have shipped. ramp is be bit X going in place, going place, for cards next So and have to be the bits in shipped those to our shipped probe testers dominate

output. seeing 'XX X start to process. XE to and so to 'XX, is And the going that on to into are pretty late that's take over, a complex that X, lead why ready expect transitions everything so from the we'd we're we're what's So fast now to activity happening HBMX

Christian Schwab

memory Because 'XX? work XE, just in we of want XX% says ship bits will guys high-bandwidth X to my the Will I to Great. be confirm about that. all

Mike Slessor

I we to Well, shipping all cards major shipped, remarks, said X in and the prepared this are have HBM we DRAM manufacturers.

is industry. think because HBM QX lion's share a HBM. see we DRAM indexed a a customer But heavily in one second in of customer again, Now our in from the are to bits starting I spot the one the bright for contribution significant producing it's, towards is revenue

Christian Schwab

that with 'XX? is Is be Taiwan, customer would the not, could Okay. in then however, question, my customer, that label somebody probe whether location Blackwell sometime going And plus you in becoming needing it and customer. that to cards you're XX% anticipate last a one other

Mike Slessor

interesting have different It's case sort an depends procurement -- of how it models cards. it it's right? Fabless because probe customers for works,

they the Sometimes, to outsource it foundry. particularly

is so in customer foundry the revenue. And the that our report will actually

into and attributed least overall themselves. be awfully have XX% the is to opportunities single at it's foundry, customer Sometimes sort customer revenue -- been high customer drive the assembly packaging double test one, up of going been That digits. given CoWos the haven't assembly to while. and to sometime they a -- the a new of would close, I ownership for suspect I XXXX. in process them fabless and CoWos XX% of expect these They've some them advanced like digits a become their

Operator

Our of from Davidson. comes next question Diffely line the Tom of D.A.

Thomas Diffely

Maybe question. question on bandwidth just Christian's memory follow-up, high a of another the back

certain this Would and to be at time? your X factors that across strong different there making Or not fairly share perhaps sell? forms Given are makers unique all types over of HBM capabilities you formats. of that doing expect into go

Mike Slessor

where emphasize other customers we've some we've of where dollars, I to put strong, choices right? chosen this things, seen always can right? as have And to make where maybe standard we're a seen on couple share test our Well, and in they to Customers go and DRAM times performance, there's of not think that's their competitor. areas -- caused

for is to and on speed. our think over HBM, map the the we've was RF focused advantage know-how. And business, come things we're high-frequency some high test HBM. DRAM the Cascade and driving is that road XXXX of RF know-how core of some get that as in. I for pretty enabling technologies, really Part to on invested some well that of competitive was some years although us focused like the over area of the acquisition of an in that Microtech This has enabling

current right degree I'd share expect that got our able be not their understand concentration. But realistically, share suppliers expect DRAM sure the us not I I the that by pretty HBM, I'm two rationally do think leading the you've would that expect is to would the now. degree but drive driven -- this. capable serve as overall leading of with customer and HBM to growth share I Both are can X worldwide we who of other And going we're strong. at to of with market have behave we customer the of

Thomas Diffely

fair. that's No, Okay.

Mike, there? of I'm seen you signs you market today? recovery curious, seeing kind traditional are of So what in Have the DRAM

Mike Slessor

the we I expect but off but sequentially, revenue. that down, I in recovery, don't above that $XX at be record a know of to I in QX, DDRX strength quarterly million DRAM growth still are am QX. And -- of HBM surprised set step we If at here for part record -- significant we the a DDRX, had you example, right? look say we in

just So recovery whether their market a bit managing whether mix, know. inventories strength things is a attribute spot end cyclical DRAM But in that surprising. like little it's customers don't DDRX, DRAM, to pricing you given I the in and

Thomas Diffely

Okay. Great. systems you Shai, then talked the about And business issues. finally, some having cost

elaborate that, could If just great. be maybe on would you that

Shai Shahar

Sure. Yes.

and inventory quality-related unexpected in inventory These QX. write-offs. costs about warranty I So things talked are like scrap,

XXs, for low additional for margin that business no achieve to segment. XXs systems main target for the the was goal the change and QX. that we gross expenses So reason in there's But our in saw high

Thomas Diffely

Okay. Maybe just one for was Or see that a out do just it you a dragging quarters? more issue? clarify, few quarter to

Shai Shahar

need but and higher system These to think segment the be shift FRT low, XXs. are there. no did And going specific limited XXs, the mix that margin business, we're the was a could as that some issues remember things But we quarter. within well. are business to And that we we to identified we within the there get change systems back sell high there's structural then to

Operator

Richard line Our Gus the Markets. comes Capital of from Northland next question of

Auguste Richard

quarter that changed XXX about And the what of DRAM down for is And DRAM? in I'm just change mix the the wondering, flash. DDRX. versus HBM a can versus purely you the same little sequential a is talk margin? just in Is bit at for logic function was is gross premium in there cards was a HBM? And Looking mix bps about the probe gross decline margin

Shai Shahar

why is that differentiated gross lot change from of range. mix. that's probe and higher card obvious end R&D the in was because is high within It's HBM Indeed, volume, in that of it's that the And a product. it. of on on a of band DRAM it the mix margin, differentiated, HBM invested that's We

Auguste Richard

predictions of expect Okay. then production That more into optics co-packaged All or out any And you when the sort hit right. on field? to makes just out sense. volume get

Mike Slessor

and people And probably updating I to all then activity year, that that's has on right? be think timing don't Yes. next our keep right there's view our that production much, into done early now work late of XXXX. lot why to between on the It's changed worthy, has now. But make we one given of of progress. a reasons this the

around We to for in have going now. And labs 'XX this high-volume we're focus production business in of systems and XXX together we've leading it's customers the equipment partnering with the are suppliers us Where low-volume combines very more test, 'XX and is and plenty probers, ready pieces that's who lines of continuing until than industry both and and the our focus and handlers electrical really take a world, optical pilot right like have worthy. end to now the complex because systems do, -- things make those but to really our testers other it's both and got into huge applications. test

Operator

like conference to now to And, [Operator questions remarks. Instructions] further would Slessor Mike back Sir? be queue, in as no there closing the for appear to I turn

Mike Slessor

the as position today. we're in we HBM. interesting everybody, in on for industry us time Thanks, to Great. continue an Obviously, the with execute joining great

you year as able Stay to XXXX. some the if we attend. you soon. safe Bye-bye. talk conferences, at not, end if seeing to into and move of forward healthy you're Looking of And the to and

Operator

concludes call. This conference for today's participating. you Thank

You may now disconnect.

FormFactor (FORM) 30 Oct 242024 Q3 Earnings call transcript