Friday, January 9th 2026

New GPU Startup Bolt Graphics Promises 2.5x Path Tracing Performance of RTX 5090

Bolt Graphics is a new computer graphics hardware startup based out of Sunnyvale, California. This company wants to shake things up in the consumer graphics hardware market with a brand-new, grounds-up graphics architecture for graphics rendering, gaming, and HPC. What makes this startup unique and refreshing is that its architecture, or at least the marketing material of it, doesn't use the most uttered two-letter acronym in human history.

Called ZEUS by Bolt Graphics, the architecture offers a contemporary raster 3D graphics, ray tracing, and path tracing hardware pipeline, using an in-house designed SIMD architecture that's controlled by a RISC-V command processor that doubles up as a CPU. The prototype graphics card features a PCI-Express 5.0 x16 host interface, and LPDDR5X as its graphics memory, with additional DDR5 SODIMM slots for the RISC-V command processor. Together, the board can have up to 384 GB of memory to deal with large data sets. The card also features a 400 Gbps or 800 Gbps enterprise network interface, and a fully-fledged BMC/IPMI circuit on board, so you can manage large clusters of the card in a rackmount render server or workstation. The most surprising aspect of the card is power, and it only features an 8-pin PCIe power input, for a total of 225 W of power draw.
Besides hardware, Bolt Graphics is offering a fully-fledged software that that includes Vulkan and DirectX 12 interfaces, including support for Unreal Engine and Unity. Applications for the architecture support all modern programming environments. Bolt innovated a path tracing pipeline.

Bolt Graphics has made some astoundingly tall performance claims that indicate an ASIC-like approach towards popular use-cases, using fixed-function hardware. In FP64 EM simulation benchmarks, the Zeus 4C is claimed to offer 300 times the performance of an NVIDIA B200 "Blackwell." In another FP64 math benchmark, the Zeus 1C (single silicon) is shown offering 3 times the performance of a GeForce RTX 5090. Perhaps the most fascinating claim is path tracing performance, and here, a Zeus 1C is claimed to offer 2.5 times the path tracing performance of an RTX 5090, 5 times with 2C, and 10 times with 4C.

The part about Bolt Graphics marketing that caught our eye is their claim that the Zeus 1C is "the fastest graphics processor ever," made possible due to a fundamental reimagining of conventional graphics processing workflows.
Add your own comment

114 Comments on New GPU Startup Bolt Graphics Promises 2.5x Path Tracing Performance of RTX 5090

#76
lexluthermiester
LycanwolfenDejavu. Reminds me of the 90's when Ati and Nvidia was staggering along no big improvment in Gaming then came along a small little company called 3dfx.
Exactly what I was thinking!
Posted on Reply
#77
Solaris17
Super Dainty Moderator
btarunrthat indicate an ASIC-like approach towards popular use-cases, using fixed-function hardware.
Yeah; I mean not hard to see. You can make a lot of things pretty quick using ASICs. Just doing everything else at the same time is hard. Thats the part I will doubt.
Posted on Reply
#78
Icon Charlie
john_AMD under Su is very careful and doesn't take risks. That's why they aren't making money, why their financials didn't exploded like Nvidia's. Their big bet is with Helios and 400 series. If they find enough buyers for Helios and they will probably find, their share price will go to over $300. If they fail, their share price will go back down to $150. Intel is also a nice bet, because it is refocusing on consumer market and Xeons and consumer market and Xeons alone can drive their valuation up to $60. If 14A works and Apple or Nvidia start making products on Intel factories, their share price can hit $100.
We aready have seen the financials on this site. It is Awful actually on the NET profits as shown by the Lost Swede (again thanks for that) to say how awsome they are.

And in my case I do not care of the price of the shares of their stock. I only care about...

Dividends I get on the stocks I invested. Again I'm an old man. I do not have much time to make any catstrophic mistake. Which is why I'm into commodies. I have a net profit goal and if I make more (which I am now and it is not that high) I'm very happy.

If you are 50 something or younger you still have time to take more risks than I.

Speaking about Risks: All things considered... I guess all those AM4 people are kind of smiling now. I sure am. Wheeeee :peace::peace::peace:
Posted on Reply
#79
kimbentsen
ChomiqHere's hoping they'll actually deliver on their promises and won't get eaten by Nvidia.
Nvidia might buy them and plunder their RAM supply to use in AI products.
Posted on Reply
#80
njshah
So the writer thinks it's promising because they didn't use AI in their presentation 😂

I swear this caveman like resistance to machine learning (ai is an inaccurate term) is laughable. It's the future, get with it or ge left behind
Posted on Reply
#81
lexluthermiester
njshahI swear this caveman like resistance to machine learning (ai is an inaccurate term) is laughable.
Get outta here with that crap. There's nothing "caveman" about desiring life to be free of AI-slop.
njshahIt's the future, get with it or ge left behind
I would tell where you can go, how to get there and what you can eat when you arrive, but the mods might prefer me not to. You get the idea though.
Posted on Reply
#82
Unregistered
lexluthermiesterGet outta here with that crap. There's nothing "caveman" about desiring life to be free of AI-slop.
The "crap" is the inane labeling of all AI as "slop". Yes, nearly all current AI output is devoid of creativity. Guess what? Nearly all human "creative" output is too ... have you watched a Hollywood film in the last 30 years, or listened to any pop music? I'll bet any sum you care to wager that the long-term effect of AI on that isn't going to make it any worse, and will almost certainly improve it. I'll even give odds.

This Neanderthalic anti-AI attitude is strongest on tech-oriented sites, populated largely with people fearing (and in many cases, seeing) their jobs evaporating to AI tools. Which makes sense. Two centuries ago, the Luddites rioting and destroying the world's first computers (semi-automated punchcard Jaquard looms) were the weavers losing their jobs to them. Luckily, we managed to rein in those lunatics.

Interestingly enough, they raised the same arguments in the 1800s as are being raised today: machine-woven cloth was "slop(py)", poor quality, dehumanizing, and would eventually erode the quality of life. Yet somehow, we survived.
Posted on Edit | Reply
#83
Dragam1337
Considering how difficult it has been for intel to just catch up to nvidia, i have some serious doubts about this startup surpassing them. But it would be a most welcome surprise !
Posted on Reply
#84
Icon Charlie
lexluthermiesterGet outta here with that crap. There's nothing "caveman" about desiring life to be free of AI-slop.


I would tell where you can go, how to get there and what you can eat when you arrive, but the mods might prefer me not to. You get the idea though.
I agree with this comment. The main reason is why it (AI) is being used the following:

www.techbuzz.ai/articles/investors-predict-ai-labor-displacement-accelerates-in-2026

From my perspective: AI is being used as a method to hide actual real performance from the investors with the usual market speak given.

It's similar Bull$h!t I am seeing with real estate. The Foundations are bad but it sure looks good when you state on how you are using "tools" to enhance your company.
Posted on Reply
#85
notoperable
Is there any empirical data to back your claims @Endymio || GTFO

btw: what did I say, lol: videocardz.com/newz/popularity-of-ddr3-motherboards-is-growing-rapidly
EndymioThe "crap" is the inane labeling of all AI as "slop". Yes, nearly all current AI output is devoid of creativity. Guess what? Nearly all human "creative" output is too ... have you watched a Hollywood film in the last 30 years, or listened to any pop music? I'll bet any sum you care to wager that the long-term effect of AI on that isn't going to make it any worse, and will almost certainly improve it. I'll even give odds.

This Neanderthalic anti-AI attitude is strongest on tech-oriented sites, populated largely with people fearing (and in many cases, seeing) their jobs evaporating to AI tools. Which makes sense. Two centuries ago, the Luddites rioting and destroying the world's first computers (semi-automated punchcard Jaquard looms) were the weavers losing their jobs to them. Luckily, we managed to rein in those lunatics.

Interestingly enough, they raised the same arguments in the 1800s as are being raised today: machine-woven cloth was "slop(py)", poor quality, dehumanizing, and would eventually erode the quality of life. Yet somehow, we survived.
Saying a healthy dosage of skepticism is "neanderthalic" is arrogant, but let me make first clear things up. I strongly advise you to read about Luddites, because those were the first victims of industrial revolution - and a cautionary/sad story not something to be sneered at. (Btw. what does that mean actually? Neanderthal had a larger Brain-pan then Homo-Sapiens, and there is rather conjecture evidence for being stupid, quite the contrary, the Neanderthal had just a smoother ride with less environmental pressure to be a sleazy car salesman like Homo Sapiens had to be to survive, that being said, maybe you mixed up words, and you wanted to write Troglodyte instead?)

Also, you need to learn about cognitive processes and creativity itself: Human brain operates on very fuzzy and imprecise data, its basically a master of working with "good enough".
AI cant be creative or visionary about shit - since it operates in its current iteration on predictive, boolean logic rooted computation and it has a very hard time to come even close to
handle good enough like human brain does, its reasonable to say that it is questionable it will ever be (I for one am sure the current LLM approach never will be the scaffold where an emergent AI will arrive, neither AGI or anything like that. You can throw ever growing capacity for computation at it (brutforcing a prediction chain to be a simulacrum of a intelligence - thats what we do now), but by nature how linguistics work - it will ever happen. You have a very shallow idea how AI - as LLM + support architecture works, to make it simpler: its still a prediction chain, its neither creative or smart - it just happens at insane speeds (capacity for computation) and tricks you about being intelligent. LLMs do not calculate 2+2 = 4 but predict the highest score for that particualar problem. Thats it, its not magic or anything new.
Posted on Reply
#86
Unregistered
notoperableSaying a healthy dosage of skepticism is "neanderthalic" is arrogant
I wasn't replying to a 'healthy dose of skepticism', but the puerile belief that "AI is slop" and "life should be free" of it.
notoperableI strongly advise you to read about Luddites, because those were the first victims of industrial revolution [sic]
No, they were not "victims". No one has a guaranteed job for life, especially when they insist upon using obsolete methods to produce outmoded products. It's rather hypocritical as well, since I doubt you're driving a horse and buggy to work, simply to keep ostlers employed, or purposefully avoiding movies with CGI effects, to ensure 1980s-era toy model-makers and special-effects artists draw paychecks. The Industrial Revolution eliminated more human misery and toil than any other event in history. Respect it for that.
notoperableAI cant be creative or visionary about shit - since it operates in its current iteration on predictive, boolean logic rooted computation
Quite untrue. There are many nondeterministic neural networks in operation today. Nor do we need AI to be "visionary" to pilot a vehicle, execute a complex surgery, or perform forensic accounting. I'll also note that most of humanity shows little sign of being creative, yet still somehow manages to contribute to society.
notoperableYou have a very shallow idea how AI - as LLM + support architecture works.
Anyone who uses LLMs synonymously with AI has already disqualified themselves from the debate. LLMs are only a portion of generative AI, and all generative AI combined is a small fraction of all AI in use today. Don't attempt to play expert in a field you don't understand.
notoperableBtw. what does that mean actually? Neanderthal had a larger Brain-pan then Homo-Sapiens ... maybe you mixed up words, and you wanted to write Troglodyte instead?
The interesting thing about language is how context dependent it is. The words "awesome" and "awful" both literally mean "inspiring awe", yet those of ordinary intelligence understand the difference; adjectives like "hot" and "cool" vary by what's being described; and if I called you a cretin two hundred years ago, I'd simply be referring to you as a good Christian. I used the word I meant.
#87
notoperable
Tell me how have you inferred my lack of knowledge, I would gladly address that. You are falling into a context trap by self admission, so please don't lecture me on semantics or linguistics if you don't understand either

Anyone synonymously addressing LLM as AI or otherwise aside of you hehe, dude, precise language and reinstating assumptions is tricky. What you are saying is matter of interpretation
Posted on Reply
#88
lexluthermiester
EndymioThe "crap" is the inane labeling of all AI as "slop".
I didn't say all AI was slop. It has it's purpose, just not installed on my personal systems. What I said above is that it has NO place being forceably installed on consumers PCs and devices. If people want it, that should be their CHOICE. For those of us who do not want it, that is our CHOICE too. AI is being force feed to us all like we're cattle.

The key point there is choice. It shouldn't forced on us.
EndymioThis Neanderthalic anti-AI attitude is strongest on tech-oriented sites,
Nice insult. People who fail to see the forest for the trees and rush into forest fires deserve to get burned and roasted alive. It is seriously, woefully ignorant, foolish and short sighted to trust AI tech before it has been properly vetted and PROVEN to be safe. Those of us who understand this very serious potential for actual danger social and culturally, are the ones advocating for strict regulations and controls for tools that can and have already been seriously abused.

Keep silly insults aimed at those of us voice legitimate and meritful concerns out of discussions like this.
Endymiopopulated largely with people fearing (and in many cases, seeing) their jobs evaporating to AI tools. Which makes sense.
True, and that very seriously problem is one of many.

AI has presented a deeply disturbing state of affairs for a great many people. The progression of things as they are is not acceptable.
Posted on Reply
#89
Octopuss
julien1428I've got one of the Devkits installed and all of their claims are true! The performance is actually much higher than their claims.


Posted on Reply
#90
Unregistered
lexluthermiesterIt is seriously, woefully ignorant, foolish and short sighted to trust AI tech before it has been properly vetted and PROVEN to be safe.
Automobiles never proven to be safe, and in fact kill roughly one million people each and every year.

Do you drive an automobile? Yeah, I thought so.
lexluthermiesterThose of us [are] advocating for strict regulations and controls for tools that can and have already been seriously abused
There isn't a tool ever made by man -- from the first sharpened stone knife crafted by Homo habilis-- that hasn't been misused at some point. And AI will be no different ... and yet it will still accomplish ten thousand times as much good as bad.
lexluthermiesterWhat I said above is that it has NO place being forceably installed on consumers PCs and devices.
No. What you said was "get out of here with that crap" in response to the OP's very valid chiding of a journalist who finds the most "unique and refreshing" thing about Bolt is that it doesn't use the word AI. in it's marketing. Do you believe that your encountering the term in an article equates to having an AI chatbot forcibly installed on your PC?
Posted on Edit | Reply
#91
lexluthermiester
EndymioAutomobiles never proven to be safe, and in fact kill roughly one million people each and every year.

Do you drive an automobile? Yeah, I thought so.
There's a daft strawman argument.
EndymioThere isn't a tool ever made by man -- from the first sharpened stone knife crafted by Homo habilis-- that hasn't been misused at some point. And AI will be no different ... and yet it will still accomplish ten thousand times as much good as bad.
As context doesn't seem to be your strong-suit allow me to help you along: AI if used in even a slightly nefarious way can have a world-wide devastating effect if the powers that be are allowed to give AI the power no damned AI should EVER be allowed to have.

The potential for abuse is huge. This is NOT tin-hat stinky, it's legitimate concern for a set of tools so power that the abuse is something to be very worried about. And no, I'm not talking about Skynet crap. I'm talking about powerful and elite people using AI for abusing their power.
EndymioNo. What you said was "get out of here with that crap" in response to the OP's very valid chiding of a journalist who finds the most "unique and refreshing" thing about Bolt is that it doesn't use the word AI. in it's marketing. Do you believe that your encountering the term in an article equates to having an AI chatbot forcibly installed on your PC?
More taking what was said out of context...
Posted on Reply
#92
Unregistered
lexluthermiesterThere's a daft strawman argument ... The potential for abuse is huge. This is NOT tin-hat stinky, it's legitimate concern [sic] for a set of tools so power that the abuse is something to be very worried about. And no, I'm not talking about Skynet crap. I'm talking about powerful and elite people using AI for abusing their power.
Every major invention and advance in history comes with risks. "proven 100% safe before use" is an impossible standard that no technology can ever meet. Did you even ever stop to seriously consider what you're demanding? How could AI tools ever be "proven" to be beyond the ability for anyone to abuse them?

This isn't some sort of game, where you garner points by virtue-signaling on message boards. It's not. AI is already saving many lives in the medical field (it potentially saved mine a year ago). The sorts of rules of regulations you're advocating would have prevented that.
Posted on Edit | Reply
#93
lexluthermiester
EndymioEvery major invention and advance in history comes with risks.
While true, AI is different as it's potential for severe abuse is significant. This is NOT acceptable. Strict regulations need to be created and enforced.
EndymioAI is already saving many lives in the medical field (it potentially saved mine a year ago).
And that is what AI is useful for.
EndymioThe sorts of rules of regulations you're advocating would have prevented that.
No, it wouldn't.
Posted on Reply
#94
Unregistered
lexluthermiesterWhile true, AI is different as it's potential for severe abuse is significant. This is NOT acceptable. Strict regulations need to be created and enforced.
I guarantee you that any regulation you postulate can either be easily circumvented, or would be a 'cure' worse than the disease.

Go ahead: try it as a gedanken experiment on your part. Post the results here.
Posted on Edit | Reply
#95
lexluthermiester
EndymioI guarantee you that any regulation you postulate can either be easily circumvented, or would be a 'cure' worse than the disease.

Go ahead: try it as a gedanken experiment on your part. Post the results here.
Uh huh. And?
Posted on Reply
#96
njshah
lexluthermiesterGet outta here with that crap. There's nothing "caveman" about desiring life to be free of AI-slop.


I would tell where you can go, how to get there and what you can eat when you arrive, but the mods might prefer me not to. You get the idea though.
I don't talk to room temp chronically online nobodies , sorry. Cope while i use locally hosted LLMs to make money
Posted on Reply
#97
lexluthermiester
njshahI don't talk to room temp chronically online nobodies , sorry. Cope while i use locally hosted LLMs to make money
There's nothing wrong with that. AI has it's uses. My objection is having crammed up our collective asses..
Posted on Reply
#98
notoperable
EndymioEvery major invention and advance in history comes with risks. "proven 100% safe before use" is an impossible standard that no technology can ever meet. Did you even ever stop to seriously consider what you're demanding? How could AI tools ever be "proven" to be beyond the ability for anyone to abuse them?

This isn't some sort of game, where you garner points by virtue-signaling on message boards. It's not. AI is already saving many lives in the medical field (it potentially saved mine a year ago). The sorts of rules of regulations you're advocating would have prevented that.
Dude, every major invention has risks and there is no doubt about it, but you are like a child who sees a new toy and screams its the best toy ever without seeing the bigger picture, so let me broaden your perspective: in current geopolitical state of affairs, we are past cold war "detente" rope pull. ICBM pointed at each other as deterrent is nice, but we have a cyber theater that doesn't give a shit about that. We are in a world where countries offensive capability will be measured not in megaton yield but in terraflop computation. Don't you still get it? First thing that happens (and is happening) is the weaponization of AI with the difference that the risks of that is carried by everyone. Are you literally that ignorant that you don't understand such simple things? Nobody asked me or you if its fine to train models with your digital footprint - now everything is turning around value extraction from data and as a nice bonus to that - AI model training. If a soverign country wants to do AI, what do you think is their choice? There is no choice at all, whoever arrives at AGI first wipes out the competition - and the general public is the data drainage for that to happen.

I for one, were not asked about me being fine with training AI models for free based on my digital footprint by any random corporation - and nobody else has, basically each individual human being is being reduced to a excel sheet row, since AI models are not AI at all, those are "just" complex boolean logic based computations with insane energy inefficiency - and now everyone is being taxed for sponsoring this.
Secondly, the whole interwoven economy is so heavily neck deep into AI that when the bubble bursts - and it will, probably with a soft landing parachute sponsored by the goverment. What you don't get is, that, who do you think is gonna pay for that in tax money?
Posted on Reply
#99
Unregistered
notoperableDude, every major invention has risks and there is no doubt about it, but you are like a child who sees a new toy and screams its the best toy ever without seeing the bigger picture, so let me broaden your perspective: in current geopolitical state of affairs, we are past cold war "detente" rope pull. ICBM pointed at each other as deterrent is nice, but we have a cyber theater that doesn't give a shit about that. We are in a world where countries offensive capability will be measured not in megaton yield but in terraflop computation. Don't you still get it? .... (snip)
While I appreciate being comradely addressed as a fellow surfer dude, I'm surprised you fail to see the logical consequences of your reasoning. Your premise is that "whoever arrives at AGI first wipes out the competition" ... yet you want Western nations to halt AI research essentially forever -- since we'll never be able to prove AI is indisputably 100% safe -- while Russia, China, and others continue and -- as you admit -- then will promptly then proceed to wipe us out.
Posted on Edit | Reply
#100
notoperable
EndymioWhile I appreciate being comradely addressed as a fellow surfer dude, I'm surprised you fail to see the logical consequences of your reasoning. Your premise is that "whoever arrives at AGI first wipes out the competition" ... yet you want Western nations to halt AI research essentially forever -- since we'll never be able to prove AI is indisputably 100% safe -- while Russia, China, and others continue and -- as you admit -- then will promptly then proceed to wipe us out.
I dont play but games, you need to address questions with answers and not with another round of questions or assumptions from how/where sphere.

I wonder though - where did you get the impression I want "to halt any AI research"? We are long past the point of no return, since you cant close the pandora's box or put the genie back into the bottle - it seems you missed that along the way.
Posted on Reply
Add your own comment
Nov 18th, 2026 04:53 CST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts